In a recent post, Science versus Ideology: Readers Comment on Ultrasociety, I encouraged comments from readers, both good and bad. Here’s one response.
In a blog post, titled rather alarmingly, Can we trust Peter Turchin?, Alex Guzey writes,
This is a short-ish critique of his book Ultrasociety. I’ll only cover its one paragraph, which, I hope to show you, is absolutely enough. The paragraph was taken from the 4th chapter titled “Cooperate to Compete”.
Here’s the offending paragraph:
Frederick Wiseman and Sangit Chatterjee sorted the Major League Baseball teams into four payroll classes, ranging from those with the biggest disparities to those with the smallest. Between 1992 and 2001, teams in the most equal class won an average of eight more games per season than those in the most unequal class.
Guzey then accuses me of massaging the Wiseman and Chatterjee results to suit my theory:
There are three data points: one that contradicts him (1985-1990), and two that support him (1991-1997 and 1998-2002). Turchin just ignores the first one and goes on to report the latter ones.
In fact, Guzey’s critique shows a misunderstanding of how statistics work. Here’s the relevant part from Table 3 of Wiseman and Chatterjee:
The results from 1985-1990 do not contradict the hypothesis that more equal teams win more games. The relatively more equal teams won (marginally) more games than the relatively unequal ones, but the relationship is weak and statistically not significant. What is more interesting and important is why the relationship becomes strong in the next two periods.
The reason is that during the 1980s there was very little disparity in pay within teams. The median Gini coefficient, measuring inequality of pay within teams, was only 0.36 in 1985 (as Figure 2 in Wiseman and Chatterjee shows). It increased during the next 10 years and reached a peak of 0.64 in 1995, then decreasing a little to 0.58 in 2002.
Note that the strongest effect inequality had on performance was during the second period, when it reached the peak. So we see a clear relationship: when overall levels of inequality are low, it’s irrelevant to performance. Beyond a certain threshold the more inequality you have, the stronger is its effect on performance.
Gini of 0.36 is pretty small, and corresponds to inequality in a country like Japan, while 0.58 characterizes highly unequal countries, like Honduras. In fact, my expectation would be that if the Gini fell much below 0.3, we would see more equal teams performing worse—because such overly equal distribution of rewards would be perceived as unfair.
I have dealt with the difference between inequality and unfair inequality in another recent post, Dimensions of Inequality.
Thus, I conclude that charges leveled by Guzey at me are unfounded.
According to Guzey’s site, he is a sophomore majoring in economics at Higher School of Economics in Moscow, the well-known bastion of neoliberalism in Russia. I am not terribly surprised that someone with this background would find the general message of Ultrasociety highly unpalatable. Several architects of the Russian economic collapse of the 1990s have found refuge in Higher School, and I have no difficulty imagining what they teach their students (see also Naked Self-Interest is a Recipe for Social Dissolution).
Thank you for your response! I do not believe that you did a fair job of replying to my critique, though.
(1) 1985-1990 results do contradict your hypothesis. By ignoring entries that don’t reach statistical significance you inflate the chance of false positives and your analysis ceases to be valid.
(2) 1985 is an outlier. Excluding it, average Gini coefficent fluctuates around .55 and does not show a steep upward trajectory.
(3) It is still unclear where did you get 1992-2001 interval — it’s not mentioned anywhere in the paper. You can’t just bundle 1991-1997 and 1998-2002 and report the average of them.
(4) Your causal claims are still based on purely correlation data.
(5) You completely ignored my concerns about unfaithful representation of the other two studies, which contradict your hypothesis.
perhaps relative equality just means the team can afford the best players?
Alex Guzey, So now it is possible to buy books through Amazon in Russia! Did you buy a printed or an e- version, I wonder?
I bought the Kindle version.
Allow me to offer a different take on the issue.
This gets to what it was in this chapter that bothered me (it was the only chapter in the entire book which I didn’t love). The issue is that you come across as an academic with a brilliant theory but little exposure to the field in question. I too have zero knowledge of the internal world of baseball, but let me offer some suspicions I would harbor before sending the data to the owners of all the teams:
1). There is a very, very good chance that they know this already, and have known it all along. In other words, they are aware of the risks and rewards of a team of superstars.
2). There is a very good chance that at least some of the teams are not seeking the goal of maximum wins. They may be pursuing the goal of maximizing the value of their team, or maximizing profit from sales and rights or whatever.
3). There is a good chance that those teams whose goal is winning a pennant are aware of the risks and rewards of a team with a few superstars, and are willing to take the risk. High risk/high reward strategies may very well have lower average payouts, but may be a good risk to win the pennant (the obvious risk of a superstar based team is that all your eggs are in one or two baskets). The logic should be obvious, but to spell it out, being a few games above average may actually be a losing strategy.
Again, these are just the first three things that come to mind in a field I am totally ignorant on.
I am familiar with rank and yank in business, and I had similar concerns with your chapter on this issue. Personally I have never been a fan of routine rank and yank, and my experience is that most executives and entrepreneurs share my revulsion. However, it is a theory on ways to motivate and avoid bureaucracy. It has been tried, but no firm tried it without having important voices and interests warning of the obvious pitfalls and dangers. I remember reading these countless times going back to the early eighties.
My take on the issue is that it was one of many fads that businesses go through adding to the long list of such things as Objective Based Management, Participative Management, Total Quality Management, Diversity Management, and so on. These goofy fads come every few years across firms (they would be a great case study on cultural evolution, mimicry and so forth). In general, they are tried for a while,succeed and fail in various ways and ten years later every firm has kept some good practices and cultural effects and lost others.
Just sharing….
Thanks, Roger, for your thoughts. You may be right about team owners. Like you I have zero knowledge of the internal world of professional baseball. But my book is not about that (it was really a throw-away comment). It’s about the general question of how inequality affects cooperation. So I was using baseball as an interesting illustration of the principle. Not even baseball, but team sports as a model for understanding cooperation. As you know, having read the book, my primary interest is in the long sweep of human cultural evolution, and why we first allowed god-kings and then got rid of them.
I wonder if pay inequality itself is the problem, or the perception of whether it is deserved – that is earned in such a way that most people would agree it is justified.
A system worth studying to answer such a question might be the armed forces of a large nation such as the US or Russia – where pay is tied almost exclusively to rank and the promotion system relies more on quantifiable criteria. As you might imagine, such a system is keenly, but at least somewhat transparently competitive.
At least from my experience, the promotion system of the US Army does introduce some almost comical distortions – such as supply officers attending jump school (airborne training) or Ranger training. But at least on the surface, your family name or race is not supposed to enter into the picture. Maybe having Patton or Zhukov as a surname would not hurt.
I guess the hard part would be making a fair determination of such an army’s effectiveness. Both US and Russian armies seem pretty good at straight-up fights against similar – if much weaker – armies; not so good at pacification.
Yes, that’s an interesting question. However, I think with modern armies their effectiveness will be dominated by other reasons–training, esprit de corps, the cooperativeness and equality of the encompassing societies, the degree of external threat, etc.
Actually when someone doesn’t understand afterward its uup to other viewers
that they wiill assist, so here it happens.