Let’s start with some basics about the stimulus bill (H.R. 1: the American Recovery and Reinvestment Act of 2009). According to my quick read (so all numbers here are subject to revision) of the Congressional Budget Office (CBO) analysis, the total estimated cost of the proposal is $816 billion over ten years. This is composed of $604 billion of extra spending, plus $212 billion of less money collected in taxes than had been planned. That’s a whole lot of money, even for the U.S. economy.
First, consider when the $604 billion is projected to be spent over the next ten years (all spending timing in Fiscal Years):
Only about 1/7th of the outlays occur in the current fiscal year. Substantially more of the outlays occur in 2012 or later than occur in 2009. How is this stimulus spending?
One argument is that we expect this to be a very long-lasting recession, so we’re going to need this stimulus spending in 2011, 2012, and so on. But, consider that if we’re heading into a Japan 1990s style decade-long slowdown, this kind of spending is exactly what they tried, and it sure didn’t work in that case.
So, if this is a “normal” length recession, the spending bill will have the classic problem that fiscal stimulus does–namely, it comes too late to do much good, but right on time to help stoke inflation and mis-allocation of resources that are suddenly in high demand as the economy enters a recovery. And if this is a very long-lasting recession, more like a U.S. 1930s Depression or Japan 1990s “lost decade”, then the problem is so long-lasting that we’re not really debating a stimulus bill, we’re debating a near-permanent shift of control of resources to the government, which doesn’t exactly have a sterling track record of success. Only if this is a “Goldilocks-length” recession of more than 1-2 years, but less than a decade (which is a pretty hard beast to find in modern American history) would this temporal spending pattern turn out to be wise.
Compare this to the projected temporal distribution of proposed tax reductions:
There is a huge debate among economists about the “multiplier,” which is basically an estimate of how much economic growth bang we get per buck of deficit spending in a recession, and about the effectiveness of fiscal stimulus generally. Keynesian theory is that the multiplier should be greater for government spending than for tax reductions. The debate about this question among economists has a lot more in common with a debate at a meeting of the Modern Language Association (if you were to just add a bunch of data and subtract the intermittent urbane charm) than it does with a meeting of the American Physical Society; which is to say that it’s long on theory and the use of analysis as rhetoric, but pretty short on empirical demonstrations of causal models that can reliably predict the impact of any of these proposals.
That said, at least the tax reductions occur pretty early, and it’s hard to see a whole of multiplying going on for spending that happens several years from now. Of course, these tax cuts weren’t exactly developed with a copy of Hayek in one hand and a calculator in the other. They are mostly a combination of disguised redistribution, social welfare and environmental spending disguised as tax credits, and the kind of rigged changes to investment expensing that have made the U.S. corporate tax the shining model of effectiveness and fairness that it is today. And for whatever it’s worth, academic economists seem to agree with the everyday observation that the 2008 tax rebates didn’t seem to head off many problems at the pass.
Next, consider how the $604 billion of outlays would be spent.
It’s easy to go through a huge proposal and find what seem like fairly ridiculous line items, so I’ll focus on as comprehensive a view as I can of the spending. The CBO reviews each Title (basically, spending area) of the bill, and calls out major items within each Title. Here are all the items that I saw them identify as individual programs with more than $10 billion of projected outlays, in the order that they call them out:
$20.0 billion to increase the maximum benefit under the Supplemental Nutrition Assurance Program (i.e., Food Stamps)
$18.5 billion for energy efficiency and renewable energy programs
$20.4 billion for programs administered by the Department of health and Human Services
$20.0 billion to renovate elementary and secondary schools
$17.6 billion for Pell grants and other student financial assistance at post-secondary institutions
$29.1 billion for other elementary and secondary educational programs
$30.0 billion for highway construction
$13.1 billion for other transportation programs
$11.2 billion for housing assistance programs administered by HUD
$19.5 billion (minimum, could be higher, as per Title XIII) for education grants to states
$27.1 billion for increase unemployment benefits
$13.3 billion to increase health insurance for unemployed workers
$11.1 billion for “Other Unemployment Compensation”
$20.2 billion for Medicaid and Medicare incentive payments to encourage providers to improve healthcare IT
What does this sound like to you? It sounds to me like a wish list for the left wing of the Democratic Party.
I tried to go quickly through the spending for all categories and crudely map them to the OECD classification system that allows for the comparison of spending across governments in the developed world. The huge categories of spending under this bill that I could map to categories other than “General Spending” are in Social Protection (~$90 billion), Education (~$90 billion) and Environment (~$55 billion). Interestingly, Defense represents only about 3% of the spending in the bill (as opposed to 12% of U.S. government spending overall, or about 3% of French overall government spending as a point of comparison) and Public Safety represents only about 1% of spending in the bill (as opposed to about 6% of U.S. government spending overall, or about 2% of French government spending overall). In other words, the net effect of this bill is to shift the distribution of U.S. government spending as a whole away from defense and public safety and toward social programs: for good or ill, to make the U.S. into more of a European-style social welfare state. Because the amount of spending is so huge, this will be a material, not notional, shift. Eventually, we will emerge from this recession/depression/whatever it’s going to be. When that happens, is this really the kind of government we’re going to want?
And this change is unlikely to be temporary. Imagine two illustrative scenarios. First, the U.S. goes through a fairly standard recession and emerges by about late 2010 into a recovery. The government, subject to normal grumbling, is mostly given credit for handling things the right way. Obama is reelected in 2012 and Democrats retain control of Congress. Or second, we enter and new Depression, or more likely, a Japan of the 1990s long-term recession. Unemployment is stuck well above 10%. The mood of the country is deeply pessimistic, and government programs are a lifeline for a good chunk of the population. In which of these two scenarios is it realistic to expect that the 2009 increases to food stamps, unemployment compensation, healthcare benefits or HUD housing assistance will really be rolled back in 2012-2015? Neither, as far as I can see.
And it is this kind of spending that will happen first, and be so locked in, even if we change course in the next couple of years. It’s hard to spend money on roads and bridges extremely quickly unless you get really reckless and inefficient (and we’ll be seeing a bunch of that, too). After cutting taxes the next-fastest thing to do is cut checks. As an illustration, compare the projected timing of outlays for two categories of proposed spending, each of which is projected to be about $45 billion: (1) “Highway Construction” plus “Other Transportation”, which I’ve labeled as Infrastructure, and (2) “Assistance for Unemployed Workers and Struggling Families”, which I’ve labeled as Unemployment:
Or consider the gigantic–as in close to $100 billion–amount of extra federal money these guys are proposing to spend on education. Study after study has shown that, at a minimum, there is no clearly demonstrable educational benefit from more aggregate spending on schools. Unless you adopt the hard-core Keynesian doctrinaire perspective that it literally doesn’t matter what we spend money on, this will be wasted. We will spend $100 billion on schools and not expect kids to read, write or do math any better. How can this be wise? Do we really want to borrow (because that’s where 100% of this money is going to come from) this much money from our kids for that?
Nobody wants to repeat the mistakes of Herbert Hoover. This is a healthy concern. Hopefully we will be able to restrain ourselves from passing trade restrictions. Trying to balance the budget or restrict the money supply right now is almost certainly a fairly crazy experiment to run. We’re going to be running a big deficit. But this proposal is a disaster.