Pun Salad

Once in a blue moon, I try to read something out of my ideological comfort zone. Gosh, hope I don't get converted! There was no danger of that here, though.

The Spirit Level purports to show that high levels of inequality make societies, and the individuals in them, worse off. The authors (Richard Wilkinson and Kate Pickett) allege that there is virtually no social problem that inequality can't make worse. Inequality decreases life expectancy, increases infant mortality, and (for those still alive) causes all sorts of physical and mental maladies. It also causes various types of social dysfunction: crime, teen pregnancy, bullying, etc. People living in unequal societies tend to trust their fellow citizens less, are less happy, and are less likely to improve their lot. And on, and on.

These assertions are supported with many graphs showing the tight correlation of social ills with an increasing Gini coefficient. Country-by-country comparisons bear out the authors' thesis; within the US, state-by-state statistics do the same.

If you click over to the Amazon page, you can read blurb after glowing blurb about the wonderousness of this book. Customer reviews are also very good.

And yet, The Spirit Level was entirely unconvincing for me. It's hard to see how it would be convincing to anyone who brings any healthy skepticism to the table.

The major stumbling block was the book's handling of what I considered to be an obvious question: is it inequality causing these social woes, or is it (instead) poverty? This is not a subtle issue. It shouldn't be any surprise that being poor sucks in a major way, and poverty is also (of course) well-associated with a raft of other dysfunctionalities.

The book at least pretends to deal with this objection early: Chapter 2 is titled "Poverty or inequality?" The authors present a scatter plot of an "Index of health and Social Problems" versus income inequality. As expected, it shows a decent positive correlation.

So then they plot their "index" against some measure of the poverty rate, correct?

No they do not. They plot their "index" against "average income". And (surprise, surprise) no evident correlation there. But waaaaiit a minute….

It's not that the authors were unable to look at how social ills and poverty are correlated. Those statistics are pretty easy to find. There's even a site (www.cdnic.org) where you can do some of these correlations between US states yourself. And you'll find that (for example) there is:

• a strong negative correlation between poverty rate and life expectancy;
• a moderate positive correlation between poverty rate and diabetes rate;
• a moderate positive correlation between poverty rate and adult obesity;
• a moderate positive correlation between poverty rate and infant mortality;
• a moderate positive correlation between poverty rate and births to unmarried mothers;
• a strong positive correlation between poverty rate and teen-mother birthrates;
• a moderate positive correlation between poverty rate and murder/manslaughter rate.

In other words, at least for US states, a lot of the ills described in The Spirit Level are moderately-to-strongly correlated with poverty.

(Aside: I should point out that "CDNIC" in the link above stands for "Correlation Does Not Imply Causation", a good saying for everyone to take to heart. Wikipedia has a page devoted to this concept.)

So: does income inequality explain social ills better than poverty? Or worse? I don't know. But Wilkinson and Pickett don't dispose of that issue in their Chapter 2, despite its title. That obvious failure haunts the rest of the book.

There are a number of other problems, but I don't want to yammer on forever. I'll just mention a few that stuck out:

There are a couple chapters devoted to crime and punishment. Unsurprisingly, inequality is fingered as a culprit. But—again, waitaminnit—US violent crime rates have been on a long-term downward trend since 1993, during which period (we're told) income inequality increased significantly. That would appear to indicate that the actual effect of inequality on crime, if any, is dwarfed by other factors.

After building their case, the authors are noticeably flaky on any actual reforms. They argue that any lasting changes must be something that incoming legislatures can't undo. (After all, you can make it legal to loot the rich, but it's just as easy to make it illegal again.) They devote many pages to employee stock-ownership schemes, trying to argue that it's a superior form of business operation. And they would never let the CEO be paid a zillion times more than the custodians, would they? So problem solved!

Anyone who even skims the business news will recognize the weakness of the employee-ownership argument. Do employee-owned firms outcompete their peers? Evidence is (to put it mildly) weak; if they did, wouldn't there be a lot more of them?

And their examples of employee-owned firms aren't persuasive. The biggest US example is probably Hy-Vee, a Midwest supermarket chain with about 55,000 employees. A couple of their other examples are less successful: United Airlines, which terminated its employee-ownership plan years ago, just before falling into bankruptcy, leaving the "employee-owners" with a lot of near-worthless stock. And Polaroid, whose employee ownership did not prevent its woes either.

Finally: The authors combine their argument with some standard environmental doomcrying to argue for a "steady-state" economy. They ignore the massive statism that would be required to pull something like that off, and handwave away any possible negative effects.

But, notably, they do find one country that "manages to combine acceptable living standards with a sustainable economy." This country "proves it can be done."

Friends, I am not making this up: that country held up by Wilkinson and Pickett as a shining example is Cuba.