Here are a few reactions to Martin Feldstein's recent WSJ commentary on the effectiveness of the stimulus package. He says it wasn't cost effective because only around 10%-20% of it was spent on goods and services, and he uses this as evidence against Obama's proposed $1,000 tax rebate:
For the stimulus before he was against it, Free Exchange: Martin Feldstein wrote back in December of 2007 that a fiscal stimulus was needed, and that a good way to design said stimulus was in the form of uniform tax rebates. For once, Congress did just what an economist wanted it to do, introducing a tax rebate stimulus plan that sent cheques to millions of households in the second quarter of this year. Naturally Mr Feldstein is appreciative, no?
No. In today's Wall Street Journal, Mr Feldstein writes that of course the stimulus didn't work, and what's more, any old fool should have known it wouldn't. I believe this is what is known as a flip-flop.
Here are the facts. Tax rebates of $78 billion arrived in the second quarter of the year. The government's recent GDP figures show that the level of consumer outlays only rose by an extra $12 billion, or 15% of the lost revenue. The rest went into savings, including the paydown of debt....
The small rise in spending in response to these tax rebates is similar to what previous studies of one-time tax cuts found. It also corresponds to what both basic economic theory and common experience imply. Although someone who receives a permanent annual salary increase of $1,000 typically would increase his annual spending by an almost equally large amount, a $1,000 rise in wealth caused by a share price increase or a tax rebate would raise spending only gradually over a number of years.
This turnabout irks Alex Tabarrok, who adds:
The poor effects of the Bush tax rebate as fiscal stimulus, however, let Feldstein now attack the Obama plan for a $1000 tax rebate. Nothing wrong with that - McCain has nothing better however - but what Feldstein doesn't say is that if you follow the logic of his two op-eds (and this is not something I would necessarily buy into) the conclusion should actually be that fiscal stimulus would work better if it ran through government spending.
At this, I can actually hear Mark Thoma slapping his forehead. This—that if we really want the stimulus money to be spent, we should have the government spend it—is exactly what he has been saying from the beginning of this discussion, and rightly so. If we're going to adopt an interventionist fiscal policy, then we may as well ensure that it works.
And Mr Feldstein's brazen reversal aside, criticism of Mr Obama's tax rebate on these grounds is a little silly. The policy has not been pitched by the campaign as stimulus first and foremost, but rather as a means to cushion households against high energy costs, and it is to be financed by a windfall profits tax on oil companies.
There is plenty to argue against there—if Mr Obama is interested in this kind of Pigovian tax and redistribute scheme then the tax would be better placed on petrol consumption than on oil profits—but to attack its value as stimulus is odd. [Note: On the question of taxing profits versus taxing consumption, see here].
Economist Mom also weighs in:
Was the Tax Rebate a Flop?, Economist.mom: ...[A]n opinion piece by Martin Feldstein in today’s Wall Street Journal ... points to what he considers a disappointing effect of the stimulus checks on household spending...:
Recent government statistics show that only between 10% and 20% of the rebate dollars were spent. ... This experience confirms earlier studies showing that one-time tax rebates are not a cost-effective way to increase economic activity.
Feldstein cites recent aggregate (GDP) data as well as a household-level analysis by Christian Broda ... and Jonathan Parker..., all of which show that consumers don’t seem to be consuming that much.
Feldstein’s disappointment comes out of a narrow measure of fiscal policy “success”... [I]f consumption is encouraged, then saving is necessarily discouraged (at least temporarily). We can’t immediately increase both consumption and saving at the same time. Only through saving can we over the longer run increase consumption and saving at the same time (through higher incomes).
So how bad is it that only about 20% of the stimulus checks were immediately spent? ... First, Broda and Parker point out that the 20% is only within the first month after receipt and that their calculation does not include any potential multiplier effects... Second, this immediate response is similar, or maybe even slightly higher than, the experience with the 2001 tax rebates, which Broda and Parker point out “have been credited with helping end the 2001 recession”–and which over six months eventually produced additional spending that was about two-thirds of the rebate checks.
But mostly, the bright side is that eventually, over even more than six months I mean, the stimulus checks will be spent, even if they’re immediately being saved. Spreading out the consumption made possible by the checks is not at all a bad thing for the economy in a broader-than-immediate-stimulus sense. What it means is that the stimulus checks were partly good for the short-term economy, and partly good for the longer-term economy. In an economy that faces current problems that are clearly not just cyclical in nature, it seems quite prudent to diversify our policy portfolio among pro-growth (longer-term) as well as pro-consumption (shorter-term) fiscal policies. ...
To the extent that the stimulus checks “fail” to deliver immediate consumption, they “succeed” in providing some offsetting increase in personal saving (or decrease in personal indebtedness) and hence some increase in future consumption. And no matter how they are used, the stimulus checks have provided a small boost to the incomes of tens of millions of American families who were surely made better off by them. In that sense the stimulus checks will probably prove to be worth the price of $100 billion in additional debt, especially if we’re convinced the policy hasn’t jeopardized our prospects for economic growth over the longer run. ...
Feldstein segues from his critique of the tax rebates to a critique of the Obama tax plans...
[It'd] an odd line of reasoning coming from Marty Feldstein... [I]t’s not clear that Senator Obama would define “success” in his $1,000 rebate proposal the same way that (supply-sider?) Feldstein would–whether Obama views the goal or purpose of the rebate as having households immediately spend the money, rather than just having households immediately have the money (to spend over time as they choose). Finally, opposing Obama’s proposals to increase taxes (or not extend tax cuts) on supply-side grounds is one thing that can be debated, but it’s more than a little surreal to see Feldstein suggest here that such tax increases would be bad for the economy on demand-side grounds because they would decrease consumption (i.e., increase national saving). ...
Finally, pgl at EconoSpeak:
Feldstein on the Tax Rebate – Life-cycle Model Vindicated, by pgl: When President Bush and some in Congress thought that their one-time tax rebate was just the right fiscal remedy for an economy about to enter a Keynesian slowdown, some of us wonder if Albert Ando and Franco Modigliani’s old life cycle model of consumption had it right – that people would save much of this one-time tax rebate. Martin Feldstein argues that they were correct:
Tax rebates of $78 billion arrived in the second quarter of the year. The government's recent GDP figures show that the level of consumer outlays only rose by an extra $12 billion, or 15% of the lost revenue. The rest went into savings, including the paydown of debt.
That’s right – very little bang for the buck. Feldstein takes this evidence and then turns on one of Obama’s proposals:
These conclusions are significant for evaluating the likely impact of Barack Obama's recent proposal to distribute $1,000 rebate checks to low- and middle-income workers at an estimated cost of approximately $65 billion. … All of the evidence on one-time tax rebates implies that the Obama plan to send $1,000 rebate checks would do little to raise consumer spending and stop the decline in employment. If the past is an indicator of what would happen, the $65 billion he proposes to spend on this plan would raise consumer spending by only about $10 billion, or less than one-tenth of 1% of GDP.
If Obama’s proposal was solely to boost consumption demand – it is even worse than this as it is a transfer of income from one household to another with the likely effect on aggregate consumption being zero if the households had similar marginal propensities to consume. But a lack of consumption is not what ails the US economy. Furthermore, Obama’s proposal is more designed to address income inequality and less about Keynesian demand stimulus. The fiscal stimulus part of the Obama plan would be from accelerating certain forms of public investment.
Feldstein closes on what I consider heresy:
The distinction between one-time tax rebates and permanent changes in net income is also important for the debate about Mr. Obama's proposal to raise income and payroll taxes. Because those tax increases would be permanent, they would cause a substantial reduction in consumer spending and aggregate demand.
Feldstein is well aware that we have a long-term fiscal imbalance with the current level of taxes being far below the current and projected level of government spending. Ricardian Equivalence types would mock at the idea that recognizing those deferred taxes into current taxes might curb aggregate demand. And even if one rejects Ricardian Equivalence, how can any economists with a supply-side orientation dismiss fiscal responsibility. After all, Keynesian maladies tend to be short-run affairs but the crowding-out of investment is something that lowers long-term growth. Feldstein knows this and usually preaches this. What on earth got him to close his WSJ oped with such heresy?
I haven't changed my view since this started. Monetary policy should be part of the stimulus policy package, but there's good reason to suspect it may not work in a crisis such as this, so we need to implement fiscal policy as well. And with fiscal policy, government spending has a larger and more certain impact on the economy than tax cuts, even more so when the tax cuts are temporary (and if you can use it on things like infrastructure spending, so much the better).
Update: Here are Andrew Samwick's comments:
Those of us who did not support this deficit spending thought it was bad public policy to take revenue from future taxpayers to give current consumers a temporary boost of confidence or cash.
So, in the cold light of day, what happened with the rebates was that the government issued more debt than it would otherwise have issued, and consumers now hold less debt than they otherwise would have. That looks like a shifting of liabilities from individual balance sheets to the government balance sheet. In other words, it was a bailout of consumer debt.
While I am not happy for my share of that additional government debt, I am glad that households that received these rebates had the good sense to save them or pay down debt--it shows us some attention to future economic needs that were simply not present in the debt-laced consumption rampage of recent years.
To clarify Andrew's position, here's a bit from a Washington Post commentary of his from January:
[W]e can learn from this experience to have a better menu of fiscal policy options the next time around. Two changes to our budget policy would go a long way toward that goal.
First, we should rule out deficit spending to finance a consumption binge. As the economy slows, the deficit will widen even without changes in fiscal policy. But an honest budget policy would be calibrated to balance the budget over a complete business cycle. Years of cyclical deficits will be offset by years of cyclical surpluses. ...
Second, we can plan well in advance. The federal government has a critical role in maintaining and developing public infrastructure, whether in transportation, telecommunications or energy transmission projects. A sensible capital budget would include a prioritized list of projects that need attention. Some would be slated for this year, some for 2009 and so on, over the useful lives of the projects. When economic growth falters, the government would be in a position to move some of the projects from later years into the present year.
This approach to counter-cyclical fiscal policy has several advantages. Perhaps most obvious is that it forces the government to establish priorities for capital projects. It reduces overall expenditures by doing more of the work in times of economic slack, when costs are lower. It also abides by pay-go rules, since projects moved up to 2008 need not be done in 2009. With a little forethought, short-term economic concerns and long-term budget goals need not be in conflict.
I don't fully agree on the first point. Not the financing part, if you do stabilization policy correctly then the budget should be balanced over the business cycle, with deficits in bad times and surpluses when things improve, the disagreement is over whether there is ever any need for additional deficit spending to stimulate the economy. Sometimes that is needed, and I think the present crisis is an example, but as I noted above using tax cuts to stimulate consumer spending is not the most effective way to stabilize output and employment. That's where the second point comes in, which I think is a good idea, though I don't mean that government spending to stabilize the economy should be limited to infrastructure projects, only that infrastructure spending can be part of the portfolio of policies that address both short-run and longer-run needs.