[QUOTE=Cecil Adams]
Precisely. My own views on the subject, presented more or less at random: (1) Notwithstanding the difficulty of the task, injecting a little science into the study of human behavior is a worthwhile goal, and it’s not inherently foolish to give prizes for such work, whether it be in economics or the social sciences generally. (2) I accept the premise (which some critics don’t) that it’s legitimate to treat people as rational actors who respond to incentives and so on. (3) I can’t say I’ve found macroeconomic analysis especially compelling. (4) I enjoy the work of people like Steven Levitt, a guy after my own heart. I’d be reluctant to base public policy decisions solely on his conclusions, but they’re certainly thought-provoking and worthy of recognition. (5) Using scientific methods in economics hardly makes it a science.
[/QUOTE]
I must have read the previous replies with all the cynicism I’ve gleaned from the pat-on “scientistists” who have responded to macro predictions AS IF they came from empirical models, whether using pseudo data or not.
It’s like this (I think). Ask someone from the World Bank if a state should invest more in education. If we cannot tell people the truth, in this day and age, about what to expect, the models are useless. How will investment in education change per capita incomes? We don’t know! Cross-country comparison doesn’t help answer the question. So I’m suggesting that the smaller the area of prediction, necessarily the narrower the data, the more valid is Economics as a discipline.
It may be stated that prizes are OK, but it better be something way different, something we know will change our lives. The big problems nowadays have to remain: growth, inequality, technology. While Richard Lucas’ Nobel for private incentives was deserved, for example, the idea of accounting completely for the benefits of certain public goods (education, say) by adding externalities has not yet been empirically proven. In other words, do I have to pay as much for my formal education in the classroom? I have such smart friends and I learn from them.
I’m not entirely sure about the word “incentives” because it’s been bandied too much, but I’ll have a go at it. I’d like to take a system, and in tweaking its parts, make it run more efficiently. To do that say you can come up with ways that exploit peoples’ frustrations, stress, problems with personal property, eg., and just suggest they try something a different way. Because sometimes we don’t even know if the incentives are driving us the right way. I mean, sometimes.
Yes, wholeheartedly I want to agree (and excuse me when I find it frustrating) just because you do an experiment doesn’t make you an experimenter. Imagine I say something like “The empirical evidence does not support …” or “my data does not show that …” What am I doing? The first assumption is that it is appropriate for me to put the question to a laboratory setting, and presumably that every phonomenon can be measured. This is just silly. There are situations that just are not made for empirical measurement; I don’t mean there isn’t data for them, there’s data. It’s just not logical to place them in the analytical context that way. People don’t seem to get that sometimes.
Thanks for your reply, Cecil, and others.