Nevertheless, I sometimes feel like social sciences especially have a pendulum model of non-progress. People go back and forth between two distinct camps and there's no real growth in knowledge. On balance, I don't think that is social science's dynamic, of course. On balance I do think it progresses. But the frequency with which it behaves like a pendulum is unfortunate.
I got the sense of the pendulum of economics recently when I began to read this paper by Roger Backhouse and Steven Medema on "government failure" in the Cambridge, Chicago, and Virginia schools of thought. I found it through the GMU Workshop in Philosophy, Politics, and Economics webpage, which I was interested in exploring after reading this recent post by Peter Boettke on PPE programs across the country (a very good development, in my mind).
The paper presents a history of "welfare economics". First, let me present my perspective on that history - which is somewhat of an outsider's perspective, but a decently informed one, I think:
1776-1912: Classical and very early neoclassical economists outlined for the world why the division of labor and free exchange maximized human welfare. They weren't anarchists - they saw a role for the state, but at the time the important point was to show the ways in which state intervention hurt people. So most of the effort was put into demonstrating the welfare-maximizing properties of the market. Arguments for the role the state should play stayed relatively ad hoc simply because that wasn't the point of the endeavor.
1912: Arthur Cecil Pigou (in my opinion, by far the most stereotypically British name a mother and father have ever conspired to anoint their offspring with) writes Wealth and Welfare, which systematizes how we think about human welfare, reaffirms earlier views on the welfare maximizing properties of markets, and introduces the essential insight welfare maximization crucially depends on whether all the costs and benefits of a transaction are borne by the parties to the transaction. If not all the costs are felt by the parties to the transaction, too much of that transaction will transpire (why not produce and trade more of it if you're not bearing all the costs?). If not all the benefits are felt by the parties to the transaction, too little of that transaction will transpire for the same reason. In other words, my view of Pigou's most important contribution was that he put the right to enjoy the benefits of property (one of several traditional elements of a "property right") front and center in the question of the welfare maximizing properties of a market.
post-1912: After Pigou, Ronald Coase made the further innovation of introducing transaction costs. Coase reemphasized Pigou's point that full property rights guaranteed welfare maximization (remember for Pigou the problems arose when there was something wrong with the property rights). But he also highlighted the fact that in theory there's no reason why property rights can't be complete. So why are they incomplete in practice? Coase's solution was that we incur transaction costs when we define property rights or trade rights to property with others. We can think of transaction costs as the costs of trading or drawing up contracts. Coase got a lot of mileage out of transaction costs, including his theory of the firm.
That has always been my understanding of "welfare economics", at least up to Coase, and I've always been a fan of both Pigou and Coase. They seem like two sides of the same coin to me. Externalities (ie - Pigou) form a huge part of my thinking now, especially with respect to the strange but compelling idea of the social contract, and I cited Coase liberally in my senior thesis. They're both great, and quite complementary (at least I've always thought so). Some people juxtapose them more starkly, which is what this paper by Backhouse and Medema has done (although I'm not all the way through it). For them, it's not a question of these innovations of internalized and externalized costs, property rights, and transaction costs - ie, the real meat of welfare economics. For them it's a question of "market failure" vs. "government failure". Pigou is put up as this guy that thought markets failed and trusted government to fix them when they did. Coase and Buchanan supposedly provide a counterpoint to this "naive" position and say that governments can fail.
This doesn't really seem like progress to me - it seems like "pendulum science" that ignores the most essential points of Pigou and Coase. Yes, Pigou saw externalities and realized that if you simply internalized the costs and benefits - if you fixed the property rights - the market would maximize welfare. So he made a case that sometimes the government can do that. But where in the world did people get the idea that Pigou thought governments couldn't fail? Pigou's whole point was the welfare-maximizing properties of the market, after all! In his Industrial Fluctuations he repeatedly criticized government interventions as being counter-productive. He is (unfairly) held up by Keynes as being the poster-boy of neoclassical distrust of government. Where do Backhouse and Medema get this idea that Pigou was naive about "government failure" and down on markets? I suppose that's just derived from the quite simple point that a "Pigovian tax" could potentially internalize costs (for me, this falls under "obvious insight derived from his earlier ideas" not "ignorant of the prospect of government failure"). Backhouse and Medema (and a lot of others, I'm afraid) see the progress of welfare economics as a fight between "pro-government" and "pro-market" forces. That seems to considerably impoverish the real contributions of Pigou and Coase, and promote this idea of science as a pendulum - as a war of the titans - rather than as a developmental endeavor. Even if you put a lot of emphasis on Kuhnian "paradigm shifts" in science, presumably those paradigm shifts improve the way we think about things (despite their haphazard nature). Kuhn never said we oscillate back and forth between two essentially ideological positions.
This is a recurring beef I have with a lot of libertarian-leaning economists. Why do they do this or think like this so often? They tend to refer to themselves as "free market economists", implicitly suggesting that others are not free market economists or are somehow pro-government to the exclusion of being pro-market. My impression has always been that aside from the rare Marxist (who has a pretty tenuous claim to being an "economist" to begin with), economists are almost by definition "free market economists". These libertarian-leaning economists imagine this massive pendulum swinging between "pro-market" and "pro-government, but (1.) the "pro-market" and "pro-government" identities are almost always straw-men (Pigou by any measure is a "pro-market" economist), and (2.) the market/government issue distracts from the most important advances: the importance of property rights to welfare maximization (Pigou) and the importance of transaction costs (Coase). Those are the important things in welfare economics. Why are we getting involved in this food-fight that's almost completely beside the point? Let's drop this illusion of pendulum science and do progressive science instead. That's what Pigou and Coase always put their effort into.