Over at Marginal Revolution we read….
It was Jeffrey Sachs and Betsey Stevenson against myself and Will Wilkinson on the topic of whether America is failing in the pursuit of happiness. The Economist magazine was the sponsor and it was held in Gotham Hall in New York, which yes could have been out of a Batman movie.
Naturally, Tyler Cowen and Will Wilkinson won the debate. Mr. Sachs and Ms. Stevenson had overwhelming support from the crowd at the beginning of the debate (the debate was held in New York after all) but that was not enough to enable them to overcome the brilliance their opponents. Hooray for Mr. Cowen and Mr. Wilkinson! Down with the liberals!
But I have to say that I found the whole debate rather disturbing. Or rather, I should say that I find people who use happiness as a metric for judging a nation’s success disturbing. It always makes me feel as if people are being prepared to accept tyranny.
But let’s ignore my feelings for the time being. I can’t really argue against using happiness as metric on the grounds that it makes me unhappy for people do so. Instead, let us consider what the implications are of using happiness as metric.
First of all, what do happiness researchers mean by happiness?
As I understand it; happiness is a fundamentally amoral concept. Saying that a person is happy/unhappy is not a moral judgment. More to the point, it should be possible to come up with a biological definition of happiness in the same way that depression is now being defined in biological terms.
I think that economists are attracted to happiness research because it holds out the promise of having an objective biological standard that you can use to measure policy outcomes. Granted, right now happiness is self-reported and thus not objective. But there is no reason that it needs to stay that way. Besides, just because happiness is self-reported does not mean that it is a useless indicator any more then people’s self-reporting of pain is useless to medical professionals. In both cases, people are reporting a real biological phenomenon and there is no reason to dismiss what they say out of hand (though allowances should be made for cultural differences).
But why should economists want a biological means to measure policy outcomes? Whatever happened to physical measures of policy outcomes (i.e the increases or decreases in economic activity as measured in goods and services)?
The answer revolves around the truism that “money isn’t everything.” Even in the purely economic sense, it is obvious that people will sometimes voluntarily give up high paying jobs for low paying jobs in order to secure for themselves non-monetary benefits. In a small enough sample size this would cause aggregate goods and services to go down. Theoretically, then, a country with a slower GDP growth could be a more desirable place to live than a country with faster GDP growth.
Because of this theoretical possibility, economists have often been criticized for assuming that GDP growth is all. The argument is made that when economists (in particular neo-classical economists) evaluate policies solely on economists they wind up promoting policies that are actually harmful to people’s well being.
It is natural, then, that economists would look for a metric that could measure “everything.” It is also natural that they would turn to biological measures to do so. Are we not biological creatures? And is not happiness what we all strive towards?
But if people want to reject GDP growth (or lack thereof) as a measure of the success or failure of economic policies because they can give a misleading picture of what is actually good for people, then they should be even more leery of using happiness as a metric. What makes people happy is not necessarily good for them any more than having more material goods is necessarily good for them.
An analogy can be drawn to people’s appetite for food. Just because people have a natural biological desire for food does not mean that we should judge a policy’s success on the extent that it allowed people to indulge their biological appetite for food. That would be a policy that would encourage obesity.
If we use happiness as a metric, we risk similarly absurd policy outcomes. For example, why not put people on drugs that keeps them perpetually happy? Even if it shortens their life span somewhat, you still have increased the total happiness of your society. Even if the drugs increased people’s life spans, I would still argue that such a policy is harmful.
Such an example might seem extreme. But in a society that wants to drug up naturally active kids so that they will fit into the classroom environment better, it is not that farfetched.
But if Brave New World type scenarios seem like an unfair way to cast aspersions on happiness as a metric, why not look at the historical record? Take Adolf Hitler, for example.
As I understand it, the popularity of Hitler’s pre-war government had a lot to do with his success at making people happy without investing the time and effort to increase absolute wealth. That is to say, Hitler’s popularity did not come by increasing the amount of wealth in Germany, but rather he changed the Germans’ perception of their circumstances.
For example, I remember reading one study (which I don’t have time to find again for this short blog post) that showed that under Hitler, labor was actually worse off that it was under previous governments. Yet his popularity amongst the labor class actually grew even as they became worse off in absolute terms than they were previously. The authors of the study speculated that this was because their jobs became more secure even as their absolute income fell. (This was a natural result of the labor laws that Hitler passed early in his rule).
Was Hitler’s success at making people happy proof that his economic policies were beneficial to the German people?
Hitler might seem like an extreme example. But regardless of your ideological beliefs, you can find examples of things that make people happy that are not “good for them.” Conversely, you can find things that you think should make people happy but don’t. Will Wilkinson proves the latter point when he says…
The debate continues in my mind. At one point in the main event Stevenson cut me off to express incredulity that welfare benefits didn’t improve the average happiness of the unemployed. I certainly wasn’t making it up. I don’t know what explains the finding, but there are a number of reasons why this isn’t at all implausible.
One might be tempted to argue that while happiness over the short term is not very good metric but happiness over the long term is. But that raises the question; what is the short term?
For example, let us say that a whole generation promises themselves all kinds of benefits when they get older but leaves the legal obligation to pay for those benefits to the next generation. Let us say that those promised benefits increase the happiness of this generation because it frees them from having to worry about their future. So for decades this policy increases happiness even though its price has not yet come due. Is happiness a good metric by which to measure the success of this policy?
I am not trying be overly dogmatic. Societies that people want to live in will generally tend to have higher GDP growth per person than societies that people don’t want to live in. Similarly, I don’t deny that well run societies are generally happier that poorly run societies.
But if what made people happy was always the best choice for a society, then all societies would be well run. The fact of the matter is that what makes you happy is not always the right choice. Stay too focused on happiness, and everything will end in disaster.