Fair Tests of Predictions: Stimulus and Austerity

-- Posted by Neil H. Buchanan

In my new Verdict column this week, I discuss the last-ditch efforts by economic conservatives to defend "expansionary austerity," the claim that cutting government spending (and, for at least some fans of the theory, raising taxes) will -- contrary to decades of accepted Keynesian wisdom, as found in nearly every economics textbook -- result in a net increase in a country's economic output and employment. This non-Keynesian result, we are told, will occur because of an increase in spending by businesses and consumers that will more than make up for the government's drag on the economy.

The most surprising economic story of the last few years has been that the governments of nearly every advanced economy have publicly embraced austerity measures, putting into disastrous practice an anti-government ideology that was surely going to exacerbate our profound economic troubles. And as Keynesians predicted, those economies have weakened, rather than strengthened. Anyone who has read my posts over the past few years, or anything by Brad DeLong or Paul Krugman (among many others), will recognize this story.

When the world's major economies continued to stagnate (or worse) in the face of austerity measures, the first line of defense from the believers in the Confidence Fairy was that the theory would start to work any time now. It simply had to work, they said, because there was empirical evidence showing that austerity is, indeed, expansionary. As I described in a Verdict column last December, however, the studies to which austerity's advocates pointed, to prove that governments can slash their way to prosperity, were simply irrelevant to our current circumstances. (For example, there have been times when some countries cut government spending at the same time that they were experiencing export booms, with purchases by foreigners allowing the countries to avoid recession. That does not, however, mean that we can simply count on an export boom -- or any other fortunate event -- to offset austerity today.)

My new Verdict column discusses the next stage of denial by the Austerions. The defenders of the faith now say that we never actually tried austerity at all, making the victory laps by Krugman and others completely inappropriate. How do we know that we never tried austerity? Because government spending in the major countries went up (or, in some cases, did not go down by a lot) over the last few years. That cannot be evidence of austerity, can it?

Actually, it can, and it is. Interested readers can look at yesterday's Verdict column to read my full argument. Short version: Austerions confuse cause and effect, thinking that a government is not slashing spending (and harming people) merely because the consequences of those cuts result in even more human need for other types of spending. That we are not seeing huge spikes in total government spending in the current period -- when we should otherwise expect discretionary spending to remain constant (or to rise), while emergency spending (unemployment benefits, Medicaid, food stamps, etc.) should be expected to increase in response to economic weakness -- is evidence of the very deliberate cuts in government spending that have been enacted in the U.S. and elsewhere.

As I was thinking about the effort to claim that expansionary austerity has not been disproved, I began to think about how I have recently been in what might seem to be a similar position. One of the talking points from the right for the last few years, after all, has been: "We know Keynesian economics is wrong, because we passed the stimulus, and it did not work." The short-hand version of the response from me and my ilk has been: "No, stimulus was never given a fair test." At least in form, therefore, this looks an awful lost like the defensive move in which the Austerions are currently engaged. The similarity, however, is entirely superficial.

[On a grander scale, Professor Dorf mentioned to me (after reading my new Verdict column) that this "never tested" meme also captures the attitude of believers in Communism. They argue, quite plausibly, that Marx's writings simply are not the basis for so-called Communist regimes in the 20th Century. They thus claim (perhaps not as plausibly) that Stalinism, Maoism, and other real-world regimes that have been labeled Communist are wrongly used to smear an idea that has never been given a chance to succeed or fail. Professor Dorf pointed out that the defenders of austerity would certainly never credit the idea that Communism was never fairly tested -- just as, I would add, those same people vociferously reject any defense of fiscal stimulus. Professor Dorf's point raises a set of larger questions that require -- at the very least -- a separate post. Pending breaking news and other developments, I plan to return to this topic soon.]

So, how can I argue that stimulus was never fairly tried and tested, but austerity was?

As a matter of theory, the case for stimulus is straightforward, whereas the case for austerity is (at best) counter-intuitive. Because of basic national accounting, we know that cuts in government spending will put downward pressure on GDP, while increases in spending push GDP upward. We also know that, under more normal economic conditions, government borrowing increases interest rates, which depresses spending by businesses and households. This, however, will not be true when the economy is weak, because interest rates will not respond to government borrowing so long as private businesses have no reason to try to expand. And private businesses will only expand if government spending rises first (or if we are lucky enough to experience some unanticipated positive event, such as an export surge). As a matter of describing what should happen in theory, then, advocates of stimulus (Stimulators? Stimulants? Stimuloids?) argue that spending increases (and some tax cuts) during a recession will put people back to work, without any "crowding out" of the government's stimulative measures.

By contrast, Austerions have to explain how private actors overcome the logic of recession. Families and private businesses are understandably terrified of the weak economy, and they reduce their spending in response. Because of the first-mover problem, each family and business waits for others to start spending, hoping that others will get the ball rolling. They have no reason to believe that acting alone will do them much good, so they continue to hunker down. When the government cuts spending, people see that yet more people will soon lose their jobs. This gives them still more reason to be prudent.

For reductions in government spending to be expansionary, moreover, we not only have to believe that families and businesses will respond in a way that is contrary to their most natural impulses, but we must believe that they will do so in a big enough way to more than overcome the drag from the government's spending reductions. That is possible, and it is ultimately an empirical question, but when austerity makes matters worse, it is not easy to say: "But there must be something wrong with what we are seeing, because it just make so much sense for consumers and businesses to spend more in a weak economy when the government is reducing demand even further."

On the theoretical level, therefore, the Keynesian story is straightforward, whereas the case for austerity requires that a lot of unlikely things happen at the same time (and with sufficient quantitative force). Still, it is possible for counter-intuitive stories to turn out to be true. As an empirical matter, how do we compare the real-world "tests" of stimulus and austerity?

First, we have to ask whether the predicate action was taken in each case. That is, did we really do what Keynesians wanted to do, and (after the supposed failure of the stimulus in the U.S.) did we do what Austerions wanted to do? As noted above, my Verdict column makes the latter case: The U.S. and other major countries have deliberately cut spending (and have refused to increase spending, where doing so would otherwise be called for), for the past several years. Government workers are being laid off. Programs are being canceled. Benefits are being reduced or eliminated. Austerity is happening.

As to the question of whether there ever was ever any stimulus spending, the answer continues to be: Yes, there was some, but not enough. The U.S. stimulus package was heavily weighted toward tax cuts that were known not to be especially stimulative. Moreover, state and local governments were cutting like crazy, making the net stimulus from the government sector quite small. People like Krugman said these things, in advance.

Second, we have to ask if we saw what we should have expected to see, under each theory. Austerions told us that private businesses and households would excitedly respond to austerity policies by expanding their own spending. That has not happened. At all. Europe is in free-fall, even though Greece and Spain (and Ireland and Italy and Portugal) have been forced to cut their deficits.

We can expect some Austerions now to claim that the problem was that people were worried that austerity policies might be abandoned. If that is our test, however, then we really have entered the realm of wishful thinking, without any hope of subjecting our policy agendas to real-world tests. If austerity policies could only work in a world where everyone believes that those policies could never be reversed or modified, then we could only test austerity by permanently abandoning democracy (see France, Greece). And not only must we abandon democracy, but we must also believe that we have permanently abandoned it, and that the people who would take over our governments will never themselves think to abandon austerity.

Back on this planet, we can ask whether the results of U.S. stimulus policies supported the claims of Keynesians. Did the recession end, as Obama's people foolishly suggested it would? (They said at the time that the stimulus was not too small, after all.) No. Does that mean that the stimulus did not work? Again, no. Importantly, the economy strengthened somewhat during the time that the weak stimulus was in effect, and it weakened again when the stimulus died out (and the U.S.'s version of austerity started to kick in, under the post-midterms political regime). Moreover, the unemployment rate responded (both qualitatively and quantitatively) as we should have expected, given the relative size of the stimulus. We did not actually "do stimulus" very well, but the economy reacted as the theory said it would.

In short, there is no equivalence between Keynesians' claims that the stimulus was never fairly tried, and the claims by Austerions that austerity was never fairly tried. The theory was always stronger for stimulus than for expansionary austerity, and the evidence supports the Keynesians' theoretical predictions (about both stimulus and austerity). We used a little bit of stimulus, and we saw a little bit of positive response. We (and especially the Europeans) have tried a lot of austerity, and we have seen a lot of negative response. Doubling down on stimulus would have been twice as good (or more). Doubling down on austerity simply reflects ideological intransigence and an unwillingness to face reality.