Question: Let’s turn to Losing Ground by Charles Murray. Please tell us about it.
Gruber: Charles Murray took the economic concept of moral hazard – the concept that if you reward people for bad behaviour then they behave badly – and turned it into prose. Reading the book moved me a notch to the right. It posed a challenge to liberals – to get more rigorous in our analysis. It showed the simple facts didn’t look so good for us and that we needed to address questions like, “Is welfare causing women to become single mothers?” Murray really challenged the way I thought.
It turned out his facts were largely wrong, so it’s really more a book to read for an example of how someone can shift the debate with potent use of clear arguments.
1. Let’s rephrase the bold part. If, as social scientists, you use bad data and make up your conclusions to fit standard right-wing economic critiques of the state, then you can convince people of standard right-wing economic critiques of the state.
How is that not the exact opposite of what social scientists should be going for?
2. I thought moral hazard had something to do with information asymmetries, not that it is “the concept that if you reward people for bad behaviour then they behave badly.” Remember that the next time you hear about “moral hazard” in the context of health insurance, workers’ rights, mortgage relief and other associated policy.
3. As a friend emailed, the concept of “moral hazard” is actually backwards here. Picture a researcher like Charles Murray, who has the choice of writing a book where the fact are “largely wrong” (call that strategy “hack”) or a book that deeply engages in the literature and data (call that strategy “social scientist”). What do we mean by largely wrong? Outsourced to the excellent intellectual history of the past 30 years, Age of Fracture:
At both the macro and the microlevels, poverty experts were quick to fault Murray’s evidence [in Losing Ground]. The incidence of poverty had, in fact, fallen sharply in the Great Society years that Murray had surveyed. In the most visible and controversial of the poor support programs—aid to single poor mothers with children—the numbers enrolled had shot up dramatically in the late 1960s, but thereafter, the percentage of welfare mothers in the population barely changed at all in the next two decades. The sharp increase in the percentage of all children who were born to single mothers was, to a far greater extent than the headlined numbers showed, a measure of the decline in the number of babies that married women were bearing. The likelihood that a poor single woman would bear a child had, in fact, not dramatically changed. The image of the dependent poor as sharply different from the rest of the population was equally an exaggeration. Although news articles on the poor were almost always accompanied (as the racial subtext all but demanded) by photographs of poor African-American urban neighborhoods, most welfare recipients did not live in inner-city ghettoes in the mid-1980s, and most were white. Even the inner-city drug economy existed in complex symbiosis with the nation at large, a point made starkly visible in the new trading marts (East Palo Alto, California, or the Williamsburg section of New York City), where drugs were handed off to the white suburbs. As for the most graphic of Murray’s examples, his parable of the choices open to Harold and Phyllis, the experts quickly pointed out that Murray had gotten the incentives wrong, even for the unusually high welfare-benefit state he had suggested was typical.
Playing the “hack” strategy gives a huge increase in utility to the player – you are the center of attention, you get in people’s top 5 books – where playing the “social scientist” strategy means you are just another researcher trying to deal with a difficult topic. Now playing the strategy “social scientist” is beneficial to all and the public wants to punish those who play the “hack” strategy and create a contract to do both these things – however the public can’t see which strategy is played with probability 1. That is an example of moral hazard.
4. Occasionally the debate on whether economics is a “real science” breaks out on the internet. A social scientist would look to complicate, not simplify, the picture. They might investigate further and find, when it comes to how poor people balance welfare and getting a job, yes incentives matters but it’s complicated. It involves things that pull in all kinds of directions, including the ideas that “the jobs proposed or imposed are too precarious and ill paid to offer a platform for economic autonomy; they do not provide medical coverage or child care assistance, making employment both risky and prohibitively costly for mothers with young offspring; the workplaces are physically and emotionally degrading; and a majority of ‘welfare mothers’ already work while receiving aid in the first place.”
All of these are economic incentives, but like a drunk under a spotlight, if all you want to see is one part of what the government does then that is what you’ll find.
The rest of the books are worth checking out though – his discussion of “Taxing Ourselves” is particularly useful.