What makes this newsletter different?
Two Lanterns Advisory was created to provide the best political risk analysis - as well as the best analysis of political risk.
The Inside Political Risk newsletter and podcast can be found at twolanterns.substack.com.
Below are past issues, stored in a more easily searchable format.
How do you predict Jared Kushner?
A new risk to be analyzed
Chris Oates Apr 31
Today’s newsletter grapples with how to analyze a scenario posed by this Politico article.
Dozens of Trump administration officials have trooped to the White House podium over the last two months to brief the public on their effort to combat coronavirus, but one person who hasn't -- Jared Kushner -- has emerged as perhaps the most pivotal figure in the national fight against the fast-growing pandemic.
Jared Kushner is not qualified to be anywhere near this crisis.
He is not a public health expert. He is not an emergency management expert. He is the son of a real estate mogul whose personal track record in real estate was not great.
As you may recall, the Kushner family owns a troubled skyscraper in Manhattan located at 666 Fifth Avenue. Jared Kushner bought the building in 2007 for a record-setting $1.8 billion, a lofty price that marked the young developer's entry into high-profile New York dealmaking while also threatening to strain his family's personal finances should the transaction not work out.
It didn't work out.
How should the political risk industry approach a dilettante staffer who occasionally enters the picture but without the knowledge of what to do when there?
Here's Jared Kushner going for the world record of most meaningless corporate buzzwords used in a single one-minute video clip
April 2nd 2020
3,197 Retweets10,710 Likes
Kushner’s portfolio
Much has been made of Kushner’s wide-ranging remit in the White House.
He has, at various times, been involved in government reform, criminal justice, immigration, and the Middle East Peace Plan. Political commentators and analysts have noted that Kushner has failed to be effective in any of these roles.
Tufts professor Dan Drezner catalogued these results in this 180-tweet thread, the latest of which is here.
I’ll believe that Jared Kushner is growing into his role as a White House staffer when he stops duplicating the efforts of others. politico.com/news/2020/04/0…
April 2nd 2020
11 Retweets43 Likes
Forecasting COVID-19 responses
The political risk world is now almost entirely dedicated to COVID-19. How bad will it be and what happens after are the two questions most important to the industry.
Typically, risk analysts begin with a scenario in which governments act according to best practices and evidence. But what happens when a state ignores evidence and endangers its citizens?
Georgia Gov. Brian Kemp said Wednesday he had just learned coronavirus is transmitted asymptomatically: "We didn’t know that until the last 24 hours." Fauci on Jan. 31: "Now, we know from a recent report from Germany that that is absolutely the case." Analysis | Georgia Gov. Brian Kemp, who resisted strict coronavirus measures, says he just learned it transmits asymptomatically“Individuals could have been infecting people before they ever felt bad, but we didn’t know that until the last 24 hours,” Kemp said. That’s remarkably untrue.washingtonpost.com
April 2nd 2020
574 Retweets1,390 Likes
Kushner as a political risk
It seems that we should treat Kushner not just as a presidential son-in-law, but as a political risk in his own right and as a stand-in for a wider political risk.
Francis Fukuyama tried to address the problem of “What is Governance” in 2013. In the abstract to his article, he writes:
there is very little agreement on what constitutes high‐quality government. The commentary suggests four approaches: (1) procedural measures, such as the Weberian criteria of bureaucratic modernity; (2) capacity measures, which include both resources and degree of professionalization; (3) output measures; and (4) measures of bureaucratic autonomy.
Jared Kushner lacks the experience needed for the jobs he is given, but is unlikely to be reassigned to a more fitting role. Kushner’s involvement in these high-profile issues therefore degrades Fukuyama’s first, second, and fourth aspects of governance.
There is also a non-zero chance that he enters the picture and duplicates the work of the existing bureaucracy and political appointees. We need to account for this.
As risk analysts, we need to include these once unimaginable factors, because these outliers are now the norm.
Towards a general theory of Kushnerism
As the reports from Governor Brian Kemp show, this issue is not restricted to Jared Kushner.
We have seen examples of good governance and bad from both Democratic and Republican officials. Andrew Cuomo, Jay Inslee, Charlie Baker, and Mike DeWine have acted in timely and appropriate manners. Bill de Blasio and Ron DeSantis, on the other hand, were both too slow to act.
However, it does seem clear that there is a tendency that active leadership reliant on experienced experts produces better governance. If an administration has an unqualified person in power, or if there’s a pattern of listening to people who are unqualified, then we must include that in our forecasts.
In the same way that a sportswriter thinks about a star pitcher’s injury history when predicting who which teams will make the playoffs, we should account for the chance that a state is “injured” by someone unqualified taking the lead.
As the political risk world takes stock of what COVID-19 has shown us after the pandemic passes, this is a lesson that we should learn.
Even the best prepared government is not ready for a disaster if it is not allowed to do its job.
----------
Thanks for reading.
Chris
Rally round the flag, but for how long?
Parsing approval bumps
Chris Oates Apr 1
President Donald Trump’s approval ratings at “at or near historic highs”. But how long will this current bump last and how will it affect the election in November?
According to CNN’s Chris Cillizza:
There's no question that President Trump is in the midst of his best polling moment in the entirety of his time in office right now.
A Washington Post/ABC poll released over the weekend showed Trump at 49% approval to 47% disapproval, the first time in that poll that those approving of his performance in office outstripped those disapproving.
Other surveys by Gallup, Fox News and Monmouth University all confirm the Post's findings: Trump at or near historic highs for job approval.
Luckily, we can turn to existing political science literature to help us predict the next president.
Is the surge real?
There has been a non-negligible increase in Trump’s approval rating. According to FiveThirtyEight’s poll tracker, he is up to 45% approval, a level he hasn’t seen since his first week in office.
We have to take the evidence that there has been a rally-round-the-flag effect, in which a leader gets higher approval ratings during a crisis.
Is the surge significant?
However, it’s important to keep this in perspective. A three point bump in the polls is not a rocketship of popularity.
Morning Consult’s tracker of major country democratic leaders shows Trump as experiencing a small trend upwards. The UK’s Boris Johnson, Germany’s Angela Merkel, France’s Emmanuel Macron, Canada’s Justin Trudeau, and Australia’s Scott Morrison, however, are shooting past him with 20-point gains.
Interestingly, Mexico’s Andres Manuel Lopez Obrador and Brazil’s Jair Bolsonaro have seen their approval ratings drop as they dismiss COVID-19 as a major public health threat.
Going back in US history, we see rally-round-the-flag effects between 15 and 50 points. The below is from “Anatomy of a Rally Effect: George W. Bush and the War on Terrorism” by Marc J. Hetherington and Michael Nelson.
Put into global and historical context, Trump’s boost actually hasn’t been that big.
How long will it last?
This is the most important question for forecasting the election.
Will this rise in approval ratings last until November?
There is evidence that it won’t. George H. W. Bush’s boost during the Gulf War had started in January and faded by October - and that was about a war that the United States won.
There’s still no certainty about the effect of a mounting death toll, stories about federal government inaction, and a major recession on presidential approval ratings.
A better example might be George W. Bush after the Iraq War. As Gallup shows, he had a 14-point spike at the time of the invasion. That soon tumbled and began to slide as the public saw how badly managed the war was. The ratings only somewhat recovered during the election season, when partisan consolidation gave him a lift.
What to expect
The last point makes me pause in predicting what will happen.
Approval bumps don’t last. Bush won re-election, but his father didn’t. Jimmy Carter was at 61% at the start of the Iranian hostage crisis and then lost a year later.
But partisanship is a strong factor - perhaps the strongest in politics today. It stabilized W. Bush at the end of 2004 even as Iraq was falling apart and it could keep Trump within striking distance of an Electoral College victory no matter what happens.
So while we don’t want to read too much into the rallying effect we see now, we can’t dismiss it either. It’s been a very long March and we still have six months ahead of us before Election Day.
----------
Thanks for reading.
Chris
How we got here
The story of why the US fell behind on COVID-19
Chris Oates Mar 27
To end the week, I’m linking to some long-reads that look at how we got here in the novel coronavirus pandemic.
We have still no idea how long this pandemic will last. But we have some evidence that the US is on a path to experience one of the worst outbreaks.
Whatever happens, there will almost surely be a question of accountability.
Why, if the WHO announced this a major health emergency on January 30, was the United States caught flat-footed in March?
The answer may be the key force affecting politics in the summer and fall. Luckily for us, some great reporting has already started to fill in the blanks. There are three key points that they collectively make.
The government knew this could happen
It’s crucial to understand that we knew this could happen.
A virus emerging in December 2019 in a market in Wuhan, China may have been unpredictable. But we knew there was a possibility of a future, flu-like virus.
As Politico reported, members of Barack Obama’s outgoing administration told members of Trump’s incoming administration about catastrophic events they may have to deal with.
Before Trump’s inauguration, a warning: ‘The worst influenza pandemic since 1918’ - Politico
The briefing was intended to hammer home a new, terrifying reality facing the Trump administration, and the incoming president’s responsibility to protect Americans amid a crisis. But unlike the coronavirus pandemic currently ravaging the globe, this 2017 crisis didn’t really happen — it was among a handful of scenarios presented to Trump’s top aides as part of a legally required transition exercise with members of the outgoing administration of Barack Obama.
Even though it was only an exercise, they predicted what would be the critical shortages facing the country.
“There’s no briefing that can prepare you for a worldwide pandemic,” according to Sean Spicer, who was there.
But this looked like it would have done a pretty good job of it.
The government had a plan for this
As Politico later reported, not only was there a briefing. But there was a literal playbook for a pandemic.
Trump team failed to follow NSC’s pandemic playbook - Politico
“Is there sufficient personal protective equipment for healthcare workers who are providing medical care?” the playbook instructs its readers, as one early decision that officials should address when facing a potential pandemic. “If YES: What are the triggers to signal exhaustion of supplies? Are additional supplies available? If NO: Should the Strategic National Stockpile release PPE to states?”
The strategies are among hundreds of tactics and key policy decisions laid out in a 69-page National Security Council playbook on fighting pandemics, which POLITICO is detailing for the first time.
This playbook is thorough: 69 pages with detailed tables, phases, and trigger points.
Perhaps more remarkable is that it was leaked to Politico in its entirety.
While I don’t know how the reporters acquired it, the obvious source is someone in the government who wants people to know that the federal government could have done better.
You must have a considerable motivation to leak an entire pandemic government document and risk being discovered.
The government did not execute that plan
All of this adds up to the conclusion reached by Micah Zenko in Foreign Policy.
The Coronavirus Is the Worst Intelligence Failure in U.S. History - Foreign Policy
In short, the Trump administration forced a catastrophic strategic surprise onto the American people. But unlike past strategic surprises—Pearl Harbor, the Iranian revolution of 1979, or especially 9/11—the current one was brought about by unprecedented indifference, even willful negligence. Whereas, for example, the 9/11 Commission Report assigned blame for the al Qaeda attacks on the administrations of presidents Ronald Reagan through George W. Bush, the unfolding coronavirus crisis is overwhelmingly the sole responsibility of the current White House…
The White House detachment and nonchalance during the early stages of the coronavirus outbreak will be among the most costly decisions of any modern presidency. These officials were presented with a clear progression of warnings and crucial decision points far enough in advance that the country could have been far better prepared. But the way that they squandered the gifts of foresight and time should never be forgotten, nor should the reason they were squandered: Trump was initially wrong, so his inner circle promoted that wrongness rhetorically and with inadequate policies for far too long, and even today. Americans will now pay the price for decades.
This is, to me, the unforgivable nature of the COVID-19 (lack of) response.
This failure has real consequences, could have been avoided, and in other countries, has been avoided. South Korea and Germany stand as the current exemplars at containment and mitigation.
This failure has real consequences, could have been avoided, and in other countries, has been avoided. South Korea and Germany stand as the current exemplars at containment and mitigation.
Bianna Golodryga @biannagolodryga
South Korea and the US reported their first cases of COVID-19 on the same day in January. South Korea, through rapid mobilization, testing and isolation, appears to avoided a disaster. Sadly, the same cannot be said about the US. https://t.co/jMZVyOcYok
Ken Dilanian @KenDilanianNBC
It’s come to this. “President Trump asked if South Korea can provide the U.S. with medical equipment to respond to COVID-19 and President Moon replied saying, ‘I will offer as much when there are surplus to share.’”
March 24th 2020
793 Retweets1,708 Likes
Why Germany’s coronavirus death rate is so much lower than other countries’ rates - Washington Post
The biggest reason for the difference, infectious disease experts say, is Germany’s work in the early days of its outbreak to track, test and contain infection clusters. That means Germany has a truer picture of the size of its outbreak than places that test only the obviously symptomatic, most seriously ill or highest-risk patients.
It remains to be seen how the US political system will respond when we start finally to realize the true cost of the failures. It is a debate that won’t soon be going away.
----------
Thanks for reading.
Chris
Waiting for unemployment claims
Orban, Herbert Hoover, joblessness, and a new dataset
Chris Oates Mar 26
COVID-19 seems to be happening on two tracks.
On the first, the response is escalating across the country as the virus spreads to more areas.
@GovernorLittle issues statewide stay-home order, signs extreme emergency declaration
March 25th 2020
121 Retweets86 Likes
On the second track, some people (mostly connected to the right-wing or sections of the business community) are arguing that Americans have a patriotic duty to sacrifice themselves for the sake of the economy (even thought that wouldn’t help the economy).
Dick Kovacevich, the former Chairman and CEO of Wells Fargo, wants people to return to work. “Some of them will get sick, some may even die, I don’t know.” bloomberg.com/news/articles/…
March 25th 2020
2,481 Retweets4,052 Likes
While they continue to debate whether to listen to public health experts, the numbers keep going up. Here are the latest daily change rates for confirmed cases, courtesy of Joel Paul (his Venmo is jpaul156 for anyone who wants to tip him for this data).
Politics
Hungary’s prime minister is using the virus to make an authoritarian power grab - Washington Post
Covid-19 is about to claim a new victim: Hungary’s democracy.
The country’s parliament is set to adopt a new law that will give the government of Prime Minister Viktor Orban a legal mandate to rule by decree, without any sunset clause and without parliamentary oversight. The government initially sought to fast-track the legislation and adopt it already on March 24, but it lacked the supermajority needed to accelerate the proceedings. The party, however, does not lack the votes to ensure that the legislation is passed through the normal legislative process a few days later.
This power grab highlights one of the dangers of a populist or quasi-authoritarian. He may seem tolerable, even constrained, by a country’s institutions. But when a crisis hits, he will be in a position to quickly take advantage of the situation.
It’s a reminder that the United States was semi-lucky to have Herbert Hoover in charge during the start of the Great Depression. Though his economic policies were a failure, he preserved the integrity of American democracy when other countries turned towards authoritarianism. From David Frum’s review of a biography of Hoover:
Few if any Americans have dedicated more of their lives to the service of others than Hoover. A wealthy man by age 40, he turned his back on opportunities to earn more—and dissipated much of what he had gained—to devote himself to humanitarian work. As a private citizen, he organized food relief for German-occupied Belgium during the First World War. He undertook an even more ambitious task of rescue in Russia and Ukraine during the civil war that followed the Bolshevik seizure of power. Then, as commerce secretary, he took charge of the response to the devastating flooding of the Mississippi Valley in 1927: floods that drove hundreds of thousands from their homes to temporary camps and left the region in economic shambles.
The question remains which other countries will see authoritarians use this crisis for their own ends in the coming months.
Economy
Jobless claims expected to spike to a record-breaking number in the millions - CNBC
We do not yet know how many people have lost their jobs over the last month.
We know that it is a lot. The unemployment claims so far coming out of states look like a month of job losses during a bad recession compressed into a single week.
Here are WV's 5 worst weeks for initial unemployment filings 1/5/1991: 5,224 1/3/1987: 5,295 1/4/1992: 5,310 1/8/1994: 5,428 1/9/2010: 5,445 1/2/1999: 5,877 3/2020: 28,000
March 24th 2020
140 Retweets277 Likes
But where it all ends up is unclear. Look at how far apart major banks are projecting for the national numbers - anywhere from 1.5 million to 4 million.
An upper limit would make a return to a normal economy more difficult. With more people drawing down their savings over the next couple months, there’s less pent up demand in the economy when it reopens. As one analyst says:
Certainly 10% unemployment is a forecast that has a great chance of becoming fact. But I still don’t know what it means because this is still a coronavirus recession. There isn’t a housing bubble that burst. There isn’t a stock market bubble that burst, signaling underlying problems… It’s really the economy has caught this deadly virus cold, and we’re waiting for the symptoms to subsist.
Analysis
Oxford COVID-19 Government Response Tracker - Oxford Blavatnik School of Government
Governments are taking a wide range of measures in response to the COVID-19 outbreak. The Oxford COVID-19 Government Response Tracker (OxCGRT) aims to record these unfolding responses in a rigorous, consistent way across countries and across time.
Systematic information on which governments have taken which measures, and when, can help decision-makers and citizens understand the robustness of governmental responses in a consistent way, aiding efforts to fight the pandemic. The OxCGRT collects information on several different common policy responses governments have taken, scores the stringency of such measures, and aggregates these scores into a common Stringency Index.
This is a great source of international data to identify what the global response to COVID-19 is. We can see that the government responses roughly track the number of confirmed cases, though with quite a bit of variation - China has now dropped its stringency requirements to 0 and Croatia reached the maximum before it had a confirmed death.
This data is able to show just how rapidly China increased their measures against the virus and how, worringly, Italy’s gradual escalation may have contributed to its failure to control the virus.
Expect to see more from this data set as I get a chance to explore it in depth.
----------
Thanks for reading.
Chris
Washington starting to unravel
Making sense of Lord of the Flies economics
Chris Oates Mar 24
The past weekend has seen a dramatic escalation in COVID-19 cases in New York City, which is now seeing the largest outbreak in the US.
The tracker from Microsoft shows what exponential growth really looks like (and also the result of being able to finally test suspected cases).
Initially, our national politicians seemed to be listening to the health experts. More recently, however, President Trump has swapped his rallies for press conferences and the Republicans in Congress have worked to pass a bill that bails out their donors and GOP-friendly corporations.
Economy
Global Economic Effects of COVID-19: In Brief - Congressional Research Service
Thus, unlike the 2008 crisis response, which involved liquidity and solvency-related policy measures to get people spending again, the current crisis is related to the amount of operating capital in the economy. While larger firms may have sufficient capital to wait out a crisis, many aspects of the economy (such as restaurants or retail operations) operate on very tight margins and would likely not be able to pay employees after closures lasting more than a few days. Many people will also need to balance child care and work during quarantine or social distancing measures. During this type of crisis, while monetary policy measures play a part -- and the Federal Reserve has once again cut rates to near zero -- they cannot compensate for the physical interaction that the global economy is dependent upon. As a result, fiscal stimulus will likely play a relatively larger role in this crisis in order to prevent personal and corporate bankruptcies during the peak crisis period. Efforts to coordinate U.S. and foreign economic policy measures will also have an important role in mitigating the scale and length of any global economic downtown.
The Congressional Research Service report on the economic impacts of COVID-19 will likely be out of date by the time you read it, but it’s a useful primer on the main issues at hand.
Unlike a purely financial crisis in which independent central banks can play a leading role, this economic crisis comes from the unusual problem of people not working.
When a supply shock combines with a demand shock, we need fiscal policy, which leads us to…
Politics
Last week, Washington appeared to have woken up to COVID-19. President Trump was having press conferences with medical experts and Congress was working on a bipartisan plan.
Yesterday, however, we started seeing politicians and commentators argue about the severity of the new guidelines. They worried that the risk of economic shutdown would be more detrimental than the risk of the virus. As President Trump tweeted, we can’t have the “CURE BE WORSE THAN THE PROBLEM ITSELF”.
This went against the trend of more lockdowns and restrictions to deal with the rise in cases. This idea of economic interests competing with public health concerns seemed like it was out of nowhere, but it wasn’t.
As we remember from this March 6 clip - let millions die to save the equity markets has been building on the American right for a while.
Josh Marshall @joshtpm
CNBC's Rick Santelli says it might be better to let everyone get COVID-19 at once & get it over with so it won't put so much stress on global equity markets. (Note this wld likely mean **millions** of deaths and the speed of spread actually makes a vast difference in mortaliity.) https://t.co/Lx3hTUnqo7
March 23rd 2020
303 Retweets633 Likes
To be clear, this advice is wrong.
Health experts all agree that isolations and lock-downs are the best way to fight the virus. So why is something so wrong-headed popping up now as the crisis escalates?
The best explanation is captured in this article from Friday.
U.S. intelligence reports from January and February warned about a likely pandemic - Washington Post
U.S. intelligence agencies were issuing ominous, classified warnings in January and February about the global danger posed by the coronavirus while President Trump and lawmakers played down the threat and failed to take action that might have slowed the spread of the pathogen, according to U.S. officials familiar with spy agency reporting.
So why would a bad idea be cropping up now? Because we’ve seen key people in the federal government make bad decisions already.
We should lower our expectations for political decision-making.
Typically, we plan for rational actors in political risk analysis. In this case, however, we have to adjust to the new normal that our actors are not following rational medical advice. Thus we have to predict for the unpredictable.
Business
Local colleges to refund upwards of $670M in unused student room and board - Boston Business Journal
Boston-area colleges are now faced with an unexpected financial obstacle — losing millions of dollars in revenue from refunds on unused room and board fees.
We often forget that academia is technically an export industry. In university towns, the area “sells” its academic services to people who often come from elsewhere. A student from Los Angeles who attends Harvard is, in economic terms, $50,000 per year in services sold from Massachusetts to California.
We do not yet know the full impact of universities nationwide shutting down for the semester. The above number alone would represent 0.1% of the Massachusetts GDP. Of course, some of that will be captured by in-state students. But the state overall is estimated to gain 22% more students than it produces, many tens of thousands of students who will not be spending money in the state.
The same is true for any college town, and we can expect to see their economies hard hit in the next few months.
Sports
One positive story comes out of East London.
Ultimate QuaranTeam FIFA tournament schedule & favourites revealed - Squawka
Leyton Orient, a fourth-division soccer team in England, decided to fill the time of their postponed season by hosting an online FIFA tournament.
They received a huge response and 128 professional clubs from across the world decided to join in.
What’s truly amazing, though, is how quickly this fun little idea turned into something more real. From Squawka:
With some of the clubs opting to use their official eSports teams or professional FIFA players to try and win the tournament, it will come as no surprise to see Manchester City joint-favourites at 8/1… Man City are joined at 8/1 by AS Roma.
This is the economics of globalization at work.
There are low barriers to entry. Leyton Orient are only a couple of years away from near bankruptcy but have created something that has captured the attention of some of the richest teams in the world.
But commerce exists on a platform with a quasi-monopolistic position: the FIFA video game series. Like Facebook, Ebay, Uber, AirBnB, Google, etc, FIFA’s dominant position in the football market makes it the natural choice for all these teams to compete against each other, which enhances its position as the best game, which makes it more likely to be used in the future, etc.
We therefore see the heartwarming story of the plucky underdog (“Leyton Orient are challenging Man City!”) within a broader story of a megacorporation entrenching itself.
This episode is ripe to be the subject of a business school case, or at least the source of some distraction for fans under lockdown.
----------
Thanks for reading.
Chris
COVID-19 and political risk
New daily links to some of the best articles on the pandemic and its consequences
Chris Oates Mar 20
I don’t often recommend Tom Friedman’s work, but these are crazy times.
The title sums up the point of his latest column: “Our New Historical Divide: B.C. and A.C. — the World Before Corona and the World After.”
Before the coronavirus crisis hit, I was toying with writing a book about 21st-century political parties, but in light of this global epidemic it’s obvious that whatever nonfiction book you’re working on now, put it down. There is the world B.C. — Before Corona — and the world A.C. — After Corona. We have not even begun to fully grasp what the A.C. world will look like, but here are some trends I’m watching.
In typical Friedman fashion, there are overdrawn metaphors and too many quotes from CEOs, but his basic point is clear and well-founded. COVID-19 is going to have an enormous impact on the world in ways that we don’t yet know.
An infected bat was sold in a market in December and by March the volatility index almost doubled what its highs from the financial crisis.
I started this newsletter as a way of exploring different aspects of political risk that often go missed as the industry focuses on the big news story of the day. I wanted to focus on depth in a field often focused on breadth.
Now, that calculation has flipped. There is, at least for now, one gigantic news story but its implications go far and wide.
Things are moving faster with deeper consequences that we may be imagining. To grapple with this, I’ve been consuming a news from different areas every day. This newsletter will carry some of the best of what I read on a near daily basis.
I will still write longer pieces from time to time and will be producing podcasts on different aspects of political risk as well. But most of this newsletter’s issues will be focused on a curated look at the changing world.
I will try to focus on the non-health aspects of COVID-19: the political, economic, and business fallout of social distancing and closed borders. There are better places to get public health information and I’m not an epidemiologist - so I’ll wash my hands of that part of the crisis.
With that, here’s the first day’s set of stories.
Politics
Intelligence Chairman Raised Virus Alarms Weeks Ago, Secret Recording Shows - NPR
The chairman of the Senate Intelligence Committee warned a small group of well-connected constituents three weeks ago to prepare for dire economic and societal effects of the coronavirus, according to a secret recording obtained by NPR.
This is the sort of story that could have huge political ramifications - perhaps more than the Mueller Report.
On Feb 27, President Trump was publicly dismissing the impact of COVID-19. At a lunch with business leaders from his home state (who paid to be part of the group hosting the lunch), North Carolina Sen. Richard Burr was saying that the illness was more like the 1918 pandemic.
This matters not just for the content but because of the fears it confirms about how the government has handled the crisis:
Trump was going against what the rest of the government knew.
Burr, the chair of the Senate Intelligence Committee, knew what was happening. He predicted travel restrictions to Europe, school closures, and the military being called to help.
Burr spoke to well-connected and wealthy donors but wasn’t willing to raise the alarm publicly.
This will be a key question for every candidate before November. To borrow from Howard Baker, what did everyone know and when did they know it?
The fact that Burr may have illegally profited by selling stocks before the market crashed will also keep this story alive.
Economy
So Far, Stimulus Is Leaving Mass Transit Behind - Bloomberg
Mass-transit systems across the country are taking a violent financial hit from Covid-19. That’s because these systems are largely dependent on fees from riders which are, of course, way down.
This is not the biggest near-term economic story. But it could be that in a decade or two that we look to this moment as the time when a huge opportunity was missed.
Because of the particular regulations and governance structures of mass transit systems, they are dependent on ridership and local taxes - unlike highways. If this crisis turns into a deep recession, that revenue will dry up.
The federal government could make this the moment to invest in transit - or states and transit authorities could take advantage of ultra-low interest rates to issue bonds.
If they do, it could lay the groundwork for modern mass transit. If they don’t, our infrastructure will continue to decline.
Business
Responding to coronavirus: The minimum viable nerve center - McKinsey
The integrated nerve center is not a formulaic panacea. It is, rather, an efficient means of coordinating an organization’s active response to a major crisis. It is endowed with enterprise-wide authority and enables leaders and experts to test approaches quickly, preserve and deepen the most effective solutions, and move on ahead of the changing environment.
As with any management consultant writing, the above would make Hemingway jump blindfolded into a bullring.
But it describes a good practice. A business is a process for taking people’s time and material resources and turning it into profits. In a time of uncertainty, that process is disrupted.
As strange as it sounds to think of a business needing a “nerve center,” a group that is tasked with finding out what new problems are arising, what needs to be done to fix them, and doing them is of critical importance.
Every business should be willing to rethink the systems they have in place and looking for where a new turn would take them into trouble. It’s good practice in normal times, and necessary practice now.
Sports
CWI offer to host England Test series in Caribbean - ESPNCricInfo
One of the first signals to the public that this was serious was when the NBA suspended its league.
Mark Cuban reacts moments after the NBA season was suspended.
March 12th 2020
13,316 Retweets71,433 Likes
This story shows what might be happening after COVID.
The West Indies cricket board has offered to host the England’s home matches this summer. This could be both the England-West Indies series, as well as the England-Pakistan series.
While there are COVID-19 cases in the Caribbean, the virus would appear, at this stage, to have taken hold far more in the UK. There have, for example, been only two reported cases in each of Barbados and St Lucia and one in Antigua.
Since sports now get most of their revenue from TV rights, rather than in-person ticket sales, the finances wouldn’t take much of a hit. In fact, it could even be beneficial to those with money on the line: the time difference means that each match would go into UK primetime.
This doesn’t mean that this plan will happen. But it’s a sign that sports leagues are willing to be flexible to ensure that their core product - professional athletes competing in front of cameras with satellite feeds - can happen.
I mean, if the NBA came back with just a live stream from a practice facility, wouldn’t ESPN broadcast it?
Episode 3: Political Risk Insurance
Chris OatesMar 18
Exclude from Top
1x1.25x1.5x1.75x2x
0:00-26:37
I’m joined by Price Lowenstein, President and CEO of Sovereign Risk Insurance, to talk about the political risk insurance industry.
The history of the industry, how it analyzes risk, and what’s going to happen if COVID-19 and the oil crash send emerging markets tumbling are all discussed.
The Democratic Primary: What just happened?
Breaking down the rapid end to a long campaign
Chris Oates Mar 11
Based on the results of last night, the Democratic nomination campaign is all but wrapped up. Barring a crazy turn of events over the next couple of months, Joe Biden will become the Democratic candidate for president in Milwaukee on July 16.
It might feel a bit like whiplash to anyone who follows politics.
For more than a year, this primary has dragged on.
There were 23 candidates who participated in at least one of the ten debates. Remember Jay Inslee? Eric Swallwell? For a moment, they were part of this process.
Then there came the actual voting.
In the space of only a month, we had a series of events that could have been their own 2-part episode of the West Wing:
Feb 3: Iowa had a breakdown of counting, but Pete Buttigieg seemed to have a boomlet.
Feb 11: Amy Klobuchar had a breakout debate and a comeback in New Hampshire.
Feb 19: Mike Bloomberg stepped out from behind his ads and got destroyed by Elizabeth Warren.
Feb 22: Bernie Sanders won big in Nevada and showed a broad coalition.
Feb 29: Joe Biden won bigger in South Carolina.
Mar 1: Buttigieg dropped out.
Mar 2: Klobuchar dropped out.
Mar 3: Biden won big on Super Tuesday.
Mar 4: Bloomberg dropped out.
Mar 5: Warren dropped out.
FiveThirtyEight has dubbed Biden the presumptive nominee, based on one of the biggest swings I’ve ever seen in its model.
While it’s not officially over, we’re at the point when we can start looking back over the race and identify how we moved from a dragged out affair that would never end to a rapid consolidation. With hindsight barely opened to us (and so the analysis still tentative), how did we get here?
Sanders found his ceiling
One of the questions lingering from the 2016 election was whether Sanders’ votes in that race were cast for him or against Hillary Clinton.
He certainly had huge support nationwide. But there’s a big difference between receiving 43% of the vote because people like you (and will continue to like you in the next election) and because people wanted to protest against the other candidate.
It seems that we now have an answer. Sanders spent most of 2019 below 20% in polls and his surge after Iowa has only brought him towards 35%.
Sanders supporters argued that a fractured field as being why his vote share was lower than in 2016. For example, in Iowa in 2016 he lost with 49.6% of the vote, and in 2020 he led the popular vote with 26.5%.
But that doesn’t explain why Biden has been able to quickly rocket past 50% when the field coalesced while Sanders has remained well short of his 2016 numbers.
If anything, the fractured field made Sanders appear stronger than he truly was because his inability to pick up votes from other candidates was masked. His ceiling was only revealed once the campaign fell to only a few candidates, but by then it was too late to change strategies.
The party didn’t decide
One of the major theories of primary elections is The Party Decides. Based on this academic work, it argues that the “invisible primary” winnows candidates before voting even starts, creating frontrunners through endorsements and overt support.
This is now being credited as the reason why Biden in winning. As Jonathan Chait in New York writes:
“And yet the last few days of the Democratic primary have given us some evidence that the argument still has some power. In South Carolina, Representative James Clyburn’s endorsement helped power Joe Biden to a blowout victory. And now Democratic officials are joining en masse to support Biden. Two of his mainstream opponents, Amy Klobuchar and Pete Buttigieg, have dropped out and plan to endorse him.”
This is a good case for party leaders having a major impact in South Carolina and Super Tuesday. But Biden had been leading the polls from the start of his candidacy until New Hampshire. The party may have helped him recover, but it didn’t give him his initial frontrunner status.
On this front, it’s the dog that didn’t bark that’s most interesting. Despite there being a number of plausible top tier candidates close to the party establishment, there was no consolidation around a candidate. The party - for almost the entire time that it mattered - couldn’t make a decision.
Biden was a former Vice President and Senator, but was perceived as so weak that moderates encouraged Deval Patrick to get in the race late.
Kamala Harris was seen as a candidate who could reach the entirety of the party, and surged in the betting markets after the first debate. But that surge didn’t build into momentum. They party didn’t embrace her, and she had to end her campaign before the first votes.
Elizabeth Warren was the next candidate to surge in the betting markets. By September, she was leading in US and US markets - implying that traders saw something in her candidacy that wasn’t reflective in her polling position. But she also failed to get the party to rally around her and faded.
In other years, you could also plausibly see the major elements of the party coalesce around Booker, Castro, O’Rourke, or Klobuchar. Yet none received the boost and lane-clearing that would have come with that.
Biden, propped up by his consistent lead in the polls and high name recognition, won more of a war of attrition than saw the party pick his path for him.
TV ads can’t win an election
We ran a natural political science experiment this cycle. According to this infographic from the New York Times, two billionaires combined for the vast majority of the TV ads in Super Tuesday states.
The result was that Steyer dropped out before Super Tuesday and Bloomberg the day after. This is not to say that money in politics doesn’t matter, but it showed the limits of how much it matters.
You win by running up the score
Perhaps the most basic lesson of the Democratic primary from 2008 and 2016 is still true. Because Democrats allocate delegates on a proportional basis (above a 15% threshold), it is incredibly difficult for a candidate to make a comeback.
In 2008, Hillary Clinton pulled off a big win in Ohio. She said that night “Ohio has written a new chapter in this campaign and we’re just getting started.”
But she only won Ohio 53% to 45%, giving her a net of 7 more delegates than Obama. The next week, Obama won Mississippi 61-37, getting back those 7 delegates. Even though Mississippi had less than a quarter of the delegates of Ohio, Obama ran up the score and maintained his lead.
In both 2008 and 2016, the eventual winner racked up a small but significant lead in delegates early and then ran about even through the end of the race.
This is the path that Biden appears to be treading as well. Super Tuesday dwarfed the delegates on offer in the early states, so that Sanders’ early win was negated by Biden’s delegate haul. Although not all delegates have been awarded from last night, it seems that the same pattern is happening there. Biden’s win in Mississippi will get him as many net delegates as Sanders’ win in California.
Concluding thoughts… for now
The primary is not over, and there will be plenty more for analysts to digest over the next four years as the race is studied further.
The early indications are a warning against always thinking that this time is different.
A former vice president with high approval ratings from core elements of the party. A candidate whose highest office was held when in the other party did not perform well. A candidate who wasn’t demonstrating a wide coalition failed to expand his coalition.
As much as we need to be aware of changing conditions, there is a reason why Bayesian analysis - making small changes away from a baseline instead of wild swings back and forth - has proven successful. It’s often correct.
----------
Thanks for reading. If you like this newsletter, please pass it along!
Chris
Episode 2: Arab Spring in Hindsight
Chris Oates Feb 19
I spoke with Kerry Boyd Anderson, a Middle East analyst, about why the analytical community was unprepared for the Arab Spring, even after 10 years of intense focus on the region.
Subscribe wherever you get your podcasts and be sure to give a 5-star rating!
NB: The Beacon has been renamed Inside Political Risk, since there are loads of podcasts already with that name. I wouldn’t want someone looking for the latest on Middle East politics to wind up confused by an “urban fantasy audio drama” or a couple of chiropractors. The content and perspective will be the same as always.
Episode 1: What is Political Risk?
Chris Oates Feb 5
The Beacon is now a podcast!
This newsletter will now have an audio component to it. I’ll be interviewing some of the most interesting people working in and around the world of politics, business, and political risk.
Episodes will be posted in this feed and also wherever you get your podcasts.
In the first episode, I talk with Ben Godwin, Head of Analysis at PRISM Political Risk Management (http://prismlondon.co.uk/) about what is and is not political risk, and what goes into the industry that explains it.
Share The Beacon from Two Lanterns Advisory
Thanks very much for listening. Let me know if there are topics you’d like to hear covered.
2019: Recapping the year in political risk
What we got right, wrong, and missed
Chris Oates Jan 14
Around this time of year, you start to see lists of the top political risks for the next 12 months.
But we rarely see lists evaluating how our previous predictions performed. Even if it is not as topical, assessing past predictions is the only sure path towards a collective improvement and evolution of political risk as a practice.
So before I write my own top risks for 2020, here’s an overview of what the political risk world got right, got wrong, and completely missed in 2019.
Share The Beacon from Two Lanterns Advisory
What was predicted at the start of the year
I looked at the 2019 top risks lists of four political risk firms: the Eurasia Group, the Economist Intelligence Unit, Stratfor, and Oxford Analytica (full disclosure, I’m an advisor to Oxford Analytica, but I only assessed the firms as a collective).
The following topics were mentioned by multiple firms:
US-China trade war intensifies
European populists gain more power
North Korea standoff
No-deal Brexit
Italian banking crisis
Cyber attacks
Iran tensions
What political risk got right
I’m using a loose definition of whether a risk materialized. If the topic dominated risk discussions throughout the year as a thing that is happening (not just could happen), then I’ll count it.
Even under that definition, there are really only two that I can mark as being realized. US-China trade war continued throughout 2019 and is only being partially unwound.
US-Iran tensions also continued through 2019 and has certainly shaped actions in the region. (While the Soulemani strike happened in the first week of 2020, I’ll include it since it came out of a cycle of escalation in the last weeks of December).
We could quibble with the classification of these risks. They did not manifest in their worst predicted forms - the US and Iran are still not in outright war and the trade war could have been worse. But they were bad enough to force decision-makers worldwide react to them.
Risks that didn’t happen
Most of the risks on this list, however, did not come close to that definition.
The United Kingdom did not leave in a No Deal Brexit. Italy did not face a banking crisis. Cyberattacks happened but none was so crippling as to include it as a major geopolitical risk. North Korea continued at a standoff, but not markedly worse.
Populists did make gains in the European elections, but so did Greens and Liberals, which was not anticipated.
3/Know this is a bit of a narrative buster but as of right now the alliance which appears to have gained the most in the 2019 European election is the centrist, pro-EU Liberals. Home | 2019 European election results | European ParliamentOfficial results of the European elections held between 23 and 26 May 2019.election-results.eu
May 26th 2019
28 Retweets33 Likes
What got missed
The biggest single story in political risk, arguably, were the Hong Kong protests.
They shut down one of the world’s financial centers, called into question the relationship between Hong Kong and China, and posed questions for countries around the world about how to react.
Another huge story for risk was that of the continuing British political teeter-totter.
While No Deal Brexit did not materialize, it was not foreseen that there would be so many failed votes on a Withdrawal Agreement, a change in leadership, a new election, a landslide victory, and then the agreement for a deal that was basically the Withdrawal Agreement except that it could separate Northern Ireland from the rest of the UK (which was one of the few red lines considered to be drawn in permanent ink).
Matthew Thompson@mattuthompson
A momentous moment. The Withdrawal Agreement has passed the House of Commons. With a majority of 99. Something that Theresa May and Boris Johnson spent the last few years trying to achieve, done in a matter of days.
January 9th 2020
3 Retweets34 Likes
Or you could make a strong case that 2019’s big story was not just Hong Kong, but all the protests. As Jackson Diehl, deputy editorial page editor for the Washington Post, put it:
But this year is exceptional for the sheer breadth and diversity of the unrest. Hong Kong, which has now had 20 consecutive weeks of mass protests, has had perhaps the most noted uprising. But the Middle East has seen demonstrations in Algeria, Sudan, Egypt and Iraq in addition to Lebanon.
In Latin America, Chile’s riots followed mass protests in Ecuador, Argentina and Honduras. In Eastern Europe, Romania, Slovakia, the Czech Republic, Serbia and Georgia have been rocked. Vladimir Putin has had to contend with the largest street demonstrations in Russia since 2012. And the list goes on.
The political risk industry did not see this coming.
That is not an indictment of their abilities. The governments of all these countries did not see protests coming either.
But it does seem like a miss big enough to notice.
Why, if it’s an industry devoted to forecasting social and political upheaval around the world, did it not foresee social and political upheaval in lots of places around the world?
Lessons learned
This is only the second year I’ve conducted this exercise, but the correct answers seem less weighty than in 2018 and the misses bigger.
I’m not quite sure why. It’s not as if 2019 was certainly going to be a more volatile year than 2018 and would test us more deeply.
It may be that social movements are inherently harder to predict than elite politics, which have fewer actors and more comprehensible trigger points. Any year with widespread protests would therefore be one with lower accuracy.
But that doesn’t explain why UK politics, which for most of the year happened in Parliament, the most elite of locations, was another big miss.
I don’t have a definitive answer, but one possible reason is that political risk often looks for what major negative events could happen. The list of 2019 risks has quite a few of them, from major cyber attacks, to conflicts in North Korea and Iran.
A string of events, like the global wave of protests, may individually not make more than page A10 of the New York Times. Collectively, they can set the tone for the year. They are harder to predict because they demand a global narrative and stochastic occurrences. We are not predicting a single risk, but rising social temperatures that can flare up into something more at somewhat random intervals.
British politics was also hard to forecast because so much of it was based on unprecedented questions.
Would a party expel some of its most senior and popular members?
Would the prime minister prorogue parliament and then have the decision overturned by the Supreme Court?
Would a party vote for a general election when they are trailing in the polls by 10 points and could set up an alternate government?
It’s not only that these questions were hard to answer at the start of 2019. They were hard to know that they would ever have to be answered.
This is not to say that political risk is an impossible field, or that anything less than 100% accuracy should be seen as a critique.
But if 2019 tells us anything, it’s that we can’t simply make a set of predictions in January and expect them to hold for the next 12 months.
Stuff will happen. Events go to places we never anticipated. People do things that we hadn’t considered.
The essence of a good prediction is perhaps knowing when to abandon it and come up with a new one.
----------
Thanks for reading. If you like this newsletter, please pass it along!
Chris
I don’t know what’s going to happen with Iran
But the first phase looks to be over
Chris Oates Jan 8
For this issue of the Beacon, I thought I’d do something a bit different than normal. The death of Iranian General Qassem Souleimani has the potential to be the biggest geopolitical event in years. But it also might not.
It’s a highly uncertain situation that is still evolving. But even nascent crises require analysis and this analysis should still be tested and held accountable.
So below are some of the highlights of the immediate reactions and what we know from the reaction to events in the few days that have happened. I wrote this issue continuously throughout this time, so it roughly tracks what I thought about the situation (and the foreign policy consensus that I read) at the time.
Share The Beacon from Two Lanterns Advisory
Friday, January 3
It’s less than 24 hours since the death of Iranian General Qassem Souleimani by US missiles was announced. The immediate reactions speak to the magnitude of the act.
Spoke to a very knowledgeable person about what Iran's response to Soleimani's assassination might be. This would be the equivalent of Iran assassinating Petreus or Mattis, I argued. No, he responded, this is much bigger than that...
January 3rd 2020
2,160 Retweets5,150 Likes
"We need to presume we are now in a state of war with Iran… and that is not something that the Trump administration appears to have been prepared for." - Brett McGurk, former special envoy to coalition to defeat ISIS McGurk on US strike on Soleimani: ‘We need to presume that we are now in a state of war with Iran’Brett McGurk, who served as special envoy on the global coalition to defeat ISIS under Pres. Obama and Pres. Trump, talks to Rachel Maddow about the serious implications of the strike on Qassem Soleimani, the head of the Islamic Revolutionary Guard Corps-Quds Force.on.msnbc.com
January 3rd 2020
144 Retweets292 Likes
Soleimani was evil but he wasn’t the leader of a stateless terrorist org. He is the highest-ranking military commander assassinated by the US since Adm. Yamamoto in 1943. This is a major, unexpected development whose full import no one can predict.Opinion | Trump just upped the ante in the Middle East. Is he ready for what comes next?The president is facing one of the major tests of his career.wapo.st
January 3rd 2020
356 Retweets791 Likes
Global tensions are high today as Iran's top military commander Qasem Soleimani was killed last night in U.S drone strike in Iraq - so what can we expect if the conflict between D.C. and Tehran escalates? @dandrezner explains how Iran is likely to retaliate.
January 3rd 2020
10 Retweets19 Likes
“Anything short of a full-blown conventional war” is the assessment of the Fletcher School’s Dan Drezner. That’s not good.
We have extremely little to go on and what statements we do have appear contradictory.
Right now, we have two separate U.S. government explanations for assassinating Soleimani. Pentagon: It was meant to deter future Iranian aggression. Pompeo: "There was an imminent attack taking place" we needed to stop. Could both be true? I suppose. But it's confusing.
January 3rd 2020
3,034 Retweets9,911 Likes
“Long in the making.” Yet another indication that the strike on #Soleimani was not in response to an imminent specific threat, but rather a situation that allowed—and the Constitution that requires—the President to have gone to Congress for authorization.
January 3rd 2020
941 Retweets1,836 Likes
In short, we’re being asked to analyze the killing of the second-most important figure in Iran when we don’t have a clear reason why he was killed, we don’t know how Iran will retaliate, and we don’t know how the White House will react, we don’t know what Congress will do, and we don’t know what Iraq’s role in all of this is - given that it happened on their territory and Baghdad has agency of its own.
In the aftermath of the assassination of Qassim Suleimani, here's why the biggest problem for Trump may come from the government of Iraq. Are the US and Iran headed for war?On Thursday night, President Donald Trump ordered the US military to assassinate Qassim Suleimani, a man widely considered Iran’s second most important leader. Following the successful strike, fury has erupted in Iran. Threat levels have escalated in both Tehran and Washington. Are the US and Iran h…gzeromedia.com
January 3rd 2020
22 Retweets51 Likes
My current sense is worry about what the cause was and effect will be. The WWIII memes trending on twitter are likely hyperbole, but there is a large range of terrible consequences that fall short of that and some sort of reaction appears inevitable.
Saturday, January 4
President Trump has threatened to bomb Iranian cultural sites, which would be a war crime. That is highly unlikely to happen - but it’s also not a good sign that this is even in the discussion. Iran has not taken any action, but the rhetoric is worsening.
Donald J. Trump @realDonaldTrump
....targeted 52 Iranian sites (representing the 52 American hostages taken by Iran many years ago), some at a very high level & important to Iran & the Iranian culture, and those targets, and Iran itself, WILL BE HIT VERY FAST AND VERY HARD. The USA wants no more threats!
January 4th 2020
63,958 Retweets271,867 Likes
Sunday January 5
Amid the fallout of the U.S. drone strike on Friday that killed Soleimani, Sunday saw the following whiplash-inducing developments unfold almost simultaneously:
The Iraqi parliament voted to oust the thousands of U.S. service members stationed in Iraq.
President Trump threatened Baghdad with sanctions if U.S. forces were forced to leave the country.
The U.S. military announced that it would halt its fight against ISIS in Iraq.
Iranian state television reported that Iran would no longer abide by the limits of the 2015 nuclear deal.
Hundreds of thousands of mourners packed Iran's streets to pay respect to Soleimani.
Additionally, Trump may have tried to invoke the War Powers Act regarding a strike against Iran? It’s difficult to know, but this seems to be an attempt to be escalatory?
Donald J. Trump @realDonaldTrump
These Media Posts will serve as notification to the United States Congress that should Iran strike any U.S. person or target, the United States will quickly & fully strike back, & perhaps in a disproportionate manner. Such legal notice is not required, but is given nevertheless!
January 5th 2020
83,843 Retweets372,317 Likes
Meanwhile, Speaker Pelosi told her caucus that the House will vote on a resolution to block military action with Iran.
Again, the lack of an Iranian response is a positive sign that it won’t escalate immediately, but it does appear that the US is heading towards action and will, at a minimum, seriously harm its relations with Iraq.
Connor O'Brien @connorobrienNH
JUST IN: Pelosi says in a letter to Dems that the House will vote this week on a War Powers resolution to block military action against Iran. She says it will be sponsored by Rep. Elissa Slotkin, a former CIA analyst who served in Iraq.
January 6th 2020
279 Retweets544 Likes
Monday, January 6
The military said that it is withdrawing from Iraq, and the Secretary of Defense then said that the letter was sent in error.
Joint Chiefs Chair GEN Milley: “That letter is a draft it was a mistake, it was unsigned, it should not have been released…poorly worded, implies withdrawal, that is not what’s happening”
January 6th 2020
2,431 Retweets5,279 Likes
But there are also reports that the United States is going to sanction Iraq. So the relations with Iraq are definitely trending downwards.
Andrew deGrandpre @adegrandpre
Stunning. One official tells The Post that the administration’s plan is to wait “at least a little while” on the making a decision to sanction Baghdad — to see if the Iraqis actually make good on their threat to evict American troops. After Trump’s threat, administration begins drafting possible sanctions against IraqCritics have questioned the wisdom of imposing economic punishments on a U.S. ally.washingtonpost.com
January 7th 2020
279 Retweets364 Likes
Tuesday, January 7
President Macron of France is attempting to mediate, or is at least acting against escalation and trying to prevent too much of the 2015 nuclear deal from being breached.
French President Emmanuel Macron called Iranian President Hassan Rouhani and asked him not to escalate tensions in the Middle East Bloomberg - Are you a robot?trib.al
January 7th 2020
57 Retweets115 Likes
Other European governments have also called for a cooling down of tensions, although UK Prime Minister Boris Johnson hasn’t been seen in public during the crisis after a vacation in the Caribbean.
Then Iran attacks bases in Iraq that house US troops. The first reactions are worries that this is the reaction that kicks off a new round of an escalatory cycle.
The Washington Post @washingtonpost
"This was an act of war": Lindsey Graham warns of U.S. retaliation following Iran missile strike Lindsey Graham threatens to take Iran ‘out of the oil business’ in wake of missile strikeGraham described Trump as “very methodical in his thinking,” saying the president has “one of the best national security teams I’ve seen since I’ve been up here.”wapo.st
January 8th 2020
174 Retweets473 Likes
But! Then it appears that this was designed to be the Iranian response to the Souleimani death and was deliberately limited to end this round of tensions. On first glance, Trump is also willing to de-escalate and take this as finished.
Iran took & concluded proportionate measures in self-defense under Article 51 of UN Charter targeting base from which cowardly armed attack against our citizens & senior officials were launched. We do not seek escalation or war, but will defend ourselves against any aggression.
January 8th 2020
18,095 Retweets47,112 Likes
Donald J. Trump @realDonaldTrump
All is well! Missiles launched from Iran at two military bases located in Iraq. Assessment of casualties & damages taking place now. So far, so good! We have the most powerful and well equipped military anywhere in the world, by far! I will be making a statement tomorrow morning.
January 8th 2020
141,218 Retweets613,832 Likes
Stepping back
I’m deliberately sending this out before Trump’s address on Thursday morning so that my prediction is not shaped by what happens (there will be an update after the address).
In this situation, the only thing that we can do is to focus on what we know - and our ignorance is one of the most important things we know.
There are four big questions that I wrote down last Friday that guided my own thinking. Some of them have been resolved (answers in italics), others not so much.
What was the cause for the strike? Was is to head off an attack, to respond to the embassy incursion, or because that option was included as a bureaucratic maneuver to make other responses more palatable, as some of the immediate reporting suggested (see these from Washington Post, Associated Press, Los Angeles Times, Politico and New York Times)? We still don’t know, but this is less important for predicting what comes next.
How will Iran respond? Hezbollah has said that it will not target US civilians, but will other Iranian proxies follow the same limits in their responses? What about Iranian state action? We seem to have the first response, which was limited.
Will Iraq expel US forces? If so, would the US actually sanction Baghdad? Would Iran see this as a sufficient response to claim victory? It looks like leaving Iraq is more likely than not, though perhaps not all troops - still undetermined.
Will Congress or Europe act as a constraint on the White House? Even without an explicit resolution pass the House, a clear sign that escalatory action would not find support elsewhere could limit Trump’s or Pompeo’s willingness to push on this policy. It doesn’t seem as if either will be a major constraint, but they also haven’t been supportive of greater escalation.
Can we have a prediction?
No.
We have a situation with at least five distinct actors. Two (Trump and Iran) could set off a new escalatory cycle of violence that the other three (Congress, Europe, and Iraq) may not be able to prevent. Meanwhile, there are other actors who with a simple mistake (like sending out a purportedly draft letter or a proxy that launches an attack on its own) can kick off a new round of confusion and misperceptions.
I think that any risk analyst with a clear understanding of their own limits knows that there is no way of predicting what will happen next. The worries of an immediate war have subsided and the reaction to the missile strikes is unknown. Iraq has emerged as an independent actor that could be a wild card in the calculation.
In a sense, we’re in a holding moment, but if I had to guess - and this is not something that I say with much confidence - I think that the US and Iran will not escalate further but that US relations with Iraq will deteriorate. But, really, we can only wait and see.
Update: It looks as if this prediction was correct. Trump has broadly agreed to deescalation. It still remains to be seen how US-Iraqi relations evolve, but the current signs are not pointing downwards.
----------
Thanks for reading. If you like this newsletter, please pass it along!
Chris
Electoral inefficiency of the left
Parsing the results of the UK General Election
Chris Oates Dec 18, 2019
The British General Election last week ended in a landslide.
Immediately, as happens after all elections now, the analysis and punditizing began. It was Jeremy Corbyn. It was Brexit. It was not tactical voting. It was the BBC.
So what was it really?
Share The Beacon from Two Lanterns Advisory
Lukewarm takes
Nate Silver is now a quasi-professional contrarian and skeptic of the kneejerk reaction. His immediate summary is a good starting point.
Generally when there's a landslide (as in the UK tonight) everyone's hot takes are mostly right and when there's a close result everyone's hot takes are mostly wrong.
December 13th 2019
205 Retweets1,713 Likes
You probably CAN blame Corbyn AND Brexit AND Labour AND going too far to the left AND [insert hypothesis here], because you need some combination of ALL of those things for them to lose to the Tories by 14 points or whatever it's going to wind up being.
December 13th 2019
301 Retweets1,833 Likes
However, even if it was a combination of everything, the relative importance matters. If we find that Brexit mattered for 90% of the loss, the Corbyn wing of Labour will probably stay in power. If it was only Corbyn, we could assume that Labour will snap back in the polls with a change in leadership but maintaining similar policies. Policy shift would happen if Labour sees it as a result of their political positioning.
Luckily, some polling data can start to answer these questions.
NEW Opinium poll The main reasons people did not vote Labour — The leadership (43%) — Brexit (17%) — Their economic policies (12%) (Research carried out Dec 13)
December 13th 2019
3,145 Retweets5,144 Likes
Conservative seats
But there’s a broader point here. Before we read thousands of pixel inches about what cost Labour with the voters, we should recognize a bigger fact. It was Labour that determined the election. The Conservative Party did not win this election as much as Labour lost it.
One of the idiosyncrasies of British electoral politics is how much the share of the votes won can deviate from the share of the seats won. Below is the Conservative results in the elections since 1983. It has swung from 18.7 percentage points more seats than votes in 1983 to 6.5 points below in 2001.
The Tories even lost seats from 2015 to 2017 while seeing a big jump in the people voting for them.
The 2019 landslide was not a result in a huge swing in votes. Boris Johnson got a meager 1.2 percentage points more Theresa May did. The difference is that his votes were much more efficiently allocated.
Electoral Reform Society@electoralreform
Across Britain, it took... 🗳️864,743 votes to elect 1 Green MP 🗳️642,303 votes to elect 0 Brexit Party MPs 🗳️334,122 votes to elect a Lib Dem 🗳️50,817 votes for a Labour MP 🗳️38,316 votes for a Plaid Cymru MP 🗳️38,300 votes for a Con. MP 🗳️25,882 votes for a SNP MP #ScrapFPTP
December 13th 2019
9,744 Retweets16,148 Likes
Labour’s collapse
It’s not a secret how this happened. It is the result of Labour losing support to smaller parties.
From the 2017 to 2019 elections, Labour lost 7.9 percentage points while the Liberal Democrats gained 4.2, the Greens 1.1, and the Scottish National Party 0.8 (the new Brexit Party took 2.0 points, which was about what UKIP had in 2017).
So 16.4% of the fall in Labour’s vote share correlated with an increase for right-wing and pro-Brexit parties, and 83% of it “stayed” within the left-wing and anti-Brexit camp. This followed the path I predicted in July:
Forecasting a General Election is therefore almost impossible. If Boris Johnson is able to bring most Brexit Party voters to the Conservatives, bringing their national polling position close to 40%, while Labour and Lib Dems stay at 20-25%, the Conservatives could receive a huge majority government, according to Electoral Calculus’ prediction model.
If the reverse happens - or Labour, Lib Dems, and Greens form some kind of alliance and coordinate to prevent vote splitting - then Labour could receive an outright majority, or at least enough to form an anti-Brexit coalition with the Lib Dems and SNP.
This doesn’t give us the answer, but it helps to identify the central question: which side of the political spectrum comes together more rapidly if a General Election is called.
If we see that Labour is repeating its 2017 performance, that’s a sign that Johnson’s decision has backfired and a second Brexit referendum is likely. If the Brexit Party stumbles, outpaced by the deeper and wider electoral infrastructure of the Conservative Party, expect a big victory for the Tories and the Withdrawal Agreement passing.
Conservative tilt
This is basically what happened.
The left was split, not helped by Labour’s Brexit position and Corbyn’s abysmal approval ratings. The Brexit Party effectively stood down as an anti-Conservative force at the beginning of the campaign.
The Conservatives and the Brexit Party thereby combined for 45.6% of the vote but the Conservatives took 56% of seats. Labour/Lib Dem/SNP/Plaid/Green combined for 50.9% of the vote but only 41.1% of seats. (Northern Ireland excluded).
But this isn’t the whole story.
It’s easy to stop here and say that Labour couldn’t consolidate their side of politics, the Tories did, first-past-the-post exaggerated the difference, and that’s all that matters.
But after going through the constituency-level data (helpfully gathered by my colleague Joel Paul), I found that Labour faced an additional hurdle, and one that kept them from the majority.
Let’s consolidate everything into two broad parties, a Right Party and a Left Party. Assuming that the Brexit Party splits 25-75 towards the Right and 100% of the Lib Dem, Plaid, SNP, and Green voters go to the left, the Right Party would still have gotten a majority in Parliament.
It would have been smaller than their majority in real life, 335 compared to 365, but Boris Johnson would still be prime minister. Even after consolidating the two sides of politics, the conservative wing outperforms its vote share by 6 percentage points.
The below chart shows how the Right and Left Parties would perform based on the share of third party votes going towards them. The Lib Dems, as the third-largest party by vote share, naturally has the largest effect. If their voters split equally, the Left Party would have only 251 seats; if all their voters went to the Left, they’d have 305 seats.
But it is only by getting a majority of the Brexit party supporters that the Left would get close to a majority.
Without pro-Brexit voters last Thursday, the Left - even with a majority of the votes - could not won the election.
Hurdles for the global left?
Now, before we go further, the usual caveats apply.
Safe seats where some parties didn’t contest the vote can skew the overall totals. If there it had been a two-party system or ranked choice voting to create this effect, the campaign would have been different. Other leaders would have changed the overall vote shares and results would have shifted.
But it is nevertheless a big gap, and it’s remarkable because it mirrors what we see in US politics. The Republican Party won control of the House of Representatives in 2012 while gaining fewer votes than the Democrats; Donald Trump is expected to lose the popular vote in 2020 but may win through the Electoral College.
As parties in advanced democracies are sorting themselves on lines of age and education, it is creating an additional obstacle for liberal parties. Their voters are more likely to be mobile and to move to growing urban areas. Each young Democrat that moves from Wisconsin to New York is effectively a wasted vote. Each older person in northern England who decides to switch from Labour to Conservative is a valuable swing voter.
This is not to say that Labour is doomed, or that Trump will win re-election. But when analyzing British politics or the US House/Senate/Electoral college, we will need to mentally dock a few percentage points from the liberal parties from their standings in the polls.
The ground, as least under the current partisan sorting, is tilted against them and our predictions must be updated to reflect this.
----------
Thanks for reading. If you like this newsletter, please pass it along!
Chris
Holiday Spirits at Risk
Politics and bars
Chris Oates Dec 9, 2019
Now is the time of the year (at least in much of the Northern Hemisphere) when people gather indoors, away from the cold and snow and drizzle, and enjoy a nice dinner. Maybe even a drink or two. It’s December. Treat yo self.
But if you’re in Boston, that drink could represent political risk that could cripple an entire industry of small business owners and/or is the largest single obstacle to a thriving sector and vibrant neighborhoods.
Share The Beacon from Two Lanterns Advisory
Post-Prohibition Prohibition
The problem with liquor licenses in Boston goes back to the 1930s. The State Legislature, dominated by the old WASP establishment, thought that the Irish-dominated city government would let alcohol run wild after Prohibition ended. The stereotype led the Legislature into giving itself authority the number of liquor licenses allowed within the city limits.
That number has stayed basically the same since 1933, a little more than 1,000 licenses for a city of 685,000 that’s the entertainment hub for a region of 4.8 million.
The result is obvious to any student of Econ 101.
Growing demand for places to drink with a limited supply of liquor licenses means prices rise. A license that officially costs, according to the city’s website, $2,800, was recently sold for $455,000.
Unequal dining
That is a problem.
The high costs of liquor licenses means that many restaurants cannot afford to have them. If there’s one on the market when you’re opening a restaurant, you would have to spend additional hundreds of thousands of dollars for the license - which isn’t something a lot of chefs and entrepreneurs have the cash for.
The result is inevitable. The licenses that existed were unevenly distributed. Wealthier neighborhoods and bars near the major business and tourist sites could afford the licenses, predicting that they could eventually make back the cost of acquiring one. Poorer areas catering to a local clientele could not and did not.
The below map created by Boston Magazine shows the situation in 2016. For those not familiar with Boston’s geography, the areas with lots of blue dots are the wealthy sections of town. The sections south of that with only scattered dots are some of the most marginalized.
Source for this map: https://www.bostonmagazine.com/restaurants/2016/12/18/reform-liquor-licenses-boston/
Reforming the system?
The liquor license regime was loosened a little bit a few years ago. City Councillor ( now Congresswoman) Ayanna Pressley lobbied the legislature to grant an additional 153 licenses, with most of them restricted to underserved neighborhoods.
@AyannaPressley “Neighborhoods like Roxbury, Mattapan, Dorchester had not seen a new sit-down restaurant for ten, fifteen, twenty years,” she said. “Just to contextualize it, the North End is a .2 square mile radius, and had ninety-nine liquor licenses.” If You Build It, Will They Come?Low-cost liquor licenses were created to revitalize neighborhoods like Roxbury’s Dudley Square. But is the strategy working?to.wgbh.org
June 25th 2018
1 Retweet2 Likes
It was a big victory for restauranteurs in Boston. But not just the ones who got the new licenses.
No, the real victory went to all the current owners of liquor licenses, who prevented the system from being opened up even further.
From that Boston Magazine article, here’s how they described it:
While officials on both sides agree that it’s high time for change, they have conflicting views of how to go about fixing the problem. It’s ridiculous to let a legislator from Springfield have greater power over the number of licenses allotted to Boston than the city’s own councilors, but a complete overhaul of the system would drain millions of dollars from the pockets of those who currently hold unrestricted licenses.
“I thought that was really a slippery slope to go down. We were really concerned with some businesses in our communities,” says Patrick O’Connor, a state senator from Weymouth who opposed the reform.
Many business owners leverage their secondhand licenses as an asset, and expected them to hold a high value that increases over time. By eradicating the current system, the value of those licenses, and the business’s assets, would drop by more than $300,000 in their resale value.
“I don’t think the solution is to render all those licenses valueless,” O’Connor says.
Assets at risk
Anyone who doesn’t own a restaurant might question this idea.
Why not return licenses to the price that they are supposed to have? Why allow them to cost hundreds of thousands more that they’re worth only because the market is artificially limited? Are you really “draining” millions of dollars or merely recognizing that they should never have cost that much in the first place?
Liquor license reform would seem to be the perfect non-partisan policy. Conservatives see the government stop overregulating small businesses. Progressives see greater equality. Consumers and voters get more selection, livelier neighborhoods, and a smoother functioning economy.
The problem is that, as the bolded section above shows, many business owners have based their operations on the assumption that there would be no reform. They have projected that Beacon Hill will not have a change of heart, will not be swayed by the next Councillor, and that they can build a business model on a piece of paper that could, in theory, lose all its value in a single day.
I should stress that I have sympathy for these legislators. No one wants to take on hundreds of business owners with lots at stake. And I have sympathy for the restaurant and bar owners. In order to do business, they have had to pay huge sums of money. It would seem intensely unfair that politicians tell you you’ve lost all that cash and that new competitors can enter the market without similar debt.
But the fact remains that - using the last transactions for liquor licenses I’ve seen and some back of the envelope math - the Massachusetts State House could destroy $340 million in value in a single sector in a single city with a single policy.
And that policy would probably be popular.
Political risk in a beer
If there’s one central thesis to this newsletter and to Two Lanterns, it’s the political risk is much more than what’s in the headlines of the Wall Street Journal.
I haven’t yet researched this enough to know what the future of Boston’s liquor laws will be, and if you’re reading this outside of the Boston area, you probably don’t care. But it is the perfect microcosm of how political risk can seep into places as safe as the neighborhood restaurant.
A law passed 85 years ago created a market that served no real purpose. Restaurants and bars had to participate in that market in order to stay in business. That market had negative effects on those in the market, those outside the market, and regular people who want to be able to grab a drink. No one knows how or if or when that market will be changed. And it’s up to politicians to do so.
Anyone looking to open a restaurant in Boston, any bar owner wondering whether to use a license as collateral, or any resident wondering whether to move to an area that is a little dead at night but is promised to pick up soon, is basing their decision on the probability that politics addresses this issue.
But are they? Is there a discount rate added to these licenses to reflect the risk to their long-term value? Is there insurance being taken out on them in case they don’t retain value? And is there the chance that something takes off that destroys the value of the licenses without politics getting involved, in the same way that Uber and Lyft drove a collapse in taxi medallion value?
I’m guessing not, which is part of the problem. To quote Bill Nighy’s character from a great/terrible Christmas movie, political risk is all around us. But we need to notice it.
----------
Thanks for reading. If you like this newsletter, please pass it along!
Chris
Railroad gauges and the NBA
A cultural Cold War emerging?
Chris Oates Oct 10, 2019
The NBA is currently in the middle of a balancing act, caught between the world’s two largest powers.
Last week, Houston Rockets GM Daryl Morey tweeted support for the protestors in Hong Kong. It was an anodyne tweet by American standards: “fight for freedom, stand with Hong Kong.” By Chinese standards, this crossed a line.
The Chinese consulate in Houston denounced the tweet. An exhibition game by the Rockets affiliate was cancelled. Chinese television stated that it would not show the Lakers and the Nets games that are being played in China this weekend. To ease tensions, the NBA immediately put out a statement implying regret about the incident and a Chinese-language statement that went even farther in seeming to take Beijing’s side over Morey’s.
But that statement prompted American elected officials to criticize the NBA for the “prioritization of profits over human rights” and putting human rights “up for sale.” It was a forceful, unified message delivered by two politicians who would otherwise never agree.
We’re better than this; human rights shouldn’t be for sale & the NBA shouldn’t be assisting Chinese communist censorship.
October 7th 2019
3,886 Retweets14,096 Likes
The only thing the NBA should be apologizing for is their blatant prioritization of profits over human rights. What an embarrassment. https://t.co/bbiwCBTwc1
Sopan Deb @SopanDeb
NEW: the NBA has released a statement on Daryl Morey: https://t.co/FOI79W31b1
October 7th 2019
3,651 Retweets13,964 Likes
The issue is ongoing and, as I write this, appears to be getting worse.
Cameron Wilson 韦侃仑 @CameronWEF
Houston Rockets artwork on an indoor basketball court in Shanghai being painted over. We live in sad times.
October 8th 2019
2,443 Retweets4,316 Likes
Video games and videos
On the same front page as this piece about the NBA commissioner’s response was also this article about how a professional video game player has been expelled from tournaments for a year and fined $10,000 in prize money for supporting Hong Kong protestors.
The American company that did this, Blizzard Entertainment, harmed a person’s livelihood for their political statements against the Chinese government. They too are now facing consequences from their American audience.
@PlayHearthstone I’ve thought about it and there’s just no other ethical choice: I’m quitting @PlayHearthstone and boycotting all other @Blizzard_Ent products until they reinstate @blitzchungHS and the casters. I encourage everyone else to do the same.
October 8th 2019
29 Retweets168 Likes
And, just to compound the issue, a leak has revealed that the Chinese-based social network TikTok censors content that mentions Tiananmen Square or Tibetan independence.
Although the company claims that it no longer censors specific topics, it has over 27 million active users in the United States. It seems only a matter of time before politicians or advocates start to ask questions about whether Chinese government censorship is blocking what Americans can say to each other.
Cross-Cultural trade
The challenge for these corporations is how to please both countries at the same time.
If the NBA prevents its employees from speaking their minds, it violates democratic values (and will anger their American fans, although the profit motive shouldn’t be as important as basic tenets of a free society).
If more players and executives tweet about Hong Kong, it will trigger heavier and heavier responses from Beijing.
There is no middle ground. The US and Chinese demands are so far mutually exclusive.
That’s not surprising. According to this article from 2015, “almost upon arrival in China, one is counseled not to discuss the three Ts — Taiwan, Tibet, and Tiananmen.”
You could say that the NBA was caught out by not knowing that Hong Kong is now on that list. But if player tomorrow tweets support for Tibetan independence, the result will be the same. The dilemma of Chinese censorship and American free speech seems intractable.
So how is it that we are seeing this issue flare up here and not for every other company involved in the $737 billion in US-China trade, given that many of them have also run into political sensitivities?
Cultural trade
It’s probably not a coincidence that the above incidents happened in the cultural/entertainment/media world.
We can assume that Beijing cares about shutting down speech on these issues from anyone it does business with. But would the Chinese government really patrol the tweets of a soybean farmer in Iowa?
Maybe, but it wouldn’t be worth the effort. No one eating tofu in Shanghai would know they came from beans grown by @DubuqueDude12345.
Sports and culture are different.
Fans are not watching basketball as an abstract set of rules and events. They’re watching LeBron James and Steph Curry play basketball. They’re rooting for the Lakers or the Nets. When Yao Ming played for the Houston Rockets, he turned millions of people into Rockets fans, who then became Tracy McGrady fans and made McGrady’s jersey the best-selling in China.
It is impossible to separate LeBron the basketball player from LeBron the person. Actors too run into this, given their celebrity, as well as writers, artists, and anyone who people would want to listen to.
As long as people have opinions and platforms to speak their minds - whether that’s Twitter, a stream like the Hong Kong gamer, or TikTok videos - cultural industries will run into these issues. Their business is the actions and opinions of people, and people take actions and have opinions. Sometimes, governments will get angry about those.
Where is this all heading?
It’s impossible to predict the outcome of this right now. There are a few scenarios:
1) Global censorship
As discussed by Matt Yglesias in Vox, Hollywood has already appeased Chinese demands. They changed a Marvel character from a Tibetan monk to Tilda Swinton to avoid Beijing banning Doctor Strange.
In this scenario, corporations calculate that the Chinese market is worth whatever criticism they receive from the rest of the world and tailor their product to their demands. We see American and European companies largely following Beijing’s speech codes worldwide to avoid any repercussions on their sales and supply chains.
2) Railroad gauges
When railroads were first invented, different companies used different gauges (the width of the rails). This made some railroads incompatible with each other, and required time-consuming unloading and loading or wheel changing to switch tracks. Incidentally, this is one of the advantages the Union had over the Confederacy. Virginia and North Carolina used tracks that were an inch and a half narrower than the rest of the South, which slowed down their internal supply routes.
If US and Chinese audiences/governments insist on mutually incompatible visions, then the cultural products may be stuck on different gauges.
Some businesses will make the changes to switch tracks, perhaps making two edits of the same movie or creating different codes of conduct for each market. This is perhaps what TikTok is doing, calibrating its level of censorship to the user’s country.
Others will choose to ride with only one system. They’ll find the costs associated with making switches are too high or flat-out impossible.
Tom Scocca at Slate has argued for this scenario (or at least for considering it). He believes that the NBA must be willing to walk away from China now, before their commitment gets even greater, and that the speech of its employees is not something that can change tracks on its way across the Pacific.
3) Modus vivendi
In this scenario, businesses will steer away from Chinese red lines when in China and China will use its firewall to block any extra-territorial comments. Both sides will accept that the other will follow their own values and seek to contain them. It’s a middle-ish ground scenario that requires China to desist from the retaliatory measures and for businesses not to overreact. It will probably also require that any Chinese-originated business like TikTok permit speech according to each country it’s operating in.
However, this scenario requires concessions from the Chinese government on some of its most sensitive issues. As their global power rises, Beijing may not wish to do so, and then we’re at Scenario 2.
4) Cultural political risks
Oil companies have long recognized political risks. There was always a chance that the host country could nationalize the oil, raise fees dramatically, or prevent profits from crossing borders. These companies can’t avoid where the oil is located, and they can’t avoid the salience of natural resources to government finances, so they accept the risk of politics affecting their business.
In this scenario, cultural industries experience the same thing. They can’t change the fact that their customers live in countries where politics conflict with their employees’ opinions. This week it is China, but they weren’t the first to criticize athletes and artists and they surely won’t be the last. Authoritarian countries and even democratic countries with exceptions to free speech (legal or informal) will occasionally hurt business in order to retaliate against some statement or position.
Political risk will be a constant concern. Cultural businesses will need to discount their investment value in countries that are most likely to block their activity and closely monitor governmental temperment to know when their finances are about to take a hit.
Am I confident in where we’re heading?
No.
We’re seeing the start of a new era of global business. A single thought from one executive is jeopardizing the $4 billion NBA China. Banning a single gamer has sparked a global boycott against a video game company. I wouldn’t be surprised to see Congressional hearings about TikTok.
The confluence of global communication networks that individuals can easily access coinciding with the rise of an economic superpower is a genuinely new phenomenon. There’s no guarantee which scenario will play out, or if different businesses go down different paths.
Much still depends on the actions of these executives, governments, and consumers in each country - which are themselves dependent on the actions of other actors. Time will tell.
One near certainty, however, is that even if each of these incidents resolve themselves quickly, that they won’t be the last time cultural industries are caught between the values they espouse and the profits they seek.
----------
Thanks for reading. If you like this newsletter, please pass it along!
Chris
Systems analysis and impeachment
How did this all happen so quickly?
Chris Oates Sep 27, 2019
Over the last week, one of the big questions in American politics appears to have resolved itself.
The following chart shows the prediction market for whether Donald Trump will be impeached during his first term. Try to spot when news of Trump’s interaction with the Ukrainian president broke.
I said back on May 1: “I’m not saying that [impeachment] will happen. But I bet that, if it does, lots of analysts will be surprised at how quickly events move.”
Lets walk through why the existing discussion back then on impeachment was missing a few things.
What happened this week?
Even as the story is cantering through one breaking news alert to another, we can study the recently historical decision by Speaker Nancy Pelosi to open an impeachment inquiry.
After so much time spent pushing back against impeachment, why did Pelosi now change her mind?
Through much of the coverage of impeachment through 2019, there has been one clear diagnosis: Pelosi was wary of impeachment because she thought that it might imperil frontline members, those who won the swing seats that give Democrats the majority.
The assumption went that as long as frontliners were against impeachment because they represented districts that leaned Republican, Pelosi would oppose it.
That analysis was, in part, correct. But it also missed a giant piece of the puzzle.
The three legs of the Democratic caucus
It’s true that the frontliners are a crucial part of the caucus. But they are no more than 40 members in a caucus of 235 and were divided on impeachment. They alone could not have stopped the decision if the rest of the caucus wanted it.
One way to divide the House Democrats is between those representatives in safe seats mostly vulnerable to primary challenges, those in frontline seats mostly vulnerable in the general election, and party loyalists who aren’t really vulnerable from anywhere. That is a massive simplification, of course, but it’s sufficient as a theoretical framework for the moment.
According to this great article detailing the changing dynamics in the Democratic caucus over the summer, the real shift on impeachment happened not among the frontline representatives, but among the primary vulnerable and party loyalists.
The following are very basic schematics of the avenues of political pressure and attitudes towards impeachment within House Democrats. They’re a little busy so I’ll explain them below.
In July 2019, we saw that primary vulnerable Democrats were in favor of impeachment in response to a pro-impeachment Democratic base. Swing voters (perceived to be anti-impeachment) did not have a large impact on these members. These swing and lean Republican voters were determinants for frontline Representatives and party loyalists concerned about the overall majority.
Over the course of August, the pressure from the Democratic base increased, drawing more primary vulnerable members into a pro-impeachment stance. We also saw some frontline members come out in favor of impeachment, either moved by the base or because they believed it to be the right thing to do.
As an aside: I’m talking here only about the political calculations. Many members appear to be genuinely moved by the merits of the Mueller report. Assume that that exists within every group. Now let’s get back to the political approach.
Once September came, party loyalists had started to doubt the strategy. According to the Intercept piece, they were worrying about the risks of non-impeachment. It could de-mobilize Democratic supporters in 2020, costing them volunteers to knock on doors in Wisconsin. It could undercut their messaging on Trump, making it seem like the accusations they threw at him were not that serious, because otherwise they would have impeached.
Generally, there was a sense that non-impeachment could hurt the party as a whole. In my schematic, they had moved to an ambivalent position before the Ukraine revelations and were becoming less receptive to frontline members.
The tipping point
In hindsight from the future, it may seem that the Ukraine story was inevitably going to lead to impeachment, a scandal so huge that no other response was merited.
However, when it first appeared, that this was not the immediate reaction.
A couple of days after the story broke, an aide to Pelosi was asked if the Speaker was moving towards impeachment. The response was “no. see any GOP votes for it?”
This story could have travelled the path similar to others that had come before it. Bad, but not worthy of that final step towards impeachment.
Democrats have had ample opportunities to call for impeachment. The release of the Mueller Report, Mueller’s testimony, Corey Lewandowski’s testimony, Michael Cohen’s testimony and other events all could have been tipping points, but they weren’t.
It may be that we will retrospectively classify the Ukraine story as a tipping point. But it’s important to note that before it came out, Democrats had become ready to tip.
House Democrats as a system
The House Democratic caucus can be viewed as a system. It designed to generate the best output for the Democratic Party collectively and each individual member, based on the inputs they receive from events in the world, base voters, and swing voters.
That’s why I said in May that impeachment could come and it could come fast. Within this system, the key determinant was a group - the party loyalists - who could be expected to move in relative unison and only when the view had changed that impeachment was now a net benefit.
Impeachment was not going to be a linear progression of individual members making evenly paced announcements on their way up to a majority of the House. Those like Adam Schiff, John Lewis, and others close to leadership would move together without much advance warning. When they changed their minds, so would Pelosi, and so would the entirety of the House Democratic system.
Because of this group’s role in the caucus and their collective decision strategy, we had been spending 2019 with a situation in which nothing would happen - until all of a sudden it would.
----------
Thanks for reading. If you like this newsletter, please pass it along!
Chris
What about the canaries not in the coal mines?
Tracking non-leading indicators
Chris Oates Sep 13, 2019
Nate Silver at 538 recently tweeted out his take on one of the most important pieces of information for the Democratic nomination.
Continues to be the single most bullish indicator for Warren IMO. She does a lot better with people more interested in politics (and Sanders does a lot worse). Those folks may be early adopters of what people who are less interested will decide later on; also more likely to vote.
Matt Grossmann @MattGrossmann
There are also big divisions based on interest in politics, but not at the top: - Biden leads across the board - Bernie is especially popular with those uninterested in politics - Warren is especially popular with those extremely interested in politics https://t.co/4ZYeTqbWd9 https://t.co/fnqGMFZM98
September 10th 2019
173 Retweets786 Likes
This gets to the heart of political analysis. Ultimately, all analysts want to find the Rosetta Stone of politics, that one data point that points the way through the informational morass and on to the glory fields of being a protagonist in the next Michael Lewis book.
But how do we know when we’ve found that clear bellwether when political data can be interpreted in different ways?
Matthew Yglesias @mattyglesias
Bernie’s strength being concentrated among people who aren’t interested in politics is probably a bearish indicator for his odds, but it also speaks to his strength as a general election candidate.
Matt Grossmann @MattGrossmann
There are also big divisions based on interest in politics, but not at the top: - Biden leads across the board - Bernie is especially popular with those uninterested in politics - Warren is especially popular with those extremely interested in politics https://t.co/4ZYeTqbWd9 https://t.co/fnqGMFZM98
September 10th 2019
30 Retweets239 Likes
Canaries in Iowa
This is something that has definitely affected the Democratic primary race. At the moment, there are few concrete pieces of information.
We know the candidates’ fundraising (though on a quarterly basis), their endorsements (which don’t amount to much yet), the crowd sizes for their speeches (extremely rough proxy), and their polling.
The last one has attracted nearly all the attention. But the historical record should make us pause for what it can tell us. Below are the top three pollers in September of the year before the election for recent cycles, with the eventual winner in bold.
2016 GOP: Donald Trump 33%, Ben Carson 20%, Jeb Bush 7%
2016 Dem: Hillary Clinton 45%, Bernie Sanders 25%, Joe Biden 17%
2012 GOP: Mitt Romney 22%, Sarah Palin 17%, Mike Huckabee 16%
2008 GOP: Rudy Giuliani 30%, Fred Thomsen 25%, John McCain 15%
2008 Dem: Hillary Clinton 40%, Barack Obama 26%, John Edwards 12%
This fits with the general sense of what an analyst might tell you about polls. It’s much better to be in the lead - especially if there is a large gap to second place - but it does not have a 100% forecasting track record.
The prediction markets appear to back this up. Despite Joe Biden’s lead in the polls, they factor in enough other pieces of information that leads them to think that Elizabeth Warren is the most likely nominee.
The contrast with Bernie Sanders is remarkable. Despite similar positions in the polls, the markets think she is more than twice as likely to win the nomination.
What is that information they’re including? What are they thinking is the key factor that makes Warren more likely to be the nominee than her polling peer? What is the leading indicator that they are seeing?
Kamala’s bubble
Thinking this over, I’m reminded of the surge in Kamala Harris’ odds after the first Democratic debate.
It was the single largest jump I’ve seen in any political betting market. The only analogy was the spike in Boris Johnson’s odds when Theresa May announced she would resign and it was clear the next PM would be chosen by the Conservative Party members and not in a General Election.
Even before that jump, I wrote that “There is a clear conception across the betting agencies that despite her polling numbers, Harris has strong fundamentals… Their odds suggest that Harris is well-positioned to sustain her campaign through the early primaries, and that her numbers will pick up as 2019 heads into 2020.”
In that first debate, Harris attacked Biden’s record and pulled off a moment that went viral. From the odds’ reaction to that moment, it looked as if the market was waiting for it. They clearly thought that Harris could be a frontrunner. They were holding off only for the first piece of evidence, the first thing that looked like a canary. As soon as they saw it, they poured their money into the markets.
The Brexit comparison
The problem, at least in this circumstance, was that it wasn’t a leading indicator. Since then, Harris has fallen into fourth place, below where she was before the debate. The major shift that they thought was imminent never actually came.
This leads us inevitably to Brexit, the drain down which all political analysis these days flow.
Boris Johnson has had the worst start to a premiership in centuries. He has lost every vote in Parliament. A Scottish court ruled that his prorogation of Parliament was illegal. His own brother implied that he’s hurting the nation. The London-based media are scrambling for the thesaurus to describe how badly he’s done so far.
And yet here are where the politics lie.
Westminster voting intention: CON: 32% (-3) LAB: 23% (+2) LDEM: 19% (-) BREX: 14% (+2) GRN: 7% (-) via @YouGov, 09 - 10 Sep Chgs. w/ 06 Sep
September 11th 2019
399 Retweets773 Likes
According to Electoral Calculus’ prediction model, this set of polls would translate to the Conservatives picking up 36 seats. After all this, Johnson would actually gain seats.
Now for the caveats. This model is a rough approximation. Drops in Conservative support may be concentrated geographically in vulnerable areas. Tactical voting could help Labour and the Liberal Democrats. The expelled Conservatives may stand as independents. The chaos of Westminster may not have filtered through to the public. I certainly am not taking this set of seat predictions to the bank.
And yet, it makes me take a step back.
The ultimate destination for Brexit is probably going to be determined by an election. Therefore, a potential canary would be related to that election.
So as long as Johnson is able to keep 35% of the vote locked down while the opposition is split among three parties, he is favored to remain Prime Minister.
If we see that electoral calculus start to change, or hear of a Remain Alliance, or a Brexit Party surge, or enough Independent Conservatives candidates standing, then and only then should we call Michael Lewis’ agent.
Until then, we aren’t necessarily seeing a leading indicator, because we aren’t looking at what is leading us to an indicator.
Political Risk Monitor
The PRM this week showed how diverse stories can impact the world of politics: Hurricane Dorian and the Bahamas, Brexit and Great Britain, Taliban peace talks and Afghanistan, and the death of Robert Mugabe and Zimbabwe. They were a wide array of questions, each of which deserves more space than is left in this newsletter.
----------
Thanks for reading. If you like this newsletter, please pass it along!
Chris
Breaking mental models with the Queen
Upending and sanctioning democracies
Chris Oates Aug 29, 2019
There are three major stories this week that speak to how to analyze political risk and how doing so can impact the future of the world.
Before I do, a reminder to pass along this newsletter. The best way to help out this kind of analysis is to grow the number of people who read it.
European standoffs
Prime Minister Boris Johnson has received the Queen’s consent to prorogue (suspend) Parliament for five weeks. This is, according to one of the best UK political analysts, a constitutional crisis.
This feels like the constitutional crisis we’ve been expecting: PM saying he has to suspend parliament, Speaker replying it’ll happen over his dead body. The Queen is now in the hideous position of being the rope in tug of war between PM and MPs, between executive and legislature
August 28th 2019
1,046 Retweets2,011 Likes
Johnson apparently believes that this sets up a win-win for him. If Parliament uses the week before the prorogation to table a vote of no confidence and he loses it, he can call a General Election for soon after October 31 and campaign against Parliament to lock up Brexit Party votes (as I predicted last month).
If Parliament does not move against him, they may not have sufficient time to oppose a No Deal Brexit, which he seems okay with.
Either way, Johnson wins.
The betting markets appear to believe that this. The implied probabilities for the following events suggest that the markets see a General Election and a No Deal Brexit as the two most likely next steps.
· Second Referendum: 8%
· No Deal Brexit: 57%
· Article 50 revoked: 26%
· General Election in 2019: 70%
However, events in Italy tells us not to bank on this.
In Rome, the leader of the far-right Lega party attempted to bring down the government to force early elections. It was a similar move to what Johnson is attempting, a gamble that would lead to a consolidation of power.
Unfortunately for him, the fear of elections and a Lega government has led to his now ex-coalition partner and an opposition party forming a coalition of their own. He’s out of luck and out of power.
Salvini sidelined as M5S and Democrats agree Italy coalition Salvini sidelined as M5S and Democrats agree Italy coalitionParties to try to form government in move that would avert snap electionstheguardian.com
August 28th 2019
1 Retweet3 Likes
Boris Johnson may have stolen a march on his opponents yesterday. But only hours later, news broke that his actions drove one of his party’s leaders to quit, risking the party’s 13 seats in Scotland and any majority they might get.
Big blow to @BorisJohnson that @RuthDavidsonMSP is set to quit. @theresa_may only clung on to office thanks to the seats Davidson unexpectedly won for the Tories in Scotland. If those seats are now at risk, which they may well be with her resignation, a general election...
Chris Musson@ChrisMusson
EXCL: Ruth Davidson 'set to quit' as Scottish Tory leader due to family pressures and differences with Boris Johnson https://t.co/Hf75mB3Tfk
August 28th 2019
329 Retweets954 Likes
We’re not yet at the point where we can confidently predict what will happen in the UK. As Italy showed, power plays can backfire.
In a few months, we may see Boris Johnson with a newly-vindicated hold on Number 10 Downing Street, or shortest-serving PM of all time. Things are still very much in flux.
Sanctioning Brazil
A few weeks ago, I sent a text to a friend who works in politics asking if he thought the United States would levy sanctions on Brazil.
It was prompted, as you might guess, by reports of the deforestation of the Amazon. The rainforest that produces 20% of the world’s oxygen is being to a tipping point. If we go past that point, scientists project that it will enter a downward spiral called forest dieback. The loss of rainforest will allow drier conditions, which spark more fires.
In short, Jair Bolsonaro’s disinterest in protecting the Amazon could be the single largest defeat in the fight against climate change.
So why did the thought of sanctions on Brazil seem like such a crazy idea? Why did I have to think that rather than read it in the second paragraph of every story on the issue?
When Iran is rumored to have done anything in the Gulf, the punditocracy erupt with colons followed by a question mark about military action. When the single greatest threat to civilization is facing a threat in its single greatest vulnerability, it’s reported in the passive.
“The Amazon is burning,” not “Farmers and ranchers are burning the rainforest.” Or, to be more political, “Brazilian president threatens world with indifference to climate change.”
There’s two responses to this question, I think.
Predictions are history carried forward
First, it’s important to remember that those who predict politics do not have good imaginations.
That is, in part, to be expected. Most of the time, the probable event happens – that’s why it’s probable. Analysts are not paid to tell what they’d like to see, but to give a cold-hearted assessment of what is likely. The only data we have is in history, so what has happened in the past is the greatest indicator of what will happen.
But this doesn’t explain everything.
You can build a rigorous model while still keeping an open mind. Before the 2016 election, FiveThirtyEight was ridiculed for suggesting Trump had a 30% chance at the presidency. But they took the possibility of a popular vote-electoral college split seriously and reflected that in their uncertainty levels.
Generally, those in the analysis business have strong incentives to stay within conventional wisdom. Few people remember those who were wrong along with everyone else, such as every TV producer who overhyped the Ebola outbreak before the 2014 midterms even though the risk to the United States was low and it was a campaign tactic.
But we do remember those who were wrong and out of step with everyone else. Without having to Google “bad political analysis,” this column came to mind. Peggy Noonan wrote that Mitt Romney would win 2012 because “All the vibrations are right” and she saw a lot of yard signs.
People rely on mental models, but these models often calcify and constrict thinking. Iran is the type of country the US sanctions, so we go to that option on the slightest provocation. Brazil, a free market democracy, is not the type of country we sanction, so we don’t think about it even when the government’s actions merit that response.
It’s important sometimes to step back from the details of an issue and think to oneself, “If a generic country was doing X that affected other countries in Y way, the most likely global response would be Z.” It may be the same result as before, but it at least forces an analyst to come to that conclusion on the merits, not just before it sounds right.
At a tipping point already?
The second response to the question of why sanctioning Brazil seemed impossible is that, perhaps, I underestimated global leaders.
Since that time, the Amazon has become a central point of debate at the G7. Senator Brian Schatz has promised to act to use Congressional appropriations to act.
Anthony Adragna@AnthonyAdragna
Dems are looking to spur Brazilian action on Amazon fires. Those efforts, led by @brianschatz, could include everything from barring commercial space launches (~$300 billion potential industry) to barring Brazilian officials' travel. Leading Source of Policy News Analysis & Coverage | POLITICO Prosubscriber.politicopro.com
August 28th 2019
4 Retweets6 Likes
It has also made Brazil the second-fastest rising country in the Political Risk Monitor, behind only France, whose numbers are high because it hosted the G-7 summit.
While I was in London, I even stumbled across a protest at the Brazilian embassy of a size usually only seen outside of authoritarian regimes with human rights abuses.
The confluence of the hottest summer on record, the leadership of French President Emmanuel Macron, grassroots efforts like that of Greta Thunberg and the Sunrise Movement, and Jair Bolsonaro providing an intensely disliked figure to rally against may have brought the global community to a point where action actually happens.
The initial commitment of only $20 million to a fund for Amazon conservation is less than required, and it has been rejected by Brazil, but it could be the data point that allows analysts to create a new mental model.
Thanks for reading. If you like this newsletter, please pass it along!
Chris
Local political risk
Of school boards and trade wars
Chris Oates Aug 15, 2019
This is an August full of political risk.
New tariffs on Chinese products have been announced and delayed.
Italy may be heading to early elections in the fall that could elevate the populist Lega to control of the government.
Hong Kong is protesting and President Trump has revealed that US intelligence is tracking Chinese military movement to that city (which cannot be what the intel agencies wanted him to do).
Donald J. Trump@realDonaldTrump
Our Intelligence has informed us that the Chinese Government is moving troops to the Border with Hong Kong. Everyone should be calm and safe!
August 13th 2019
24,573 Retweets85,042 Likes
However, I want to take this issue to highlight an episode of political risk that was completely missed by the industry when it happened: a vote by School Committee of Somerville, Massachusetts, not to approve a new high school.
The story of Powderhouse Studios
The full details of this episode is contained in this excellent Boston Globe article from the Hechinger Report. I’ll recap the main points as they pertain to political risk.
In 2012, a non-profit that ran an educational summer camp was noticed by the mayor of Somerville, MA, a city of 75,000 bordering Boston and Cambridge. He encouraged the founders to consider expanding their work (using projects to teach STEM and the arts) into an entire school. He recommended using the Innovation Schools initiative as their framework. This was a part of the state’s educational system which allowed for non-traditional approaches most commonly used at charter schools to be developed within the public school system.
In 2016, the group won a $10 million grant from XQ, a foundation supported by the widow of Steve Jobs. This would help them fund their school, which they were calling Powderhouse Studios.
Everything seemed good for Powderhouse Studios. It had the support of the Mayor, would win the approval of the teachers union in January 2019, and major financial support. The plan to eventually teach 160 high schoolers who preferred this approach was on track.
Then, at a School Committee meeting on March 18, 2019, when final approval was supposed to come, the plan was voted down unanimously.
As the Globe described the vote: “As recently as January, no one would have predicted such an outcome. But there it was.”
Is this political risk?
This is not the usual example of political risk. It is not a coup, a terrorist attack, a populist president, or an escalation of a trade war. But if you break down the elements, it bears all the similarities of many definitions of political risk:.
Does it involve politics? Yes. The Mayor and the School Committee are elected officials and the decision was part of the normal process of politics and governance. The origins of the plan depended on a government initiative.
Did political actions impact an organization/business? Yes. The vote killed the idea for now. About $1 million of a $10 million grant has been spent without results.
Was this an area where predictions about politics shaped people’s decisions? Yes. The article states that the grant was given because there seemed to be political support for the school. The creators of the school probably wouldn’t have worked so hard on the idea if the mayor hadn’t encouraged them to. Assumptions about the political landscape were crucial for the entire episode.
Lessons for political risk
I don’t know who is responsible for the reversal in political support for this plan or what’s happening next. The article makes it sound as if there was miscommunication about what the funding could be used for and concerns that the main high school would be hit with budget cuts to accommodate the new school.
But it does highlight how political risk pervades areas far removed from Italian no confidence votes and tweets about Chinese military movements. An idea for a new high school was dependent on a political process and a miscalculation about political support meant that a million dollars and countless hours were possibly wasted.
Even at the most local level, political risk can be pervasive. It’s a reminder that as the industry focuses on international trends and patterns, political risk can be happening right down the street.
Political Risk Monitor
Movements in Hong Kong have propelled China to the top of the overall PRM table and actions relating to Kashmir have made India the top mover of the past two weeks.
Other notable countries for political risk in the past two weeks have been Kyrgyzstan, where an ex-president has been arrested, Ireland, at the center of Brexit debates, and Brazil, where Bolsonaro has attracted attention for the devastating rate of deforestation in the Amazon.
Bookie Tracker
The betting markets are showing high levels of uncertainty about UK politics right now.
The market to be the next Prime Minister reset after Boris Johnson took over at Number 10 last month. We are now asking who comes next after Johnson, and the market is saying that it does not know.
Labour leader Jeremy Corbyn has been given an average implied probability of 28%, followed by Labour Shadow Brexit Secretary Keir Starmer at 9%, Liberal Democrat leader Jo Swinson at 8%, Brexit Party leader Nigel Farage at 7%, and Labour Shadow Business Secretary Rebecca Long-Bailey at 5%.
The top ten contenders to be the next PM, according to the betting markets, have a combined probability of 80%.
Back in May, before Theresa May announced she was stepping down, the top ten candidates had a combined probability of 99%.
In other words, the betting markets in May were relatively confident that they knew who the next prime minister would be out of a shortlist of contenders. Now, they are saying (in effect) that the second-most likely candidate is a shruggy emoji.
This market overall is a vote of low confidence for Corbyn.
Despite being the leader of the opposition, other members of his party are strong contenders to be the next PM. His relatively low probability, despite forecasts of elections in a few months, is a sign that the markets think that there is a very good chance that Conservatives will either win enough seats to form a government, that his own party will replace him before Labour is able to take power, or that a non-Conservative government will form without him in charge.
Thanks for reading. If you like this newsletter, please pass it along!
Chris
Predicting what will predict Brexit
Parliamentary math on the future of the UK
Chris Oates Jul 25, 20191
Boris Johnson became Prime Minister of the United Kingdom and Northern Ireland yesterday.
As the betting markets showed, he had been the presumptive favorite since the leadership race was announced. He quickly took a lead and, except for a drop that coincided with Michael Gove’s bounce, maintained his lead through a Jeremy Hunt bubble, a Rory Stewart boomlet, and Jeremy Hunt again.
Three years after the referendum on Britain’s EU membership, it has a PM who campaigned for the winning side. As Johnson put it in his acceptance speech, “we are going to get Brexit done.”
"I say to all the doubters: Dude, we are going to energise the country, we are going to get #Brexit done" New Tory leader Boris Johnson says as PM he will bring "a new spirit of can-do" to the country #NextPrimeMinister updates: bbc.in/32VonES
July 23rd 2019
748 Retweets1,559 Likes
This raises the question: How?
John Springford @JohnSpringford
ICYMI, I can't see a middle ground that's acceptable to a Johnson led government and the EU. We're heading for no deal or a general election Boris Johnson and Brexit: What to expectThere are no compromises on the backstop acceptable to the EU or a Johnson-led government. A general election fought by the Conservatives on a no deal ticket is therefore very likely.cer.eu
July 23rd 2019
31 Retweets43 Likes
Politics is arithmetic
I’m regularly annoyed the journalistic tendency to portray politics as a series of dramatic narratives and heated policy debates when, most of the time, it’s about simply counting votes.
Brexit is no different. There are only three outcomes for what could happen.
Britain leaves the European Union without a deal on October 31.
Britain leaves the European Union with an agreement.
Britain does not leave the European Union.
The problem with the Boris coverage is that choosing one of these options almost certainly requires a majority in Parliament.
The House of Commons has voted against a No Deal before and has taken steps to ensure that Johnson cannot follow that path against its wishes.
MPs vote to block the next UK PM shutting down Parliament in order to push through a no-deal #Brexit, by 315 votes to 274 bbc.in/32BneSI
July 18th 2019
1,557 Retweets5,032 Likes
Parliament has also voted against the Withdrawal Agreement as negotiated by now-almost-former PM Theresa May.
Parliament has also voted against holding a second referendum or repealing Article 50.
The question for Boris Johnson is not, therefore, what he will do. It’s what he can get Parliament to agree to when it currently does not agree on any approach.
Complexity within a simple question
Okay, so we’ve gotten Brexit from the wooly coverage of the PM campaign down to the central question of what would change the parliamentary math. Now we run into the problem of predicting the actions of 650 individual politicians, many of whom are not following party leadership, limiting our ability to forecast their actions. They are now led by a politician whose former boss said he “would not take Boris’s word about whether it is Monday or Tuesday,” so we can’t rely on his statements.
However, political risk analysts can’t simply throw our hands up and be done with it. Our job is to provide clarity about what is more or less likely, and what we would be likely to see en route to that direction.
We have to be able to start to unravel the threads, or at least clarity what are the worst tangles.
Forecasting Parliament
One of the most detailed approaches I’ve seen comes from the writer and analyst Jon Worth, who has created a series of flowcharts, with probabilities at each decision point, to generate overall probabilities of the different outcomes.
The latest full chart can be found here and I’ve included a snippet below of what would happen after a vote of no confidence in Parliament. The highlighted numbers are the probabilities that each path would be taken from the previous scenario.
Worth argues that the highest probability is a snap General Election before the fall. If the Conservatives take a majority, they could push through the Withdrawal Agreement with enough votes to spare, especially if the party campaigned on it and new members felt an obligation to do so.
Other analysts are starting to talk about that possibility and bookies are now offering shorter odds on an election in 2019 than in 2022 when they’re next scheduled.
Five-party politics
Unfortunately for analysts, even being correct that there will be a General Election doesn’t do much to tell us what will happen. UK politics is extraordinarily scrambled right now. Polls show that four parties are roughly equal in polling intentions, with the Greens in a historically strong position.
Westminster voting intention: CON: 25% (-) LDEM: 23% (+3) LAB: 19% (-2) BREX: 17% (-2) GRN: 9% (+1) via @YouGov, 23 - 24 Jul Chgs. w/ 17 Jul
July 24th 2019
457 Retweets1,016 Likes
However, the UK votes in a first-past-the-post system that means we cannot simply extrapolate national polling to seat totals. In 2015, the UK Independence Party received 12% of the vote for 0.2% of seats, while the Tories turned 37% of the vote into 51% of seats.
Electoral Calculus, a great polling analysis and forecasting site, finds that even small changes in vote totals could have huge swings in how many seats are won. A 1 percentage point difference in vote share is estimated to change the Conservatives seat total by 4% of seats.
This is especially important given the volatility in the electorate. In the 2017 election, Labour shot up the opinion polls during the campaign as previously disenchanted supporters rallied behind the chance to win seats and take control of the government.
Labour and the Conservatives now face the reverse situation. The European Parliament elections in May saw seats proportionally allocated, so the parties could easily fracture. The Brexit Party halved the Conservatives’ strength and the Liberal Democrats and Greens did roughly the same for Labour.
Forecasting a General Election is therefore almost impossible. If Boris Johnson is able to bring most Brexit Party voters to the Conservatives, bringing their national polling position close to 40%, while Labour and Lib Dems stay at 20-25%, the Conservatives could receive a huge majority government, according to Electoral Calculus’ prediction model.
If the reverse happens - or Labour, Lib Dems, and Greens form some kind of alliance and coordinate to prevent vote splitting - then Labour could receive an outright majority, or at least enough to form an anti-Brexit coalition with the Lib Dems and SNP.
This doesn’t give us the answer, but it helps to identify the central question: which side of the political spectrum comes together more rapidly if a General Election is called.
If we see that Labour is repeating its 2017 performance, that’s a sign that Johnson’s decision has backfired and a second Brexit referendum is likely. If the Brexit Party stumbles, outpaced by the deeper and wider electoral infrastructure of the Conservative Party, expect a big victory for the Tories and the Withdrawal Agreement passing.
These are two widely divergent scenarios, but both show the way to what matters most in the months before October 31.
This situation will be monitored throughout the summer and fall, and future issues will assess whether this analysis was correct or, like so many predictions about Brexit so far, wildly off-base.
Chris
The EU after Brexit is here
What the Summit taught us and Export Development Canada
Chris Oates Jul 11, 2019
The Beacon is focusing on the latest EU summit this week, with a Tip of the Cap to Export Development Canada.
Before we continue, pass the Beacon along if you like it! Every two weeks, more coverage of the political risk world and how it’s being analyzed.
The next phase of the EU
In what was a geopolitical lifetime ago, I was a doctoral student, examining the European Union’s foreign policy processes. Here’s one passage from my dissertation that I was reminded of last week.
After September 11, [UK Prime Minister Tony] Blair, [French President Jacques] Chirac and [German Chancellor Gerhard] Schröder held a separate meeting before the European Council summit on October 20, 2001, attempting to agree first on a position that satisfied the “Big Three” of France, Germany and Britain and then pressure the rest of the EU to agree with them. They planned to repeat the exercise on November 4, but the prime ministers of Italy, Spain, the Netherlands, and Belgium (which held the Presidency), and High Representative Javier Solana, “forced their way, almost literally, to the table.”
Britain in 2001
It is one of the rare instances when “almost literally” is correct.
Blair invited Chirac and Schröder to dinner at Downing Street so that they could come to an agreement between themselves on Afghanistan, but had to invite other important EU actors who refused to be left out of their discussions. London was, on that day, the center of EU politics that other European powers demanded to be invited to.
Throughout my dissertation’s case studies (non-Iraq foreign policy with American involvement from 2001-2005), Britain was a central player. Even as the Iraq War caused tremendous divisions across the continent and the opposition of its traditionally close allies, Britain was nonetheless a leader of a significant bloc. As a graphic from my dissertation shows (this was a very non-graphical thesis), the UK position had support from governments representing a majority of the member states or those soon to join.
July EU Council Summit
Last week, the European Council (elected leaders of the member states - the PMs and Presidents across the continent) met to discuss who would get the top jobs: President of the Commission (the EU’s executive branch), Council President (the person who chairs Council meetings), the President of the European Central Bank (in charge of monetary policy), and the head of the EU’s foreign policy.
The summit was a fascinating display of how quickly politics can change. The Spitzencandidaten system to link the Commission presidency to the European Parliamentary elections that happened in May fell apart, and various packages of candidates were proposed and rejected throughout the summit.
According to the FT, halfway through negotiations, “a visibly angry [French President Emmanuel] Macron had emerged from stalled talks to rail against the “divisions” and “hidden agendas” that made it impossible for the bloc to reach decisions. The EU was once again displaying its vacillation at a time when the world around it was in upheaval.”
After an all-night session that broke so that the heads of government could grab some sleep, a new set of candidates was proposed and accepted. Macron was now happy with a slate of candidates who would likely support his vision of further integration within the EU for the next five years.
Britain-less Europe
The FT and Politico have great breakdowns of the summit. Among the crucial actors for determining who will run the Union were the following:
Macron
Members of European Parliament
The Visegrad 4 (Czech Republic, Slovakia, Hungary, and Poland)
Prime Ministers affiliated with the European People’s Party (Ireland, Croatia, and Bulgaria)
Various German politicians, from Angela Merkel to her coalition allies
Notable in this is the absolute absence of Britain. PM Theresa May was present at the summit, since the UK is still a part of the EU, but the only headline I could find of her participation was the following, written by The Sun’s Associate Showbiz Editor.
Of course, the decline of UK influence in the EU didn’t happen all at once, and the 2001 dinner on Afghanistan wasn’t a normal meeting (the UK was especially important in foreign affairs, and more so when it was a transatlantic issue). But it’s still remarkable how total the drop off has been.
It also shows the effect of the 3rd largest member state being sidelined. France and Germany have more clout and those pushing back against faster integration have lost their largest ally.
The EU is still at an unstable moment. The UK could choose not to reverse the Brexit decision. Macron’s domestic situation could sap his influence in Brussels. A recession could tip the eurozone into another crisis.
But at the moment, it looks as if the EU is on the cusp of a new integrationist moment, if only because one of the major forces arguing against integration has decided to leave the table and head for the sofa.
Tip of the Cap
This article in Risk Management from Derek Baas details how Export Development Canada developed their approach to political risk assessment. It is one of the rarest specimens in the political risk world - a real-life case study about how an institution has dealt with the topic over decades, complete with diagrams and classification of countries.
What jumped out to me was how they classified countries for risk scoring. The classifications were likely right at the time they were made (it appears to be the mid-2000s), but significant changes have happened to many since the list was made. Hungary, Syria, Egypt, Ukraine, Tunisia, and Turkey were cited as representative examples of their classifications, but have all seen major disruptions since then.
It’s a reminder that even in rigorously designed risk assessment systems, frequently updating the metrics is as important as overall methodology.
Thanks for reading. If you like this newsletter, please pass it along!
Chris
Presidential assumptions
Warren, War, and Worrying about Journalism
Chris Oates Jun 26, 2019
Two big stories in the Beacon this week.
What Elizabeth Warren’s rise means for betting markets.
Are we going to war with Iran?
And a Tip of the Cap to an article showing why journalists need forecasting skills.
Warren upending the bookies
When I started a newsletter, I knew that this would happen sooner or later. I’d find something interesting, write up what it means, hit publish, and then watch as events proved me completely wrong.
In the last issue, I wrote about how Kamala Harris’ performance in betting markets was far higher than her polling numbers. It showed, to me, that the experts saw fundamental strengths to her candidacy that they didn’t see in other candidates like Elizabeth Warren.
That didn’t last long.
The story of June in the betting markets has been the rise of Warren from a distant fifth place up to second.
What’s most interesting to me is that this is not simply a reflection in polling. Warren is rising, but her average of 12% in RealClearPolitics’ tracker still trails Sanders’ 16%, and the other major candidates are within a point of their averages in early June.
Instead, it appears that the conventional wisdom - to the extent that the markets reflect it - has flipped. Warren had for months been given odds lower than the position polls would suggest. Perhaps it was the ancestry issue, perhaps the image of a Harvard professor, but whatever it was, there was an assumption that she faced a steep climb. Now, the opposite is true. Despite still trailing Sanders in the polling averages, the assumption is that she has the inside track of non-Biden candidates.
This shift in the odds gives us a couple insights into how the betting markets treat politics. It’s certainly still early in the overall process, but here are three takeaways so far.
1. The “imperceptibles” of politics can shift quickly.
Warren is not some unknown quantity. She is running on a platform of reforming the economic system, which was broadly the throughline of her 2018 Senate run, her 2012 Senate run, her 2008 work on the bank bailout oversight panel, and her academic work dating back to the 1980s. If you asked a political analyst in January to predict what a Warren campaign would look like, this is pretty much it.
What changed was the perception of a policy-driven campaign from a liability to an asset. There was certainly an impetus in this shift, as some good polling in early June prompted re-evaluation. But mostly it was the result of analysts realizing that their assumption may have been incorrect, and rapidly revising their forecasts.
Predictions are built on assumptions and these are often not subject to evaluation until it becomes too apparent they’re flawed. When the unexpected happens in politics, it’s often not due to a crazy fluke, but the result of clear evidence being discounted. Brexit, for example, was a shock because undecideds (more than 10%) were assumed to be breaking for Remain.
Of course, this could all change if another assumption looks shaky, like polling being the best predictor of success.
2. Betting markets can help us see lanes.
Warren’s rise came at the expense of only a few other candidates. From June 1 to June 26, the top 6 candidates moved by:
Warren: +9
Biden: +0.5
Sanders: -5.5
Harris: -2.7
Buttigieg: +1.3
O’Rourke: -2.2
Warren’s increase roughly matches Harris’ and Sanders’ combined decrease. Her odds are more closely correlated with Sanders’ - the days of her biggest jump coincided with his sharpest drop - but she has certainly eaten away at Harris’ chances as well.
Betting markets can show us a more fine-grained look at the traditional idea of candidate lanes. If the above patterns continue, we will see that Warren is about twice as much in Sanders’ lane as she is in Harris’, and we can make inferences about what will happen depending on the order of candidates dropping out after the early primaries.
3. We still don’t know about Biden.
What’s remarkable about the dramatic movement for three of the top five candidates in the past month is how little it’s affected the frontrunner, who has been chugging along steadily since he announced.
There hasn’t yet been any event to seriously prompt a re-evaluation of his chances higher or lower. We’ll soon see if the debates serve that role, or if his lead will remain stable into the weeks leading up to Iowa.
Political Risk Monitor
The story of the past two weeks in the risk world has been the US and Iran, with Oman dragged into it as well.
There’s no other way to put it. This story is scary. The President of the United States almost launched a military strike against a country that has the capacity to strike back and we are still not sure what process was involved with the decision.
Breaking News: President Trump pulled back from strikes against Iran late Thursday, hours after approving them in retaliation for the downing of a U.S. drone Strikes on Iran Approved by Trump, Then Abruptly Pulled BackMilitary and diplomatic officials still expected a strike as of 7 p.m. Thursday in retaliation for Iran shooting down an unmanned American drone.nyti.ms
June 21st 2019
4,738 Retweets7,830 Likes
The two countries are now locked in a rhetorical escalatory cycle.
Iran's president is mocking Pres. Trump, going so far as to say that the White House is "afflicted by mental retardation." The Latest: Saudi Arabia says it shot down Yemen droneThe Latest: Saudi Arabia says it has shot down a drone launched into the kingdom by Yemen’s Iranian-allied Houthi rebelsabcn.ws
June 25th 2019
821 Retweets2,766 Likes
Donald J. Trump @realDonaldTrump
....Iran’s very ignorant and insulting statement, put out today, only shows that they do not understand reality. Any attack by Iran on anything American will be met with great and overwhelming force. In some areas, overwhelming will mean obliteration. No more John Kerry & Obama!
June 25th 2019
21,517 Retweets83,231 Likes
The question remains whether any of these words will turn into action. A detailed analysis of the situation requires a longer space than we have here, but it does highlight one very important distinction: probability vs certainty.
The probability of a military conflict with Iran remains low. The costs involved - in lives, resources, and regional instability - are higher for both sides than the benefits. If the US strikes Iran, it will accelerate Tehran’s pursuit of nuclear weapons. If Iran attacks US forces and neutral ships, it will alienate potential partners in alleviating sanctions.
But the certainty of that prediction is also low. While, true, it doesn’t seem like a good idea for either side to attack the other, people make mistakes all the time. The close call last Friday could easily have happened and we would right now be in a very different situation.
We are now in a gray area where military conflict is well within the realm of possibility and could happen before any analysts have time to parse the tea leaves.
Tip of the Cap
Helen Lewis’ article in the New Statesman gets at the heart of why political risk and rigorous forecasting matters, even in journalism.
Miscalling the 2015 & 2017 election results. Complacency over the rise of Trump. And the vox pops that skew our idea of “real Britons”. Why political journalism gets it wrong, from the “view from Versailles” to the Confident Posh Man problem: Why political journalism keeps getting it wrongWhen I started at the New Statesman in December 2010, we were six months into Britain’s first experience of coalition government since the 1940s, a surprisingly strong alliance between David Cameron’s Conservatives and Nick Clegg’s Liberal Democrats. Since then, we’ve had two more general elections,…newstatesman.com
June 20th 2019
143 Retweets378 Likes
Lewis sums it up clearly, drawing on her experience in the 2015 UK general election:
“In 2015, we assumed that the Conservatives could not win an overall majority. The likeliest result, most journalists thought, was a Labour minority administration backed up by the SNP. That affected the questions we asked: Ed Miliband, would you do a deal with the Scottish Nationalists? Would you give the SNP another referendum? How much money would you give to Scotland to get their votes? And it also affected the Conservatives’ “squeeze” message to bolster their own support…
Cut to election night 2015. The exit poll arrives – and it’s an overall majority, albeit it a slim one, for David Cameron. All those questions about the SNP were not just redundant, but might have influenced the outcome. At the same time, insufficient consideration was given to what the Conservatives might do with a majority.”
Good forecasting is important not because we want to know what will happen. It is important because if it doesn’t occur, it will be replaced by bad forecasting.
This is one of the most important lessons I teach in my company’s workshops: we are always acting on the basis of our predictions. A hedge fund that has an Asia desk but not an Africa one is predicting that Africa won’t be worth their investments. A government that doesn’t invest in public transit is predicting that their current infrastructure will be sufficient for the next ten years. And a journalist asking a question is making a prediction that other questions aren’t more important.
Of course, some journalists also make explicit predictions and their need to be better is even more alarming.
Totally coherent Iran policy
June 18th 2019
976 Retweets5,718 Likes
Thanks for reading. If you like this newsletter, please pass it along!
Chris
Predicting the outcome of secret agreements
Mexican trade wars, visualizing uncertainty, and Kamala Harris' enduring odds
Chris Oates Jun 13, 2019
One of the great parts about being in political risk is that there’s always something new.
I once wrote an article based on Queen Elizabeth’s husband receiving an Australian knighthood. You may have a day jumping between predicting for next month and for 2045. A project might require a deep dive on countries on every continent.
But I don’t think I expected that I would analyze an international agreement that we only know about because the President of the United States waved around a piece of paper and a Washington Post photographer was able to make it semi-legible.
“the Government of Mexico will take all necessary steps under domestic law to bring the agreement into force with a view to ensuring that the agreement will enter into force within 45 days.” @realDonaldTrump #Mexico agreement. Second photo flipped @washingtonpost @postpolitics
June 11th 2019
2,155 Retweets3,558 Likes
What are we to make of this?
On the one hand, we should take this seriously. This agreement appears to have been crucial to warding off escalating tariffs on US trade with Mexico, which could have cost 400,000 jobs and $41 billion. If this agreement was the deciding factor, and a huge amount hinges on whether it’s enacted, then it’s important that we understand what’s in it.
On the other hand, how much can we make out of a document that the Washington Post can only get to this level of specificity?
“[UNREADABLE] other party. The parties further intend [UNREADABLE] an agreem[ent] [UNREADABLE] to burden-sharing in relation to the processing of refuge[es] [UNREADABLE].”
Mexico’s Foreign Minister has denied that there is an agreement. He characterized the recent talks as ending with a 45-day delay to show that existing practices are working. Another 45-day period for negotiations would follow if the US objected to the results. It’s less of an agreement than a 90-day cooling off period. This account broadly fits with the one paragraph we can read from the photographed paper.
So how to interpret an agreement that isn’t an agreement, where the only sources we have are two politicians with their own agendas and a few fragments snapped by a fast-thinking photographer?
What happened last week and what will happen in 45 days? A few points:
In the absence of hard evidence, we should go with the most likely scenario that fits the facts. There was significant pressure to avoid the tariffs and this agreement was a largely face-saving measure to get the US to back down. Mexico made some commitments, but nothing major and it avoided potentially devastating tariffs.
Migration flows tend to drop during the summer, so both sides will likely be able to say that the agreement worked without any other significant changes.
Those two points suggest that there will not be a renewed threat of tariffs in August. However, Trump appears to believe that the agreement is significant and was reached via the threat of tariffs. It could be that this becomes a template for other situations.
In short, we don’t know what will happen. But my best guess is that this could be the end of one threatened trade war and the justification for many others.
The next one may be here by the end of the month.
President Trump says he will raise tariffs if China's leader Xi Jinping fails to meet him at the G20 summit cnn.it/2R9UgnK
June 12th 2019
17 Retweets38 Likes
Political Risk Monitor
The PRM was focused on the US-Mexico trade war over the past two weeks. Mexico was the top mover, getting up to third-place in the overall rating. That’s its highest position yet, a reflection of the consternation that the almost trade war set the political risk industry.
Elsewhere, Sudan re-emerged as a major topic of discussion, as Khartoum saw a general strike and attack by paramilitary forces on civilians. The relative discussion between Mexico and Sudan is a reflection of the political risk world’s priorities and coverage choices. The possibility of tariffs saw far more discussion than the reality of fatal protests.
That is not to condemn all coverage of the trade disputes - the leading section of this newsletter would not be able to hold up against that. But it’s to remember when tracking the political risk industry that there is a constant bias towards matters that affect wealthy nations, international trade, and ones that are speculative. It is relatively easy to write a column based on a source at a DC think tank about what might come. It is hard to report what is happening in Sudan right now. Articles from there, and the reporters that make them happen, should be appreciated all the more.
Tip of the Cap
Jessica Hullman of Northwestern and Matthew Kay of the University of Michigan have a project focused on visualizing uncertainty. Their first two entries on Medium are great studies of one of the most fundamental problems of political risk: how do we accurately communicate what we know, while keeping the other person aware of what we don’t know.
Jessica Hullman @JessicaHullman
New on Multiple Views: First two posts in @mjskay's and my series on uncertainty visualization We start with what we mean by uncertainty in vis, then walk through techniques from vis research and beyond, discussing trade-offs bit.ly/2Ms5tkBUncertainty + Visualization, Explained (Part 2: Continuous Encodings)Matthew Kay and Jessica Hullman (the MU Collective).bit.ly
June 4th 2019
81 Retweets266 Likes
This problem came up a lot in 2016. I remember being asked frequently what would happen with the Brexit referendum and the US presidential election. I, like most others, forecast Remain and Clinton victories. I was much less confident about the Brexit vote, because of the high number of undecided voters in the polls, but it was difficult to describe this context succinctly.
“Remain will probably win, but there’s a strong chance that Leave wins. Clinton will also probably win, but Trump could, but that chance is less than Leave, but the two aren’t related, so Brexit going the other way shouldn’t be an indicator...” does not trip off the tongue. You probably got bored halfway through that sentence.
When we try to visualize information, we further strip out the chance for context, as a general sentiment is reduced to a single point on a scatterplot. Hullman and Kay’s work on new graphical displays, like bell curves in a line chart, combine context and information in a single view. It’s a fascinating approach and I’m looking forward to what else they come up with.
Bookies bullish on Kamala
One of the constant disconnects between the Bookie Tracker and polling averages has been Kamala Harris.
Despite now trailing in the polls behind Elizabeth Warren and tied with Pete Buttigieg, she has consistently led both in the Bookie Tracker.
A simple metric is the ratio between polling averages and odds from the betting agencies. For Biden, Sanders, and Warren, it’s roughly equal. For Harris and Buttigieg, it’s more than 2:1. Both candidates are projected to win the nomination much more than their polling numbers say, at least compared to their rivals.
Kamala’s position becomes unique when including the markets for the Vice Presidential nomination.
I am not putting a huge amount of faith in the VP markets yet. Tulsi Gabbard, Hillary Clinton, and James Mattis are not credible candidates. But Harris’ position here, combined with her overperformance in the more liquid presidential nomination markets say something about how the expert oddsmakers and the markets view her.
There is a clear conception across the betting agencies that despite her polling numbers, Harris has strong fundamentals - the same that would lead to her being chosen for VP. FiveThirtyEight’s analysis of the five corners of the Democratic Party concluded that Harris had the widest appeal to the constituencies that would decide the race. The betting markets appear to reflect that.
Their odds suggest that Harris is well-positioned to sustain her campaign through the early primaries, and that her numbers will pick up as 2019 heads into 2020.
Thanks for reading. If you like this newsletter, please pass it along!
Chris
How the Champions League explains Australia’s election
Austria, Race for 10 Downing, and Abraham Lincoln’s political risk index
Chris Oates May 30, 2019
This month has seen some major upsets.
First, Liverpool came from behind against Barcelona in the Champions League semi-finals. Down 3 goals in the first leg, they scored 4.
Congratulations to @LFC. It was an absolute delight to be there to witness one of the greatest nights in the history of European football. A truly astonishing performance. 👏👏👏
May 7th 2019
7,567 Retweets94,653 Likes
Then, Tottenham, down 3-0 at halftime of the second leg, scored a 96th minute winner against Ajax, almost destroying the BT Sport studio.
Tottenham Hotspur @SpursOfficial
🤯🤯🤯 Scenes in the @btsportfootball studio! #UCL ⚪️ #COYS https://t.co/xV8cDOv2P9
May 9th 2019
2,286 Retweets13,183 Likes
Finally, the Coalition won the Australian federal elections despite trailing in every poll.
#UPDATE Australia's ruling conservative coalition appears to have secured a shock election win Saturday, with national broadcasters predicting the party has defied expectations to retain power u.afp.com/JMdv
May 18th 2019
28 Retweets21 Likes
Obviously, these three events are very different, but all of them were described in similar terms: huge surprises against prior expectations. Everyone who followed them knew what was most likely to happen: a Barcelona-Ajax Champions League final and Prime Minister Bill Shorten.
But - of course - most likely is not a 100% certainty.
One of the main difficulties that political risk faces is how to talk about probabilities attached to forecasts when the event we’re discussing will be a one-off.
What does a 60% chance of something happening mean compared to a 65% chance? What about 70%? 75%? 80%? When does a toss-up become a likelihood and when a likelihood become an inevitability?
It’s a difficult question to answer, but it’s also one that we all instinctively understand.
If you had told a soccer/football fan who hadn’t followed the Champions League that Spurs overcame a 1-0 deficit in the second leg of the semi-final to win, they would have been intrigued, but not overwhelmed. Teams do that all the time.
If you had told them that Spurs scored 3 goals in the second half to win, that fan would have been flabbergasted. That doesn’t happen often.
If you had been watching the game with the clock winding down and Spurs still needing to score, you’d have been running around the building like Steve Nash when Lucas Moura put away the deciding goal. That much on the line and that little time left - that doesn’t EVER happen.
The reaction from huge Tottenham fan @SteveNash when that goal went in 😂 https://t.co/g0NRmIVdY6
May 8th 2019
2,098 Retweets8,891 Likes
So without being explicit about the probabilities, we roughly know the relative chance of events. And the numbers given by analysts match the relative surprise fans felt:
Tottenham winning (down 1-0 at start of the game): 25%
Liverpool winning (down 3-0 at start of game): 7%
Tottenham winning (down 3-0 at halftime): 4%
Tottenham winning (down 1 goal with a minute left): Almost 0%.
The amazing thing about the Australian election is that, despite coverage saying it was a “miracle”, the odds of Scott Morrison winning another term was around 20%.
The huge shock was closer to the well-within-the-realm-of-possibility odds of the start of the Tottenham-Ajax game than the minuscule chances that produced such strong reactions elsewhere. For American sports comparisons, Australian Labor was like being up 2 runs in the top of the 5th inning in baseball or up by 7 points at the start of the 4th quarter in basketball. Fans would be confident, but a loss would not have been met with bewilderment.
So why did Australia seem like such a shock?
There are a number of possible reasons: elections matter much more than sports (though some may debate that); the polls had been relatively consistent for so long; news coverage framed the election as a done deal.
At the top of the list - and at the center of political risk - is sample size. Anyone who has watched more than a handful of games has seen one where the 20% underdogs came back to win. We have in our bank of memories a large number of comparisons and our reactions are roughly in accord with the result.
When there are fewer presidential elections in a lifetime than baseball games in a weekend, we can’t use that same technique.
We can’t easily distinguish between an improbable win for the Coalition and a shock. Similarly for the United Kingdom with Brexit (20-40% chance in the weeks before the referendum) or the US with Trump’s election (29% according to FiveThirtyEight).
There’s no easy way to rectify this situation, other than with an increased appetite for understanding the uncertainty inherent in predictions - maybe with another data point for surprise when another underdog wins on Saturday in Madrid.
Tottenham Hotspur @SpursOfficial
📍 🏴...✈️...🇪🇸 📍 #MarchToMadrid ⚪️ #COYS https://t.co/zNn7a9usdq
May 29th 2019
553 Retweets4,597 Likes
Political Risk Monitor
The trade wars have driven China to the top of the Political Risk Monitor, and tensions with Iran kept that country in the top five, while elections have seen India, Britain, France, Europe, and Australia in the top 10.
The biggest mover of the past two weeks has been Austria, where Ibizagate and a no confidence vote drew attention. The impact of the European elections on the political risk world can be seen with half of the top ten movers coming from the continent. It’s a feature of the field that - although we knew the contours of the polls going into the election and we still don’t know which group will supply the European Commission president - analysis is concentrated around pivotal news events.
Tip of the Cap
There have been few times as uncertain and volatile in American history as the Civil War. Thanks to a fascinating paper from Kristen Willard, Timothy Guinnance, and Harvey Rosen published in 1996 that I recently read, we now know what parts were most volatile for those living during it.
The three examined the market for converting dollars to gold from 1862 to 1879 and used it as a proxy for political volatility. If the war was going badly for the Union, costs would rise and the likelihood that the government would redeem greenbacks at fair value would decrease. If the war was going well, the value would rise. Traders in the market paid correspondents near the front to telegraph news as quickly as possible, and the money at stake made it one of the first places where reports from the front would be known. Newspaper correspondent Kinahan Cornwallis described it as such:
“Immediately after the receipt of news of an unexpected victory, or defeat, the Babel of voices in the ‘Coal Hole’ arrested the attention of passers-by in William street, and through the tremendous uproar and the confusion of sounds, some voice could be heard louder than the rest, exclaiming, perhaps -- "I'll give a half for a hundred -- half for a hundred-- five-eighths for a hundred." "Sold" would be the responsive cry of some one near.”
The paper finds that there were three perceived turning points in the war: events that the market did not anticipate but lead to lasting changes in perceptions about the war.
Antietam and the Emancipation Proclamation (indicating longer war)
Gettysburg and Vicksburg (shorter war)
Jubal Early’s retreat from Maryland in 1864 (shorter war)
Major battles that dominate history books - Second Bull Run, Chancellorsville, the Wilderness - certainly mattered. But they had been discounted by the market in advance and didn’t drastically change the price of dollars.
The paper is fascinating, as well as Cornwallis’ short book. It’s a great example of how using current political approaches can bring out new insights for historians.
Johnson’s bubble bursting?
The Bookie Tracker has been useful in tracking the ebbs and flows of the race to become the next Prime Minister of Great Britain. In the absence of polls like the US presidential nominations, betting agencies are some of the best aggregators of information for forecasting events.
The odds showed a big jump for Boris Johnson after Theresa May set her resignation date. He was the presumptive frontrunner, and there was rapid escalation in his odds to more than 50%.
Since then, we’ve seen him come back down. He’s still leading, but the gap is shrinking and Dominic Raab, Michael Gove, and Andrea Leadsom are closing fast. I’ll be tracking the race every Monday, Wednesday, and Friday on Twitter (more frequently as the votes in Parliament get closer) so you can follow along there.
Thanks for reading and feel free to pass this along!
Chris
Caught out by trade wars
S&P drops, bonds fail to see the guns of August, and first insights of the Bookie tracker
Chris Oates May 16, 2019
Stock markets tanked on Monday as China announced that it would respond to last Friday’s announcement of new American tariffs with its own round of tariffs.
S&P, Dow suffers worst day since Jan. 3 as China imposes retaliatory tariffs S&P, Dow suffers worst day since Jan. 3 as China imposes retaliatory tariffsU.S. stocks finished sharply lower on Monday as the U.S. and China escalated their longstanding trade spat. The S&P 500 fell 2.4% to end around 2,812....on.mktw.net
May 13th 2019
17 Retweets10 Likes
This is an escalating trade war between the two largest economies in the world, and a major cause for economic concern, but why did the markets fall that much only on Monday? The pattern has been that after the US increased tariffs, China generally responded. If I were in finance, I would have sold off shares on Friday to get in front of events.
You could even have started selling earlier. A quick search for “China” in President Trump’s twitter feed in the past couple of weeks could have told you that this new round of tariffs (or something else bad) was coming.
And yet the market is continually caught out by these trade war announcements.
On days with a major ratcheting up of the US-China trade war (according to a list compiled by Business Insider), the S&P 500 has fallen by an average of 0.40%. On the days of an announcement and the two following days (since some of the announcements happened on a weekend and so that we capture any immediate bounce back), the markets have fallen by an average of 0.17% on each trading day.
This is compared to the average day of the S&P 500 in that time (March 1, 2018 to today) which saw an increase of 0.02% and the median day, which saw a rise of 0.07%.
The day before the announcements saw the S&P rise by an average of 0.10%, so it is not as if all the trade moves were continuing daily trends.
The other major indices are the same. The announcement day/announcement 3-day period changes for the Dow Jones Industrial were -0.40%/-0.19% compared to a median day of +0.04%. The NASDAQ was -0.47%/-0.12% versus a median day of +0.07. Even the Russell 2000, which tracks small cap stocks less vulnerable to export worries, fell an average of -0.12%/0.00%, compared to median days of +0.08%.
This all means that on the days when the trade wars hit, the S&P lost an average of 93 billion dollars in market capitalization and that losses continued in the subsequent days.
This isn’t investment advice (I want to stress to any regulators who might be reading this), but it does seem that the market is particularly bad at forecasting this trade war, even when it’s tweeting relentlessly that it’s coming. It seems that there are opportunities out there for those who can anticipate these moves faster.
Political Risk Monitor
The PRM this past fortnight has shown a major change. On the back of the trade war, the United States has been displaced as the top country in overall mentions for the first time since the monitor started in March.
Iran has continued to rise in the PRM, as John Bolton appears to be pushing for a confrontation despite allies stating that there is no cause for war.
This is one of the reasons why I created the Political Risk Monitor. It’s easy to read one news story about an issue and then see that same issue pop up over and over, either because our brains are wired to see patterns or the algorithms behind platforms push the same content. Since first seeing Bolton press on Iran in recent weeks, I have continued to follow the story, but worried that I was being led down a rabbit hole while other, more important issues rose up.
Using the Political Risk Monitor has given me a sense check. Yes, Iran is a growing story. Yes, it is one of the most important issues in the world right now. It doesn’t tell us what will happen there - for that, qualitative expertise is needed - but it does tell us that the world of political risk is paying more attention to it.
Tip of the Cap
In researching whether the markets are failing to predict the trade war, I came across a paper from Thomas Chadefaux from Trinity College Dublin about markets failing to predict actual wars. His paper found that the yield on sovereign bonds, which reflect the perceived risk of a country, rose sharply after conflicts began. In other words, the risk of a war was not priced into the bonds until the war started.
There is the counterargument that, even if bonds priced in 95% of the risk of war, there would still be a jump as the last 5% of the risk materialized. We could be looking at the market operating as it should and should not be overemphasizing the jump.
Chadefaux counters this with evidence of bonds before World War I (a surprise) and World War II (a looming conflict). The bond markets saw yields rise in the run up to 1939 but jumped drastically in 1914. This suggests that what we see overall since 1816 is an accurate picture of repeatedly mistimed predictions, because in the most obvious case (World War II) the markets were able to see it coming.
It’s a useful piece of research to caution us against using the financial markets as a main indicator of political events, given their apparent track record.
Bookie Tracker
The Bookie Tracker’s first two weeks have proven that the betting universe does follow world events. Joe Biden’s announcement and subsequent polling numbers have pushed him to the top of the Democratic field.
Jeremy Corbyn has seen his chances of being the next Prime Minister drop, likely a result of the Conservatives moving away from calling a new general election as their poll numbers drop before the European Parliamentary elections.
The Bookie Tracker will be run every Monday, Wednesday, and Friday with the results posted to Twitter, and will soon be adding the markets for the next US president.
A longer explanation and academic rationale of the project is here and you can be in touch with requests for other markets to follow.
That’s all for now. Thanks for reading and be sure to share!
Chris
When is the future too close to the present?
Impeachment, Venezuela, Financial Stress, and the new Bookie Tracker
Chris Oates May 1, 2019
Is 2019 already over for forecasters?
Risk firms have been tracking for the past two years is the probability that President Donald Trump is impeached. That question is heating up now, with newly announced candidate Joe Biden weighing in on it and the Attorney General’s summary being called into question.
Joe Biden: Congress would have "no alternative" to impeachment if President Trump blocked the Mueller probe cnn.it/2GLmReg
April 30th 2019
35 Retweets98 Likes
AG William Barr denied knowing special counsel Robert Mueller's stance on report summary two weeks after Mueller confronted him cbsn.ws/2J3YcEe
May 1st 2019
19 Retweets34 Likes
It is, in its own way, a classic political risk question. It cannot be easily resolved by a historical comparisons. It is highly improbable by one view, and yet could be inevitable by another. And it could have huge implications for the policy direction of the United States.
This question seems to be well-suited for the three techniques that I collectively call the wisdoms of the crowd: prediction markets, forecasting tournaments, and betting agencies.
The three have been found to be especially useful at assigning probabilities to events that have a lot of attention. Every individual may be skewed or biased in their perception of an event. Collectively those biases should even out. They also provide incentives to get the probability right. Prediction markets and betting agencies offer monetary rewards. Forecasting tournaments offer prestige and social status by climbing the leaderboards.
But for impeachment, these fields are highly divergent. PredictIt has the chance of Trump being impeached at 21%. Bet365 places it at 25%. But the Good Judgement Open has a crowd forecast of 5%.
There may be many explanations for this. Perhaps the forecasting tournament is just better at this question, or vice versa. Maybe people don’t want to lock up money for bets that won’t pay off for a couple years, although that would indicate downward price pressure as people want to buy longer odds, and it certainly doesn’t make sense given how many shares have already been traded.
1.2 mil prediction market shares traded so far on Trump impeachment in his first term. Current YES forecast: 27% predictit.org/markets/detail…@realDonaldTrump @DonaldJTrumpJrPredictItpredictit.org
April 24th 2019
1 Retweet2 Likes
If I had to guess, it’s because of the timing of the questions.
GJO is asking whether the House will pass articles of impeachment by October 1. The other two are during Trump’s first term, by January 2021.
To me, this doesn’t merit the huge drop-off in probabilities. That would mean there’s only a 5% chance of articles of impeachment before October, but it’s 16-20% after that. Basically, 5% for the next 5 months, and 16-20% for the 16 months after that.
That seems off. To say that there’s basically an equal chance (by month) for the immediate future, as Congress is launching investigations and all eyes are on subpoenas, as there is during the most intense months of the Presidential election and a potential lame duck period? You can make an argument that there is still a relatively high chance of impeachment until August 2020, but I would think that the most probable scenario is for investigations to pick up steam over the next few months and impeachment not long thereafter. Impeachment at a later date is possible, but I wouldn’t say that after October 1 is three times as likely.
This could seem like I’m making way too much out of some prediction numbers. And that may be true. But I think it actually is a window to one of the toughest problems for any political risk analyst, which is knowing when something big is about to happen.
There is a bias in all people to assume that the future will look like the present. This bias is accelerated for the very near-future. Today’s episode of The Daily included the line “a few months ago no one could have predicted” the coup in Sudan.
I would argue that it is very difficult for people to believe something big is going to happen within a very near-term window. We assume that change could come, but probably won’t happen now. But if you extend the window out - either indefinitely or to a long horizon - it’s easier to see what could occur.
Saying that Congress will pass impeachment in the next five months is hard to visualize. They’d need to start soon, the process might drag on, there’s the summer recess, etc, etc.
But could Congress do something at some point sometime in the future? Of course! By some point in the future, anything is possible.
Neither may be correct. But it is important to be aware - and this is just an educated guess as I continue to look into this - that analysts predicting major, era-defining events may be too conservative in their near-term predictions and too aggressive in their long-term predictions.
What does this mean for impeachment? I’m not saying that it will happen. But I bet that, if it does, lots of analysts will be surprised at how quickly events move.
Political Risk Monitor
At the time of writing, Venezuela may be in the middle of a coup. It came a little out of nowhere, especially to the world of political risk as tracked in the PRM. While Venezuela is the fifth-most mentioned country over the last two weeks, it actually declined by 2 points in the last two weeks from its six-month average.
The potential coup only occupied the last day of a two-week period analyzed below, so this monitor tracks the time leading up to it. It tells us that, although Venezuela has been consistently highly covered by the risk world, the latest uprising was not preceded by disproportionate coverage. It’s another data point for the unpredictability of the exact timing of these types of events.
Tip of the Cap
I recently read this paper by Christopher Gandrud and Mark Hallerberg which used natural language processing to create an index of the perceptions of financial stress in Economist Intelligence Unit country reports.
It’s an interesting way of reconceptualizing political risk. Rather than just coding events as either financial crisis or not financial crisis, they are able to create a scale of how bad the financial situation of the country seemed to policymakers at the time.
This data is available on GitHub and lets us examine which crises were spotted by the EIU in advance and on which were they late to the game. In the below graphic (located here), you can see how the FinStress index (the black line, with higher being worse for the country) tracked with times that other researchers have coded as financial crises (shaded areas).
This tells us that the EIU was quick to identify the Greek financial crisis but that it saw the United States more as a steadily worsening situation.
Why this happened is for another research paper. But that’s a question that can be addressed, now that this data exists and allows us to test past political risk analysis.
Bookie Tracker
As mentioned in the first section, betting agencies routinely run markets for political events. They’re good mechanisms to see what experts with money on the line think are the probabilities for various events. However, it can be hard to know how their predictions are changing unless you are routinely checking gambling websites (which probably is a bad sign for your bank account).
Thus, the Bookie Tracker. This will scrape betting odds from a betting aggregator and find the average and standard deviations of the implied probabilities of events. (Implied probabilities are derived from the odds offered, so a 4 to 1 bet means that the bookie thinks there is a 20% chance of it happening.)
The Bookie Tracker will be run every Monday for political markets that have enough attention to attract multiple bookies. The averaged probabilities tell us what the bookies think is likely to happen, and a high standard deviation means that there is considerable disagreement on what the probability should be. In the example below, Joe Biden has been given just over a 25% of winning the Democratic nomination. Kamala Harris is in third place, but with the widest range for her odds of any candidate.
The markets being tracked right now are Democratic nomination, next British PM, Australian federal election, and a second Brexit referendum. More will be added or dropped as events and the bookies dictate. Future posts will have some analysis of what the betting agencies are telling us about politics (and I’ll figure out which one misspelled Amy Klobuchar’s name but was still more bullish on her than the others).
That’s all for now. Thanks for reading and be sure to share. And if you haven’t filled out your Veepstakes survey, submissions are still open to win an Amazon Gift Card.
Chris
Predicting the next Sudan
Coups, embassies, IKEA analogies, and the first data from the Veepstakes
Chris Oates Apr 18, 2019
What’s happening in Khartoum?
Sudan saw considerable turmoil in these last two weeks. President Omar al-Bashir, who ruled Sudan for nearly 30 years, was ousted by a military coup on April 11, following months of protests. The military junta which took over the country is unstable. Its first leader after the coup lasted only a day in office and it is not known how long they will hold power before a transition happens to a civilian government - if at all.
Naunihal Singh @naunihalpublic
I put down my thoughts about Sudan at greater length. TLDR: look at intra-military factors if you want to understand why 3 powerful military men in Sudan have been removed from power in 3 days. Also, this creates risks & opportunities for democratization washingtonpost.com/politics/2019/…
April 16th 2019
50 Retweets53 Likes
Predicting when a regime will fall is inherently difficult. Leaders can hold onto power despite nationwide protests, as in Syria. They can also resign after only a small incident that few people saw coming. Listen to Mike Duncan’s Revolutions podcast for repeated examples of this. Some of the best work I’ve seen on forecasting coup attempts has come from Jay Ulfelder, who tracks coup attempts and has built a model to try to predict the probability of one.
Chart of coup attempts since 1950, now updated with Turkey's putsch (data source: uky.edu/~clthyn2/coup_…) #rstats
July 20th 2016
47 Retweets54 Likes
There’s much more on these models at his old blog and at CoupCast (which doesn’t appear to have been updated in about a year). The basic idea is that coups are quasi-random events. We don’t know the exact probability of whether a coup will happen in Country X tomorrow, but we know that it is more likely to happen in Mali than Germany. Using historical data to identify some of the variables most associated with countries that witnessed coup attempts, we can build models to target countries most- and least-likely to experience a coup in the next 12 months.
This is a big step forward into our understanding of the environments that make a coup like that in Sudan more likely. But there is still a big forecasting question to be determined. What is the likely result of a coup attempt once we’re in the middle of one? Will Sudan become a military dictatorship? A civilian-led democracy? Something in between. This is a question that shapes the day-to-day actions of governments around the world when instability breaks out. It also leads to the below tweet, an oddly constructed diplomatic photo op that goes out of its way to not support the person the Ambassador is shaking hands with.
Met Deputy Head of the Transitional Military Council today. Not to endorse or confer legitimacy. But to stress steps UK wants to see taken to improve situation in Sudan. Top request was no violence and no attempt to forcibly break the sit in.
April 15th 2019
384 Retweets506 Likes
While we have made advances in the broad understanding of coups, knowing when a tipping point has been reached within a coup and the new regime is here to stay (and therefore can be tweeted about securely) is an ongoing challenge.
Political Risk Monitor
There was broad parity among top countries (below China and the United States) in the past two weeks in the PRM. This appears to be driven by a number of issues that hit in different countries, from the continual Brexit negotiations, to the Israeli elections, and the Sudanese coup.
Sudan was the biggest mover in the PRM, jumping 19% from its six-month average, followed by France and Libya. The French spike is due in large part to the Notre Dame fire. This reflects the inclusion in the PRM of standard news organizations, which covered this story in depth. But it was not only a component of the journalists. Notre Dame led Eurasia Group’s newsletter yesterday, and it is a reminder that these non-political stories can very quickly shift the political landscape, and deserve attention.
Oddly enough, Ecuador jumped in the PRM, due to its decision to expel Julian Assange from its embassy in London, which also led enough news stories to get “Julian” into the wordcloud. Wikileaks is still important enough to move the risk world.
Tip of the Cap
I recently read this paper by Jason Dana, Pavel Atanasov, Philip Tetlock, and Barbara Mellers. It looks at how prediction markets compare to self-reported beliefs (like those seen in forecasting tournaments). One of the most interesting findings, illustrated in the below tweet, is that beliefs outperform prediction markets more than 6 months out from the end date of the question, but that prices in the markets outperform the beliefs near the closing date of the question.
Pavel Atanasov @PavelDAtanasov
Do prices reflect the best available info to market participants? We elicited market orders and unincentivized probability beliefs concurrently from traders. Price signals underperformed early on, caught up over time. Our new paper out in JDM Journal. journal.sjdm.org/18/18919/jdm18…
March 30th 2019
13 Retweets52 Likes
The great part of this paper is not just that it examines the utility of two of the newer mechanisms to predict political risk. That itself would be helpful to set a baseline for future research. But it also draws some conclusions about when each technique has the best results.
One of the problems in predicting the future of politics is not always lacking techniques to do so. There are a number of methods that have been developed over the years that can help synthesize information and reach a conclusion. We have a toolkit, but we don’t have instructions about when to use each tool. Political risk can be like buying something from IKEA, but there’s no instructions and someone hands you a hammer, a screwdriver, and some pliers. You might choose each correctly, but it’s probable you end up with a mattress laying on a broken Skulsfjord. Papers like this can start to build a set of best practices for the field.
Veepstakes - Attributes
Picking a Vice President is a great way to test different political risk methods. It has a highly uncertain result, with varying theories about what the best candidates are, and everyone has some thoughts about it.
The first way we’re approaching the problem is by targeting the attributes that make someone a VP candidate. The survey is here and everyone who submits their information is entered to win an Amazon Gift Card. You’re asked to select attributes that are indicative of a VP candidate, in levels of High, Medium, and Low predictiveness.
A longer explanation of the Veepstakes is here.
The first set of results have been interesting. The highest-scoring attributes have been Swing State Native, Female, and Governor. Does this mean that former Michigan Governor Jennifer Granholm could be the next resident of the Naval Observatory?
The scores will continue to be updated as new answers come in, and we’ll release a list of potential candidates - according to the scoring system - in future issues of the Beacon.
That’s all for now. Thanks for reading and be sure to share!
Chris
How to say we know we don’t know
Mueller Report, Brexit, inequality east of the Elbe, and kicking off the Veepstakes
Chris Oates Apr 4, 2019
The Importance of Being Uncertain
The big political risk news of the last two weeks (of the non-Brexit category) was the completion of the Mueller Report. According to PredictIt, the chance of President Trump being impeached in his first term fell from 32% to 17% afterwards.
That’s odd, since no one outside of Attorney General Bill Barr and Deputy Attorney General Rod Rosenstein had actually read the Mueller Report.
It may be true that Trump is less likely to be impeached now - the initial perception of news is important in framing later debate and the actions of members of Congress and Barr’s Memo is not something to be completely dismissed.
But if you were, say, a general analyst of US politics working at a major news outlet, you shouldn’t conclude that we’ve seen a major inflection point in American politics when we still don’t know most of the facts. You may see your own newspaper publish a major scoop that contradicts a central pillar of your argument.
Analysis: There are still other clouds overhead for President Trump. But the end of the investigation without findings of collusion with Russia fortified the president for the battles to come, including his campaign for re-election. A Cloud Over Trump’s Presidency Is LiftedThe end of the special counsel’s investigation without findings of collusion with Russia fortified the president for the battles to come, including his campaign for re-election.nyti.ms
March 25th 2019
90 Retweets339 Likes
Members of Mueller’s team have told associates that Attorney General William Barr failed to adequately portray the findings of their inquiry Some on Mueller’s Team Say Report Was More Damaging Than Barr RevealedMembers of the special counsel’s team have told associates that their findings are more troubling for President Trump than the attorney general indicated.nyti.ms
April 3rd 2019
1,472 Retweets3,289 Likes
Maybe the eventual release of the Mueller report doesn’t lead to impeachment, but it does contain enough damaging information that it becomes a political albatross for Trump. Maybe it triggers even more state investigations. Maybe it shows his campaign rebuffing Russian actions and earns him plaudits.
I - along with nearly every other person in the world - simply don’t know.
That we don’t know is an important piece of information for our forecasts. Political risk is not simply making predictions. It is also indicating which predictions we’re certain about, and which are little more than guesses, and clearly communicating our confidence levels. It may not be a forecast, but it is necessary context for any forecast.
For me, this was the major takeaway from the initial analysis of the Barr Memo. Not that the conclusions were wrong, per se, but that we were not in a position to draw a conclusion, but too many in the US media still did and we were reminded of the importance of caution in political forecasting.
Political Risk Monitor
The world of political risk turned its attention to the elections in Israel and Ukraine these past two weeks.
The saga of Brexit has fluctuated over the past few months, but these last two weeks have brought it to the forefront.
It is almost the perfect story to demand attention. Constantly uncertain, frequent updates, continually requiring analysis, and located in a country that matters for economic and geopolitical reasons. One of the benefits I hope the Two Lanterns Political Risk Monitor provides is a way to track what the industry is talking about versus what you think matters, so we can start to see what issues could be over- or undercovered.
Tip of the Cap
The graphics team at Oxford Analytica did a fantastic job with this piece looking at income disparity between capital regions and the rest of the country in Eastern Europe.
Oxford Analytica @oxfordanalytica
Several capital districts in #EasternEurope exceed the #EU average for GDP per capita but poorer regions are being left behind, as our #infographic shows: ow.ly/2Hl250nQ4eb
March 24th 2019
1 Retweet1 Like
This deserves extra attention because it is one of the very, very, very rare examples of maps not distorting an underlying argument.
US politics ties itself in knots every four years trying to visualize the Electoral College. Do you paint the states red or blue - which exaggerates rural states - or make every electoral vote the same size, which doesn’t look like real geography? Or do you create a 3-D interactive map that takes a long time to load?
Source: https://www.wired.com/story/is-us-leaning-red-or-blue-election-maps/
The dilemma exists for any map that shows issues relating to a population. Things that affect rural areas are overemphasized with real geography. Urban issues require unusual maps to be seen properly.
This is why the Eastern Europe inequality map is so great. The article argues that capital cities have higher income than rural regions, and that this trend exists across countries in Eastern Europe. A quick look at the map - helped by keeping the real geography - proves this without requiring us to know how many people live in each area. It is the difference within the geography that matters for the underlying point of how politics in these countries may evolve, and the visuals immediately capture what’s happening.
Graphics are a big part of communicating political risk, but they’re also very easy to screw up. Kudos to Jeremy Smith and his team for this one.
Veepstakes - Phase 1
Prof. Anand Menon told Vox that “One of the very curious things about British politics as it stands now is we face one of a number of very implausible outcomes. We might have an election. We might have a referendum. We might have no deal. The prime minister’s deal might be accepted. They’re all massively implausible, okay? But what we know is that one of them is going to happen.”
In a sense, that’s the heart of many political risk situations. What do you do with a question where every outcome has a low probability, a high uncertainty, or a complicated path, but you know that one of them has to eventually happen?
Well, there are a number of approaches, each with their own pros and cons. But rather than a theoretical exegesis, we can run a series of experiments and find out how the methods selected shape the results produced.
The first of these experiments is predicting who will be the Democratic nominee for Vice President in 2020.
This is like a classic political risk problem. VPs are chosen by the presidential nominee, so we have to forecast the actions of a political leader. There’s little polling information on them, so it requires qualitative analysis. There are multiple theories about what makes the best VP so there’s no obvious choice. Plus, it’s a high-profile office so everyone is familiar with it.
For the first phase of the Veepstakes, we’re determining what qualities make someone a likely pick for VP.
I’ve created a survey (click here) for your thoughts on what are attributes most predictive of who becomes a likely VP selection, broken down into High, Medium, and Low levels.
When talking about potential VP picks, if you say “Well, Jeff Merkley will be a contender because he’s a Senator, Hickenlooper is only a Governor, and Seth Moulton probably not because he’s just a Representative,” then your choices are Senator = High, Governor = Medium, Representative = Low.
You can choose 3 attributes for each level. The attributes be scored based on responses and we’ll create a list of the most likely candidates according to the scoring system.
Later this year, we’ll use other techniques to predict the VP, and it will give us a chance to compare how the methods lead us to different outcomes.
Everyone who contributes is qualified to win an Amazon gift card and the results will be posted here. A longer explanation of the Veepstakes is here and the first results will be sent out in the next issue of the Beacon.
That’s all for now. Thanks for reading and be sure to share!
Chris
A newsletter for and about political risk
Hello and welcome to the first issue of the Beacon!
Chris Oates Mar 20, 2019
Hello!
I’m Chris Oates, CEO of Two Lanterns Advisory and Lecturer at the Pardee School of Global Studies at Boston University. This is the first issue of a biweekly/fortnightly asset for the world of political risk.
Why a newsletter for political risk?
Let’s be honest: political risk is a weird job.
You’re asked to make predictions about highly uncertain issues, ones that experts disagree about, and on a time horizon where everyone may be retired before you find out if you’re correct. Those working in the field are scattered between countries and industries. Political risk analysts might be called as such - they’re business strategists or policy planners, or just citizens deciding what candidates to vote for.
In a way, political risk is a highly specialized skill with few true practitioners, and in another, it’s the background of every economic and social decision.
This newsletter is trying to do two things.
First, it’ll be an update on political risk, offering the community of people interested in the field useful information that they can use to improve their own analysis. Most editions will follow this template.
I’ll cover a news story, with preference to the ones where there’s a lot of disagreement within the analytical field.
The latest data from the Two Lanterns Political Risk Monitor (more on that below).
Tip of the Cap to an interesting approach to political risk I’ve seen or read about.
A series of rotating projects (again, more on that below) and reader feedback.
Second, and perhaps more importantly, I want this newsletter to be a way to build the broader political risk community.
If you have thoughts, comments, or questions on a particular issue, let me know. This newsletter is not just a bunch of information I’m spewing out, but should be (to the extent possible) a dialogue among those working in or interested in the field.
That was a long introduction, so I’m going to skip the news story for this week and jump straight to the monitor.
Political Risk Monitor
One of the most difficult aspects of political risk is that we may be deep in the weeds on a project on Russia and Ukraine and finish it, only to find that India and Pakistan are on the brink of war and our boss is demanding a briefing immediately.
The global environment is too much for any one person to stay on top of every story simultaneously. Luckily, the broader political risk community follows all these stories, and they tweet about it.
The Two Lanterns Political Risk Monitor tracks political risk-relevant Twitter feeds. It analyzes when they mention countries for key trends:
Two-week country mentions are scored on a Distance to Frontier scale, meaning that the most-mentioned country sets the frontier and the rest are rated as (Their mentions) / (Frontier country mentions) * 100. The top-mentioned country will always be at 100. This tells you what countries are mentioned most and the ratio between them.
Biggest increases in two-week Distance to Frontier scores from six month scores. This finds which countries are seeing the most attention in recent weeks from their usual political risk state.
Sentiment analysis of tweets to find whether these feeds are reporting more negative or positive news than before.
Wordcloud of tweets, because who doesn’t like wordclouds?
If you have suggestions for what should be included in the Political Risk Monitor, please fill out this form.
Tip of the Cap
Brexit is one of the most confounding issues today. Even with the rivers of ink and pixels spilled on it, it’s hard for an analyst to know what will happen in Westminster, let alone the average citizen who wants to know what’s going on in their country.
That’s why I want to highlight this Guardian article by Josh Holder, Sean Clarke, and Antonio Voce from February that does a great job of taking the reader through how the Brexit votes show the new divisions in the House of Commons. I may be partial to this approach, since it is similar to how I used co-sponsorship information in the Massachusetts State House to understand the networks of influence there (with an explanation of what that means for political risk).
It argues that the usual divisions in Parliament - parties - have broken down. Instead of clear party divisions, we have four groups that break down only partially on party levels by mid-February: Labour whip, Europhiles, Conservative whip, and Europhobes.
All of the 7 Labour MPs who quit the party yesterday fall into the "Europhiles" cluster in our interactive. They split from the Labour whip after backing the EEA amendment. (Luciana Berger did not vote in the first 'meaningful vote' division so is slightly separated)
February 19th 2019
3 Likes
It’s similar to this work by Nigel Marriot, who finds seven clusters of MPs, from Hard Oppose to Hard Brexit.
Ok here's my analysis of the #MeaningfulVote today & possible votes tomorrow & Thursday @anandMenon1 @sundersays @GoodwinMJ @ElectionMapsUK. As before I analysed how MPs have voted rather than what they say will do using Hansard data. We have 7 factions in @UKParliament ... /1
March 12th 2019
15 Retweets26 Likes
Which one of these is the “best” way to analyze Brexit is up for debate. Good arguments can be made for four or seven groups, or that other methods would in fact be better.
What appeals to me so much about this approach is that instead of a mass of uncertainty thrown at a page, this gets the reader a good part of the way to understanding what’s happening very quickly. More research would be necessary to be a true Brexit expert, but this is a great way to introduce the topic without an overwhelming amount of text.
Rotating projects
I’ll be running a series of reader-based projects in this newsletter. Think of this like a PredictIt for topics that don’t lend themselves to prediction markets or where we want to, as a collective political risk field, experiment with new methodologies.
The first project, starting next week, will be a Veepstakes.
It’s far too early to know who will be the Democratic Vice Presidential nominee, which makes it the perfect opportunity to test out different political risk methodologies.
Let me know what you think other projects should be.