Benjamin Radcliff is a professor of political science at the University of Notre Dame. His current research focuses on how public policy affects human happiness. His recent book, "The Political Economy of Human Happiness," argues that generous welfare states and strong labor market protections produce happier citizens than do more laissez-faire policies. We spoke on the phone Thursday afternoon; a lightly edited transcript follows.
Dylan Matthews: Before we dig in, I wanted to ask about the distinction between "happiness" and "life satisfaction," since many happiness researchers tend to distinguish between the two. What's the difference, and which do you focus on?
Benjamin Radcliff: Sometimes the psychologists will want to say that if we ask people about happiness we get a more affective (i.e. emotional) response, whereas life satisfaction, and this is what I tend to focus on, is more cognitive and thus more foundational and less subject to change. We simplify down to the basic idea, which is not "are you in a good mood" but "do you actually enjoy your life"? So, we're not talking about the happiness of the smiley face but the truly basic question of, when push comes to shove, how much do you enjoy being alive? In other words, to what degree do you find your life positive and rewarding? In any case, when you're dealing with broad aggregate data they correlate massively. Empirically, the distinctions between them are fairly modest.
A recent New York Times op-ed by AEI president Arthur Brooks argued that happiness is largely determined by either genetic factors or recent life events — and that the effects of the latter are generally small in scale. So, how much room is there left for policy to influence, short of out-there stuff like genetic engineering?
That's a good point. There is a debate about the extent to which genetics (or social comparison, which Brooks didn't talk about much, which economists are obsessed with), dominate our well-being. Fifty percent determined by genetics -- other people would say higher, others would say lower -- but that's probably a good middle consensus figure. It is certainly true that some large percentage of our wellbeing is determined genetically or through personality structures that don't conform to our commonsense view that happiness is determined by the quality of life that you experience.
That other, more commonsense tradition of thinking about happiness is associated with what's called livability theory, that if you live in a better society that takes care of people better you'll lead a happier life.
One footnote from rereading the Brooks piece, which in many ways is a nice article, however we parse the percentages, he claims that the key to happiness is choosing to concentrate on basic values that should promote happiness.
There's a lot to be said for that kind of thing, but hidden in there is the assumption, which comes from the market fundamentalist view that seems to inform his basic world view, that happiness is something we choose, that we decide for ourselves, instead of something that happens to you. For a lot of people who don't have the privilege to live comfortable upper-middle class lives like Brooks or me, life is not solely what you make it. Your quality of life is largely determined by your life circumstances. The obsession with happiness as an individual-level phenomenon that individuals can determine is dangerous and dubious.
You argue that social democratic or left-leaning policies are more conducive to happiness. What sorts of things are you talking about? Government spending? Regulations? Both?
I have organized my research around two dimensions of policy. The first is the size of government, i.e. of what it is government does, from the tax burden to the generosity of the welfare state to the total impact of the government in terms of its overall consumption on GDP. The second involve institutions that protect people in labor markets, which means labor unions and economic regulations (the minimum wage, mandated vacation time, etc.), which provide a degree of sovereignty and power for workers in their employment relationships. Two sides of the coin: the general scope of what government does to make life more secure for people and the stuff that works specifically in terms of peoples' work conditions.
Both types of policies contribute to what social theorists call "decommodification,” meaning limiting the degree to which in a capitalist economy people have to act as commodities in order to survive. You have to sell your labor power on the market. Decommodification measures how much people can opt out of the labor market, whatever the reason, and provides a way of judging to what extent have we made them free of market commodification.
More decommodification makes people happier, and it does so for rich and poor people, men and women, and controlling for just about any other thing. Similar empirical results obtain when considering total social spending on education, health care, total government consumption, the tax burden, a well-known OECD measure of employee protection legislation, even indices on the size of government and labor market regulation from the conservative Fraser Institute. The smaller the government, the less happy people are.
Another variable I find of interest is labor union membership and density, i.e. do you belong, and the percentage of all workers who belong the unions. People who belong to unions are happier, and, more importantly, union density is strongly related to levels of happiness for union members and non-members.
Even when Stockholm looks like this, Swedes have their welfare state to keep them happy. (Scanpix Sweden / Reuters)
How does that compare to the effects on happiness of non-policy things like, say, the effect of being married or unemployed?
The literature would tell you that being married has a huge positive impact on wellbeing, while unemployment has an equally powerful negative effect. They thus make nice benchmarks for comparing the effect of other variables. My results suggest that the effect of the political variables is much larger by orders of magnitude.
Does regime type matter more than level of development? Will moving from the U.S. to Sweden make someone happier than if they moved from, say, Bangladesh to the U.S.?
My research focuses only on the industrial democracies, so I can't really make that comparison. Within that limited range of countries, I'm interested in GDP as a control, but it turns out to not be generally very significant or important because the sample has been truncated (we're looking only at rich countries). In terms of regime type, again, I am looking only at stable democracies, so I do not explicitly consider regime differences (aside from their public policy regimes).
The obvious objection with any non-experimental research like yours is that it could just be that happier people live in countries with bigger welfare states, but that the latter doesn't cause the former. What makes you think that these policies are actually causing a meaningful increase in happiness?
On the empirical side, you do everything you can to control for these differences. For instance, in many of the statistical models I include dummy variables for every country and every year. Whatever it is that makes Sweden different from the U.S., aside from the variables in the model, you're controlling for it with the dummy variables, which do a very good job of absorbing the variance. Thus, the fact that, say, happiness is always higher in Denmark than the U.S. can be accounted for without being able to specify the potentially limitless number of reasons why that might be the case, aside from the factors you are focusing on. The year variables do the same thing with regard to time. The data are from 1981 to 2007 -- if there's something going on in 1995, we can fit a dummy variable that controls for that worldwide trend at that particular time.
More generally, like most social scientists, I have tried to structure the work so as to address all possible counterarguments (both statistical and conceptual) throughout the text, in footnotes, and appendices. I feel confident that the results are robust. In any case, fundamentally all statistical results are just evidence that has to be embedded in a theory that makes the results sensible. I've tried to make the arguments about why it is that these kinds of institutions promote happiness that is grounded in the happiness literature. So, to belabor the point, one develops a large pile of statistical evidence, while endeavouring to answer counterarguments, and by then you have confidence in the results to the extent that the theory that supports them.
Can you talk a bit more about what you mean by "decommodification"? Do you mean not being reliant on work to live — not being a commodity yourself — or the carving out of certain things (human organs, say) that just aren't commodities you can buy and sell?
A society is decommodified to the degree to which people are not entirely dependent on labor market participation in order to survive — principally because they are aged, because they are ill, or simply because jobs are scarce, but also, potentially, so that they can take time to care for a new child or an ailing family member, etc. My research suggests people lead better lives in those societies that are the most decommodified. The reasons are easy enough to understand: There’s a famous quotation observing that a capitalist economy, whatever its many positive aspects, creates a situation in which people have to behave as commodities in order to survive. It doesn’t take great insight to realize that people do not enjoy being reduced to commodities, so a society that limits that necessity is likely to be a better one in which to live.
Now, to be sure, the market economy absolutely contributes to human well-being in other ways — no one can deny that — but we have a macro- vs. micro-problem. At the macro level, capitalism works well. I would agree with Brooks that the market society is one of humanity’s greatest achievements. But at the micro level it depends at the very core of its logic, as even Adam Smith was at pains to point out, on the idea of using other people (employees) as a means to making profits for oneself. The people we hire to do work are just mere commodities in the profit-loss calculations, no more worthy of special concern than barrels of oil or bushels of grain. The last chapter of my book discusses these moral tensions that capitalism creates. My conclusion is that the social safety net, labor market regulations and labor unions all limit the degree to which people become mere commodities, and thus are more likely to lead fulfilling lives.
