Friday, July 22, 2011

Popular (But False) Economic Myths

As the soap opera on the Debt Ceiling continues I thought I'd write not on that but rather on some of the underlying assumptions that are being battered around as truths but really are complete nonsense.

Balanced Budgets

Personal Budgets

At the level of the individual, the best model is the life-cycle model of Modigliani and Brumberg. The theory posits that earnings over the life-cycle are humped, rising from when the person enters the labor market to middle age, plateauing, and then falling off at or near retirement. Consumption over the life-cycle is flatter than earnings. Early on there is borrowing. In midlife the early loans are paid off and further saving ensues. In latter life there is dissaving where accumulated assets are sold off to finance consumption during retirement. Rough budget balance happens over the life-cycle but not in any single year. Uncertainty regarding future earnings and end of life make it impossible to get more than rough balance. For this reason, rational behavior predicts holding a cushion to self-insure. Bequests may emerge for that reason rather than from a fundamental desire to jump start the next generation. On the flip side, it is possible for people to outlive their estates, an increasingly important problem.

This theory at the individual level coupled with a demographic model of population growth was articulated in the early 1950s and is remarkably robust. At the individual level it predicts that net worth also will be humped and should peak at or near retirement. Some data on net worth confirms that seniors have higher net worth on average than those who are younger and participating in the labor market. That is a good thing and as it should be.

The theory so described makes sense for middle class people who do earn in excess of their reasonable consumption expectations at mid life. The theory breaks down for low wage earners and the unemployed whose income never rises above covering the bare necessities. They do not accumulate at all, even at mid life, and so may have close to zero net worth. Such people live from paycheck to paycheck. So these people are near budget balance - on a monthly (if that is the pay period) basis. Credit markets function much differently for these people. It may allow them to expand consumption temporarily. It may also push them closer to personal bankruptcy. As we all know, a greater fraction of the population is in this second category than was the case 30 years ago because incomes, particularly for those lacking a college degree, have largely stagnated.

Government Budgets

There is the question of whether government budgets can be conceived in a way similar to personal budgets. Let's mention three factors that provide challenges in doing so - changing demographics, uncertain economic growth for the economy as a whole, and the business cycle, which is also hard to predict. Then, of course, there is politics (I like to call it sausage being made.) The life-cycle model relies crucially on the rationality of individual decision making. There is no reason to assume that the sausage produced is the rational outcome of collective decision making. So what I will write here is first aspirational, as if rationality did arise, and only after that will I discuss how the argument should be modified to accommodate politics.

Government receipts will vary over the business cycle. It is sensible that government spending should be flat or perhaps somewhat counter cyclic over the cycle. This means deficits during economic troughs and surpluses during booms. The popular notion of "irresponsible government" almost certainly has things backwards. It is not irresponsible to run a deficit during hard times. It is irresponsible to make long run increases in spending during good times instead of using the the surplus funds to either pay off past debt or to accumulate savings to be spent during the next downturn. Since one business cycle happens over several years, having annual budget balance constitutes an overly constrained solution, one that is the analog to the individual case of a low income person who lives paycheck to paycheck. It may be necessary if the government entity becomes too impoverished, but otherwise is not the ideal solution.

Now lets turn to the making of sausage and the core problem of "excessive exuberance" during booms. Recognizing the problem, it then may make sense to constrain the spending in some way. Budget balance as it is normally conceived, however, doesn't really do that because the problem is fundamentally about over estimating future receipts, using current receipts as the future forecast. (Similarly, during troughs future receipts are under estimated, which is why some want to make draconian spending cuts.) A do-able way to achieve this would be to use a moving average of past receipts, say over the last four years, and tie current spending to that. This would both dampen exuberance and lessen the necessity of cuts during downturns. Further, in the bargaining that happens between government and its employees about wages and benefits, it would help to fix future expectations and so make those agreements more reasonable long term. Do note there is no perfect way to do this. The choice of four years of receipts in the moving average is arbitrary. Business cycles are not uniform in duration. So this is a rough accommodation only, but it is better than annual budget balance as it is normally conceived.

One other point here is the possibility of temporary spending increases that are dictated by other than business cycle reasons. I'm thinking particularly about going to war, such as in Iraq and Afghanistan, but it also might be due to a response to a natural or man-made disaster at a scale well beyond which normal government spending can accommodate. It then might be reasonable to have temporary tax surcharges to finance the increased spending. At present, the model is to deficit finance such spending. Viewed as a sausage being made problem, the issue is whether it is too easy for the politics to generate a solution where spending does increase on a temporary basis without any matching revenue. A tax surcharge makes voters more aware of the opportunity cost of such spending. Politicians then have a different calculation to make, about balancing the necessity of the spending with the unpopularity of the tax increase. In turn, knowing this sort of calculation will arise in the future, they may be more responsible with current day decisions that impact the likelihood of such temporary spending increases down the road. Just as means of illustration, if the clean up of the Gulf Oil spill were financed by a temporary tax increase that was advertised as such, legislators may have been more reticent about subsequently encouraging off-shore drilling or about insisting on tougher safety measures should off-shore drilling be allowed. The deficit finance approach to the problem provides the opposite incentive.

Tax Rates and Economic Growth

Break government spending into three components: (1) investment in the real economy (basic research, education, and infrastructure), (2) social insurance (Medicare, Social Security, Unemployment Insurance, etc.) and (3) military and security. I won't have much to say about the third item. I list it here just to argue that it too is clearly necessary and it needs to be accounted for.

There are three key points to be made in this section.

First, government investment in the real economy promotes economic growth. If such investment were free, we would definitely want more of it. Of course, it is not free and must be paid for - with taxes. The question then is whether we are getting a good balance between such government spending and tax revenue collected and how can one tell. Looking at tax rates only and saying they must be reduced without considering the growth impact of the spending is not a balanced way at addressing the solution. Further, where we are talking about basic research and education, the economic payoff is surely many years down the road while the taxpayer pain is now. This creates a potential for bias and under investment in the public good.

Second, there is the additional impact of lags in policy effects that are not due to the return on investment but rather to how government budgets impact the real economy. Bush 1 raised taxes, but the macroeconomic benefits from that didn't accrue till the Clinton Presidency was well underway. Financial deregulation happened a lot under Clinton, with impacts mainly under Bush 2. Bush 2 lowered taxes.... well you get the idea. These lags are not perfectly predictable. And macroeconomic phenomena are the consequence of multiple causes. But many politicians speak with a certitude about lowering tax rates and promoting economic growth. This is Republican orthodoxy. It is hokum.

Third, on the social insurance front, think of that in conjunction with life-cycle consumption-saving of the individual. For those people who do live in accord with the Modigliani and Brumberg model, it is important to note that social insurance and self-insurance qua savings are substitutes. If social insurance benefits and contributions are cut in equal measure, private savings should increase to offset those reductions. Absent an efficiency argument that says social insurance is more efficient than self-insurance or vice versa, the overall consequence should have no impact on economic growth at all. You can have higher taxes and spending or lower taxes and spending and get the same growth. It is just a relabeling thing; that is all. For those who live from paycheck to paycheck, however, there is a real consequence. This is precisely why government spending cuts borne by the low income in the population should be resisted.

Of course, there is a serious problem now with social insurance. It is not balanced. People can expect to get out much more than they've paid in (particularly on Medicare). Balance does need to be restored, by what I've called "moving the goal posts," done by a combination of means testing the insurance premium or the benefits and raising the retirement age.

Keynesian Stimulus

There is much debate on whether in the presence of high unemployment government spending in excess of tax receipts boosts GDP by increasing aggregate demand and if the answer might be both yes and no because it matters if the stimulus is achieved by raising government spending or by lowering taxes. In political terms, did the Obama stimulus work? Let's grant that the "what if there was no stimulus" scenario can't be laboratory tested, so the answer to the question is necessarily hypothetical. Nonetheless, we can use the life-cycle framework to consider the consequences of the stimulus on individuals. Doing so, there is the further issue to consider of what happens to the economy when a stimulus is applied and unemployment is at normal levels. Again in political terms, the Bush tax cut was also stimulus. Did it work? Is it possible that one of these did work and the other didn't?

Coupled with these questions about stimulus is the proper role of monetary policy (stimulus here refers to fiscal policy). A fairly good case can made that for ordinary business cycle fluctuations monetary policy is the better instrument to use and fiscal policy should focus on long term objectives rather than fine tuning of the economy, because changes are harder to implement and the lags are longer with fiscal policy. The case for fiscal policy to lower unemployment, then, is strongest when interest rates are already near zero and expansionary monetary policy is limited in its effectiveness.

There is also the question on measuring whether a policy was effective by seeing if the consequence is mainly on lowering unemployment or instead mainly on raising inflation. So what I mean here is that an ineffective policy would have consequence neither on unemployment nor on inflation. An effective and good policy would lower unemployment. An effective but pernicious policy would raise inflation. In the late 1970s, when I was in graduate school, the feeling was that both fiscal stimulus and monetary expansion would raise the inflation rate, which would be captured by how fast nominal wages were rising. In the last decade the inflation rate by this measure has been quite modest, but there have been bubbles, most notably in housing, and hyperinflation in certain sectors of the economy, health care for example. You might call this overheating the economy.

When an individual experiences a tax cut, the alternatives are either to spend the additional after tax income or to save it. In a fully rational framework the individual would calculate the increase in lifetime income as a consequence and then adjust current consumption according to the life-cycle model. If the expectation is that government has to raise taxes in the future to pay for the tax cut it is giving now the effect on the individual's lifetime budget constraint is nil so current consumption should remain unchanged and hence the tax cut increases saving. However, the argument here is made under the usual ceteris paribus (all else equal) assumption. If contrary to that assumption the individual experiences substantial capital gains or losses to their wealth, from changes in the stock market or in the values of their homes, that too will effect lifetime wealth and such changes may swamp the direct effect of the stimulus.

We have more history to look at in tracing the consequence of the Bush Tax Cut than in doing likewise for the Obama stimulus. Here is a bit of that history. We had a modest recession in 2001, a consequence of the burst of the bubble. In that year we had the first of the Bush Tax cuts. The stock market declined well after the official recession ended. It reached it's low in 2003. So people may have thought of the years 2001 - 2003 as gloomy from an economic view. (Of course, 9/11 contributed to that view.) There was a second Bush Tax cut in 2003. The personal saving rate had already drifted down substantially in the 1990s - as the stock market appreciated the saving rate dropped. See the diagram below, which comes from a report out of the San Francisco Fed.

personal saving rate

Here is more than 30 years worth of data on personal saving rate, a monthly time series. It is from the St. Louis Fed. There are blips up in the rate, but mainly a downward trend after the second Bush Tax cut until the burst of the housing market bubble, especially during 2004-2007. So my conclusion is that this stimulus was effective, but it was pernicious.

The Great Recession started in 2007 and lasted through 2009. In 2008 the personal saving rate began to rise. This was before the Obama stimulus. It was an individually rational response to the decline in the economy, but unfortunately it was collectively pernicious, due to what is known as the paradox of thrift. The economy needed more spending then, but it was getting less. So it contracted.

The Obama stimulus started in 2009 and there was further stimulus in the lame duck session at the end of 2010, via extension of the Bush Tax cuts and relief on the payroll tax. We know the spending part of the stimulus took a while to work its way through the system. We've seen State governments making painful cuts this past year as the stimulus concluded. Through the third quarter of 2010 personal saving rates were about two points higher than in summer 2008. They have since come down a bit. It is hard to say looking at this what role the Obama stimulus had on those personal saving rates, since they started to rise before he took office. Two possible explanations are: (a) this was in anticipation of the stimulus, in which case it was offsetting the tax cut part and (b) this was anticipation that the economy would worsen as measured by the unemployment rate, which would rise. I don't see how you can parse one explanation from the other based on the information available. We do know the economy has limped along in this interval. But, quite conceivably, it could have gone into free fall with the Great Recession still ongoing. That's impossible to know.


Political rhetoric clearly has its purpose and politicians talking about the economy should be seen as such. But repeat the same phrase again and again and it appears to take on a meaning of truth. It would be better to learn our economics elsewhere and to let long term thinking about the economy and economic policy to disregard these popular myths. Clearly that is not possible in the next couple of weeks but perhaps it will be possible next summer during the election season.

No comments: