Einstein’s insanity
19 February
The recent anniversary of the stimulus law reminded me that I need to recap my views. I will highlight a point I think is being overlooked in the historic debate and comment briefly on the pending “jobs” bills.
Recap: my views on fiscal stimulus
Unlike many critics of the stimulus law, I think that fiscal policy can increase short-term economic growth, especially when the economy is in a deep recession. In other words, I think that fiscal stimulus is a valid concept. This does not mean that I think that every increase in government spending, or every tax cut, (a) increases short-term economic growth or (b) is good policy.
In the world of fiscal stimulus there is a tradeoff between speed and power. Putting money directly into people’s hands is fast – people immediately change their behavior. We saw this on a much smaller scale with Cash-for-Clunkers: as soon as the incentive was out there, people sprinted to the dealership to take advantage of it. “Putting money directly into people’s hands” includes tax cuts, payments to individuals mislabeled as tax cuts, and entitlement spending like unemployment benefits and food stamps. The challenge with these options is that if the payment stream is temporary, people don’t spend 100 cents on the dollar of the money they get. They instead save a fairly big portion of it (probably between 60 and 75 percent). So if you’re trying to increase short-term consumption, a temporary tax cut or temporary benefit spike is fast but not powerful. Then again, in the current recession where many family balance sheets were decimated by a severe decline in the value of their home, rebuilding those losses with increased saving is a good thing, too. If the Administration had instead put $862 B directly into people’s hands, you would have seen more immediate spending and economic growth than we did, even if people had saved most of it.
In contrast, government spending is powerful but painfully slow. If the government spends $1 on building a road, eventually that entire $1 will enter the economy and increase GDP growth. Your bang-for-the-deficit-buck is extremely high. The problem is that bang-for-the-buck doesn’t help us if that bang occurs two or three or four years from now. Last year’s stimulus contained $8 billion for high speed rail and $1.5 billion for intermodal transportation grants. Not one dollar of that $9.5 billion has yet been spent. If you look at all $48 B in transportation spending in last year’s stimulus law, as of mid-January less than 20 percent had entered into the economy. Power does you no good if it comes too late.
There are other differences between spending stimulus and cash-in-people’s-hands stimulus. The Administration and its allies emphasize what they see as the non-stimulus policy benefits from their increased government spending (environmental, education). I would instead prefer that people be allowed to spend and save the money how they best see fit. My preferred path also has less waste and bureaucracy. These debates are unrelated to the macroeconomic questions and are a distraction from them.
As with a double espresso or an energy drink, there’s a post-stimulus letdown problem with any temporary stimulus. We’re going to see some of that effect this year.
Last year’s stimulus law
I agree with the Administration that last year’s stimulus law increased economic growth above what it otherwise would have been. I agree that employment is higher than it would have been without a stimulus.
At the same time, I think their estimates of how much the stimulus helped are exaggerated, confused, and misleading. The law was poorly designed and inefficient. The money has been slow to spend and filtered through federal and state bureaucracies. As a result, the macroeconomic benefits were diluted and mistimed (and timing is everything in stimulus) and tens of billions of dollars are being wasted.
Most conservatives and Republicans say what they don’t like – the stimulus law. Few argue what they would have done instead. It’s easy for outsiders (including me) to complain about the actions taken. Ask a critic: If you were in charge, would you instead have done nothing? Do you think you could have sustained that position politically throughout 2009?
Given a decision last year to do a big fiscal stimulus, I would have preferred, in this order:
- putting all the money into a permanent reduction in income and capital taxes;
- putting all the money into a a temporary reduction in income and capital taxes;
- putting all the money into transfer payments;
- what Congress and the President did.
Given the policy preferences of the President, his team’s big policy mistake last year was to let Congress turn a reasonable macroeconomic fiscal policy goal into a Congressional spending toga party. Given his policy preferences, the President should have insisted that Congress put all the money into (2) and (3) above. He would have had a bigger macro stimulus bang earlier.
Team Obama’s even bigger mistake was in overselling their policy. They made the classic mistake of overpromising and underdelivering. The political system is punishing them and their allies for this error.
Their messaging had and still has two fatal flaws:
- They created expectations of a net result, rather than an increment. This is the infamous graph in which they “promised” that their policy would result in an unemployment rate that would now be below 8%.
- They are attributing whatever economic recovery is happening solely to the stimulus law.
Only the first has been sufficiently debated. No one knows how big of a positive effect the stimulus had, since no one knows the counterfactual of what economic and job growth would have been had there not been a stimulus law. This unknown aspect always exists. The Administration tried to game it with their unemployment graph and their “jobs saved or created” measure, and they have gotten politically burned because of it.
The second flaw is important and insufficiently discussed. I’ll show it with a picture.
Problem #1 above is that we can’t measure the increased economic growth since we don’t know the baseline from which we’re measuring. We don’t know what numbers we should put in the rounded rectangle on the right.
Problem #2 is that the Administration suggests that all of that growth resulted solely from the stimulus (blue box). It is instead the result of the natural recovery that happens in any business cycle (green box), plus the stimulus, plus all the other policies that are contributing to economic growth. We cannot know how much of the economic recovery is the result of policy, and how much would have happened naturally in the absence of any policy changes. We also cannot know the relative importance of various policy changes in contributing to increased economic growth.
As an example of an alternate viewpoint, I think the financial crisis beginning in September 2008 turned a mild recession into a severe recession. The severe recession was caused by a financial wound. I believe the most important policy actions to address the severe recession and its cause were (1) preventing the collapse of the financial system, (2) recapitalizing the banks, and (3) pumping tremendous amounts of liquidity into the financial system. These actions (which were concentrated in Q4 2008 and Q1 2009 and spanned the Bush and Obama Administrations) treated, cleaned, and dressed the financial wound. This put the patient on the path to a painful, slow, natural recovery.
I think that much of the recovery we have seen would have occurred even without the $862 B law, because I think the first three white boxes and the green box are what mattered most.
Fiscal stimulus could have helped a lot more than it did had it been designed correctly. And I do believe that it was a contributing factor, but nowhere nearly as much as the Administration would have you believe.
Reasonable people can disagree on the relative contributions of the business cycle and policy, and on the relative contributions of different policy actions taken. At the same time it is clearly incorrect to attribute all of the economic improvement to the stimulus law or even to policy.
Einstein’s insanity
Insanity: doing the same thing over and over again and expecting different results. (Albert Einstein)
The Department of Transportation’s “Progress report on the American Recovery and Reinvestment Act of 2009” (p. 33) shows that after one year fewer than 20% of transportation infrastructure stimulus dollars have entered the U.S. economy. So what does Congress want to do in their next stimulus (sorry, “jobs”) bill? Increase transportation infrastructure spending, of course.
The pending House and Senate stimulus/jobs bills are political exercises more than policy efforts. The Reid bill is particularly laughable. In a nearly $15 trillion economy, a $15 billion bill is a rounding error. Its effects will be so small that no one will be able to measure it.
These bills have three problems:
- Their effects would be small relative to the size of our economic problem. This doesn’t mean you shouldn’t do them, just that you need to be careful you don’t lead people to believe that things will get much better if you’re successful. You need to be careful not to overpromise. The Administration’s track record here is poor.
- The policies they are including are, like last year’s law, slow, wasteful, and inefficient. Here we go again.
- They are trying to directly stimulate job creation rather than just trying to increase economic growth. This is hard to do.
Traditional macroeconomic stimulus focuses on increasing economic growth. If firms expect that they will sell more stuff in the future, then they will hire people to ramp up production. Short-term stimulative monetary policy tries to increase the purchase of stuff that is sensitive to interest rates. Short-term stimulative fiscal policy tries to increase broadly the purchase of goods and services in the economy. Increased job creation, should it occur, comes as firms hire workers in anticipation of greater demand for what they sell.
The pending legislation is basically trying to bypass this and instead directly change the hiring decisions of firms by temporarily reducing the costs of hiring a new worker. They are taking the future path of economic growth as a given and trying to increase the number of workers who will be employed within that growth path.
Will private employers hire a lot more people if the costs of doing so are temporarily lowered? I am highly skeptical. Even in the best case, the number being bandied around by the Administration (+600K jobs over the next year) is small relative to the size of the unemployment problem. That might knock half a percentage point off the unemployment rate.
When faced with a choice between doing something inefficient, wasteful, and directionally correct, or instead doing nothing, Congress will always choose the former. That’s what they are doing now. Does the Administration and Congress expect a different result from doing largely the same thing on a smaller scale?
(photo credit: Einstein tongue at Wikpedia)
Related Posts
(best matches are listed first)- Fiscal stimulus camps
- How to enact a bipartisan stimulus (and why it won’t happen)
- Will the stimulus come too late?
- The wrong health reform will hurt the economy
- The President’s new stimulus proposal
- What does it mean to focus on jobs?
- Summary of Chairman Bernanke’s testimony
- New projection: 2.3 million fewer people working in 2010
Another problem with current stimulus spending is it's neither short-term nor immediate. In suburban Detroit, for example, stimulus money is being used to construct a new bus and train station while several bridge repair programs have been cancelled. The long-term cost of the stimulus will probably be much greater than $800 billion regardless of its effect on the deficit.
I have a problem with stimulus spending as practiced by our government both before and after O. What we see and saw is throwing money at an economic problem and prayer. Nothing else. If a long term plan to grow the economy – a plan believable to capitalists (those who would invest and actually drive the growth) – were activated at the same time, then I think a stimulus might be warranted. Just saying "Let's give everyone $700 cash — well, not everyone, just the poor; we'll take it from the middle class and rich — is not smart in my mind. That seems like a delaying tactic — "Can I get out of this presidency before all hell break loose?". While I don't really think that was in anyone's mind that is how it came over. It had zero effect on the economy as far as I could tell.
Using a stimulus to employ people doesn't grow the economy. The reason is the stimulus will run out. Furthermore, the stimulus could exacerbate a down economy because the money frequently requires matching funds from a state or local government. If the state is goes broke during the lifetime of the stimulus, what happens to the workers? It seems to me funding infrastructure has the same effect on the economy as food stamps and unemployment. All three just pay people from government coffers until the economy improves. If the government can not figure out how to get out of the way and let the economy improve, then the stimulus will either have to be renewed or people will be getting awfully mad when the stimulus money runs out.
Stimulus is only justifiable with a credible economic recovery plan. We don't have such a plan.
To Keith Hennessey:
You wrote "I think that fiscal policy can increase short-term economic growth, especially when the economy is in a deep recession. In other words, I think that fiscal stimulus is a valid concept."
I interpret "fiscal policy" as borrowing, spending, then taxing later to pay off the debt. Would you explain why you think that such government action creates economic growth? Would you include a reference to the measured results and past successes, at any economic scale, that shows this result?
More:
I am puzzled by the phrase "short-term economic growth". Do you mean that the growth appears and then disappears? Why would a short-term result be more valuable than "slow but steady recovery" for example?
More:
As an abstract idea, government could be as wise an investor as private investors, borrowing money, applying the resources to make something, and paying back the loan from the extra value created. However, has government ever been that wise? And why should the government do this instead of private individuals doing the investing and management that they routinely do? What makes the government wiser than the millions making up the free market?
Obama's and the government's most harmful economic myth is that savings are bad, and that only spending improves an economy. Spending transfers current production, but savings buys the equipment and business expansion that directly produces jobs. The reason that government taxation and borrowing is bad is exactly because these use up savings and direct savings toward bad investments.
Cargo Cult Economics
People do not prosper because money moves around or stops moving around. The money is the result of an efficient management of resources to produce what people want (can pay for). Building more houses did not make the society rich, because many people couldn't pay for them. The government got us into a mess because they encouraged everyone to buy a house, no matter what.
…In contrast, government spending is powerful but painfully slow [but] bang-for-the-deficit-buck is extremely high.
~~~
Is this documented in practice or only theory/presumption?
(1) Government "spending" can be redirected by politics, since money is fungible.
E.g., here in NYC the Transit Authority is broke (extremely so!) and the politicians, to get the Transit Workers Union its politically mandatory pay raise, redirected $350 million of stimulus funds supposedly for building the fabled Second Avenue Subway to an 11% pay raise for the union over three years. (Sending $350m to the construction budget freed $350m from it.)
What's the multiplier for that?
(1) The CBO paper examining cost-effectiveness of various job-creating options gives "Investing in Infrastructure" as creating 4 to 10 job years per million dollars of cost, and "Reducing Employers' Payroll Taxes" as creating 7 to 16. Which is more. (With much faster effect too.)
… Continuing, since your editor cut off my prior comment for excessive length:
It also seems to me that "spending on a road" may not be so bang-for-the-buck powerful if it goes just to raise yet further the pay of a politically favored group that already works on the road — as by providing a bonus extra repaving for Rocky's Road to Nowhere", well known to NYS taxpayers.
There's another question about "bang-for-the-deficit-buck" I've never seen anyone answer. BLS says the average wage is about $42,000. The claim for the $787b stumulus (now $860b) was that it created one job year per $116,000. How do I know if that is a good deal?
Assume a job pays $42,000 per year. How much should we add to the debt to create it for a year? What should the limit be? $42,000? … $116,000? … $150,000? … $200,000… $250,000? If we don't have an objective answer to this, how do we know what the heck we are doing?
… Continuing, since your editor cut off my prior comment for excessive length:
It also seems to me that "spending on a road" may not be so bang-for-the-buck powerful if it goes just to raise yet further the pay of a politically favored group that already works on the road — as by providing a bonus extra repaving for Rocky's Road to Nowhere", well known to NYS taxpayers.
There's another question about "bang-for-the-deficit-buck" I've never seen anyone answer. BLS says the average wage is about $42,000. The claim for the $787b stumulus (now $860b) was that it created one job year per $116,000. How do I know if that is a good deal?
Assume a job pays $42,000 per year. How much should we add to the debt to create it for a year? What should the limit be? $42,000? … $116,000? … $150,000? … $200,000… $250,000? If we don't have an objective answer to this, how do we know what the heck we are doing?
Keith,
I'm glad you put permanent income at the top. People's behavior is not changed unless there is confidence that the change in the tax code is indeed permanent. This simple idea, that does work, (it worked under W, the change in receipts was dramatic that the deficit was falling), is what Republicans are so often criticized, having only one idea, tax cuts. But it is just that idea that led to the implementation of the Earned Income Tax Credit or EITC. The EITC has done more to lift people out of poverty than all of the Great Society programs combined. It's cost is likely negative except in recession. It rewards work over no work. It reduces burdens on other entitlement programs. Long live Uncle Miltie!!
Keith, I really respect your work, but I think you are dead wrong on the effectiveness of a Keynesian stimulus.
I wrote a recent paper titled "Why Government Spending Does Not Stimulate Economic Growth: Answering the Critics" at http://www.heritage.org/Research/Economy/bg2354.c...
I'd love to find out what you think of my argument that government spending *cannot* stimulate short-term growth.
Best,
Brian Riedl
The Heritage Foundation
Early in 2009 an interesting paper appeared, you can find it here:
http://www.volkerwieland.com/docs/CCTW%20Mar%202….
The New Keynesian macro model used by the authors showed that the stimulus was ill designed and that the administration's economic analysis of impacts was seriously flawed because it used an outdated methodology that showed a never ending series of essentially flat multiplier effects (highly unrealistic). It turns out that this little paper showed exactly what to expect — limited impact and a crowding out of both consumption and investment. This is exactly what the economy has experienced.
I should elaborate that the drivers of real GDP growth have been exports and inventory rebuilding. I don't think the stimulus bill had any impact on either of these components. In fact net government purchases of goods and services was negative in the 4th quarter. According to the above paper the 4th quarter of this year the stimulus bill will have a net impact of .65% on GDP but Romer and Bernstein estimate the impacts will be 3.6% a 6-fold difference. We can see that the divide is wide and administration credibility is diminishing as the real world confirms the limited impact of the stimulus bills.