Cantor’s game

 

 In Sunday’s Washington Post, Eric Cantor, a former House Majority Leader in the United States, explains that Republican politicians condone such lies as tales of rigged elections because of a “classic prisoner’s dilemma.  If the majority of Republican elected officials work together to confront the false narratives in our body politic…all Republicans will be better off. If instead most elected Republicans decide to protect themselves against a primary challenge through their silence or even their affirmation, then like the two prisoners acting only in their own interests, we will all be worse off.”

Cantor forgets that the prisoner’s dilemma is a dominant solution.  Regardless of what other GOP politicians do, your best strategy is to go along with the false narratives. Even if you and everyone else know that working together will benefit all, you will defect.

Sometimes a little math can actually help. The table below gives the payoffs for your game with another Republican politician.  In each cell, the first number is the payoff to you, and the second number is the payoff to the other GOP leader.  For example, if you denounce lies and the other Republican condones lies, then the payoff to you is zero points, and the payoff to the other Republican is 3. What will you decide to do?  Well, look at the second column, in which the other politician denounces lies.  If you also denounce lies, you will receive 2 points.  But if you condone them, you will receive 3 points. Since 3 is better than 2, you will tolerate false narratives when your fellow Republican denounces them.

Now suppose instead that the other GOP leader condones lies.  If you too denounce them, you will receive zero points.  If you condone whoppers, you will receive 1 point.  Again, your best strategy is to condone.

Since the other Republican faces the same dilemma as you do, he too will tolerate lies.  (Check it out for yourself by comparing payoffs to the other Republican in each row.) So the solution to the game is for you and your peer to condone baseless stories…even though you both would have gained by denouncing them (each player would get 2 points rather than 1).  This is the solution even if you and the other Republican know that the two of you would have gained had you both denounced lies.

   

You/Other GOP leader	Denounce lies	Condone lies
Denounce lies	2, 2	0,3
Condone lies	3,0	1,1

 

Cantor continues: “…Denouncing the false narratives and the conspiracy theories is the first step to winning back the college-educated, suburban and young voters Republicans have lost.” Without a doubt. But the problem is to convince every GOP politician that all others will act decently if she does. Lacking a solution to that problem, the Grand Old Party will blow itself up.

A similar game occurs in much of Central Asia as well as in the trans-Caucasian region and, for that matter, in Belarus and Russia.  Members of the party in power will assert that elections are free and fair, especially when they aren’t.  If all members will concede that elections are unfair, the party may gain supporters in the long run for its honesty. But if one member of the ruling power comes clean, he will lose the President’s support, regardless of whether his colleagues also tell the truth. So the safe thing for the member is to go along with the gag.   – Leon Taylor tayloralmaty@gmail.com

 

Good reading

Eric Cantor.  Many of my fellow politicians won’t tell voters the truth. The result was Jan. 6. Washington Post, January 31, 2021.

Avinash Dixit and Barry Nalebuff.  Thinking Strategically.  Norton, 1993. 

 

The GDP horse race

 

 

A Facebook poster writes: “Our GDP growth for the 3rd Qtr was 33.1 percent. We had the best economy ever until COVID and it is recovering well….Thank you Pres. Trump!”

 

At first glance, an annualized growth rate of 33% looks impressive.  It’s ten times the long-run growth rate.  Good news for Central Asian exporters.

 

But most of this growth is catch-up.  The economy tanked by 31% in the second quarter, thanks to the lockdowns occasioned by the pandemic. In the third quarter, as the pandemic slowed down, cities and states reopened their economies.  That’s why gross domestic product shot up.  (GDP is the market value of what the US produces. It’s the usual measure of the size of the economy.) The 33% growth rate is not a typical result for the Trump administration, despite the poster’s implications.

 

Another problem with the 33% figure is that it’s short-run.  It just measures how much growth occurred since the prior quarter.  As we’ve just seen, such growth may owe mainly to temporary events.  (At least, I hope that the coronavirus crisis is temporary.) 

 

To judge the success of Donald Trump’s administration, we’d like to gauge lasting changes in the economy.  We can do this by calculating the average annual rate of GDP growth over Trump’s four-year term and compare it to the performance of past presidents. 

 

Unfortunately, we have data only for the first three years of Trump’s term.  So, for comparison, I calculated the average annual rate of growth in GDP (adjusting for price changes) for the first three years of the first term of Presidents for the past 30 years.  This long-run growth rate was 3.2% for Bill Clinton, 1.9% for George W. Bush, .6% for Barack Obama, and 2.5% for Trump. 

 

The economy grew fastest under Clinton; but Trump did not do badly, if we exclude the pandemic.  Economic growth was weakest under Obama, but of course he inherited the Great Recession when he took office in 2009.

 

My calculations use data from the Bureau of Economic Analysis in the Department of Commerce of the US government.  The BEA is a treasure trove of free statistics.  You can check it out at bea.gov .

 

Of course, no one—not even the President—controls the US economy, so the newscast references to “the Trump economy” are misleading.  Neither is the size of GDP a precise indicator of human welfare.  But all that is fodder for another post. –Leon Taylor tayloralmaty@gmail.com

Cascades and ambush

 

In the United States, truth has fallen out of fashion.  A December survey by National Public Radio asked 1,115 Americans if they thought that "a group of Satan-worshipping elites who run a child sex ring are trying to control our politics and media," as claimed by the alleged government insider known as Q.  About one in six respondents agreed with Q, and nearly two in five said they weren’t sure that Q was incorrect.  Even with a margin of error of 3.3%, these results astound: Less than half of the respondents said Q was dead wrong, although Q's theory is obviously ludicrous.  QAnon believers include two Congressmen. And adherents were conspicuous in the mob that stormed the Capitol on January 6, when the Senate was certifying the Presidential victory of Joe Biden. What gives?

QAnon may owe its success to an "information cascade," a series of decisions that depend on your observations of what other people do rather than on the facts.  Here’s a simple example.  Dmitriy, a high school graduate, mulls whether to go to KIMEP University.  He reads the prospectus and (correctly) decides to go.  Dmitriy’s friend Sonya doesn't read the prospectus and (wrongly) decides not to attend.  Next is Vladimir, who decides what to do based on what his friends do. Dmitriy attends but Sonya doesn't, so Vladimir concludes that the chances that going to KIMEP would be the right move are 50-50.  He flips a coin. It comes up tails, so he decides not to go. Anastassiya sees that only one of her friends (Dmitriy) went to KIMEP and that the other two (Sonya and Vladimir) did not. So she figures that the chances that going to KIMEP would pay off are just one in three.  She decides not to go.  And any other friend observing this growing sequence and basing her decision on what friends do, will decide not to go...although by hypothesis the right decision for everyone is to enroll.  What's curious is that the outcome depended on a random event.  Had Vladimir's coin come up heads, he and his successors would have gone to KIMEP. 

In the same way, Internet users who trust their friends more than they do scholars and journalists may subscribe to the QAnon conspiracy at an ever-accelerating rate although it is a sheer skazka (fairy tale).  The problem is that the social media enable cascades—not only in the US but in Central Asia. In 2014, social media messages falsely claiming that three banks in Kazakhstan were nearly bankrupt led quickly to bank runs withdrawing hundreds of millions of dollars.  –Leon Taylor tayloralmaty@gmail.com

 

Good reading

Anderson, Lisa R., and Charles A. Holt. 1996. "Classroom games: information cascades." Journal of Economic Perspectives, 10(4): 187-193.  DOI: 10.1257/jep.10.4.187

Bikhchandani, Sushil, David Hirshleifer, and Ivo Welch. 1998. "Learning from the behavior of others: conformity, fads, and informational cascades." Journal of Economic Perspectives, 12(3): 151-170.  DOI: 10.1257/jep.12.3.151

MacKay, Charles.  1841.  Extraordinary popular delusions and the madness of crowds.  London: Richard Bentley.

Rose, Joel.  2020. “Even if it’s ‘bonkers,’ poll finds many believe QAnon and other conspiracy theories.” Retrieved from https://www.npr.org/2020/12/30/951095644/even-if-its-bonkers-poll-finds-many-believe-qanon-and-other-conspiracy-theories).