One of the things that distinguishes the Keynesian left in macroeconomics-- Paul Krugman, Brad DeLong, etc.-- from the more skeptical, and diverse, right, is the weird cocksureness of the Keynesians. Hardly anyone predicted this crisis. The exceptions, like Nouriel Roubini, are few enough that luck rather than insight seems the likeliest explanation. Just because macroeconomists didn't see it coming doesn't mean they don't have some idea of how to fix it. But you would expect some self-doubt, some "Hmm... it seems we didn't understand the economy as well as we thought, maybe we should question our old models." Which seems to be the reaction of the right. The Keynesian left, by contrast, is sure they know the answers, and the only problem is a lack of political will, and irresponsible free-thinking by some economists that weakens the profession's unified voice, and the resistance of the public to somewhat counter-intuitive arguments.
Moreover, there seems to be little recognition that Keynesianism has already been tried on a huge scale. What were the Bush tax cuts, and their renewal, and the major upswing in spending in the past decade, and ARRA, and the Federal Reserve's flooding the economy with money, but a lot of Keynesian stimulus. We've been deficit-spending some 10% of GDP for the past three years. Yet unemployment is still 9%. You can spin counter-factuals about how unemployment would be 16% without the Keynesian measures taken. But how do we know? If the economics profession, or even just the Keynesian left wing of it, had unanimously predicted disaster in the run-up to 2008, they might have some credibility in making these counter-factual predictions. As it is, the public is entitled to conclude that when Keynesian stimulus has been pursued on such a massive scale and the economy is still in such bad shape, it probably isn't working.
That's not to say there's no merit in the Keynesian story, however, and I should mention here that this post is inspired by listening to the first few lectures from Brad DeLong's Econ 1 course at Berkeley, published at iTunes U. DeLong argues (channeling Malthus, Keynes, and others) that Say's Law does not hold, and a general glut is possible, because there can be excess demand for financial assets. If everyone wants to reduce present consumption in favor of future consumption, resources just end up unemployed, and government has to step in to provide the financial assets. He says there might be excessive demand for different kinds of financial assets: cash, government bonds, high-quality private bonds, etc. In 2008, the market learned that mortgage-backed securities, which had been deemed safe, were not. They lost faith in those and all sorts of other private-sector assets, and turned to government bonds.
So far, this story makes sense. But it's a bit incomplete. First, why does demand for financial assets suddenly surge? DeLong has a story. It's basically that there was a certain financial innovation-- mortgage-backed securities-- that seemed smart, but didn't work out in the end. But that seems like a supply-side story-- there was a technological shock to the efficiency of financial intermediation-- which is dressed up as a demand-side story and thereby becomes ad hoc.
Second, it doesn't deal explicitly with investment. After all, if you want to consume less today and more tomorrow, there's an alternative to buying financial assets. Make stuff. Build machines, build houses and offices and infrastructure, invest in R&D. Stuff that costs you today and pays off tomorrow. Now, admittedly it takes a little while to mobilize investment. The people who are in a position to make the plans are probably not the people with the money to lend. It takes time for them to find each other and to design suitable contracts. If in the meantime deflation takes hold, that makes such calculations far more difficult. So it might make sense for the Federal Reserve to flood the economy with cash in the immediate aftermath of a financial crisis, to keep prices stable for the convenience of those planning new investments. But Keynesian stimulus should be an affair of months rather than years.
Third, since the Federal Reserve did flood the economy with cash, DeLong argues that cash wasn't (or not for long) the kind of asset for which there was excess demand. Rather, the economy wanted safe long-term bonds. That doesn't make much sense. Cash should be a very good substitute for safe long-term bonds. It's safe, it just pays a zero interest rate. So to satisfy the market's demand for safe long-term bonds can only mean giving safety-seeking investors a better interest rate. But that's exactly what you don't want to do. If there's excess demand for safety, you should raise the price of future consumption relative to present consumption as much as possible. You should force investors to settle for cash, or else to put their money into physical capital investments that will earn a return by being productive. Ideally, you would want to drive real interest rates on safe assets below zero if there are still unemployed resources at a zero interest rate.
My hunch is that the real secret of prolonged economic downturns is that it's all about political risk. Political risk is hard to observe. But suppose that the odds that a long-term investor in illiquid physical capital will face confiscatory taxes have risen 20% in the past five years. That could induce a steep fall in investment indeed! Money that would have been invested instead seeks safety, so demand for safe financial assets surges. And it's not hard to see why political risk has risen, given the surge in soak-the-rich rhetoric, the election of a president who wants to raise taxes on the rich, Obamacare, the surging deficit that will have to be paid for somehow, the defeat of immigration reform that would have increased the number of people around to pay it, etc. People want to transfer income from the present to the future, but they are afraid to invest it, so it's hard to find things to do with all the resources in the economy.