The Limitations Of Economic Theory In The Current Environment

Includes: EU, FEU, GREK
by: Brian Romanchuk

I keep most of my writing fairly theoretical, and not too focused on reacting to the latest events or controversies. However, it seems striking how the major questions at present are driven by factors that are not readily dealt with by economic theory. In particular, the situation in Greece is almost completely a question of political economy. Additionally, the amount of slack in the United States labour market is of critical importance from the standpoint of most theories, yet we have no good way of measuring it.

The Role Of Theory

Economic theory provides a framework to think about economic events, but I have limited hope that theoretical models can provide highly accurate forecasts for the near term. I believe that most academic economists would take such a view as being unremarkable, but this is not the case in market economic analysis. The financial media wants economists to have strongly held point forecasts about what will happen to the economy within the next few months, and once again, demand creates its own supply. Using a somewhat novel economic theory helps economists with their branding, which leads some to attempt to stretch the predictive power of their preferred theories to the limits.

Whither Greece?

At the time of writing, the outcome of the referendum in Greece is unknown.

My view is that the best outcome for Greece is for a "no" vote, and the rest of the Euro area cutting Greece out of euro are payment mechanisms. The Greek government will have no choice but to take control of the situation, and release a "parallel currency", although it may be billed as a temporary step. Some means of blocking the flight into euro currency notes to relieve the pressure on the banking system would be needed. Furthermore, some external intermediaries will need to step in, and be willing to trade foreign deposits for Greek deposits, in order to allow external commerce to once again function. Since Greece will need to default on its existing debt, the motivation of such intermediaries would probably have to be political (for example, a foreign central bank aiming to develop commercial relations with Greece).

The alternative appears to be a collapse of the Greek government, with the situation reverting back to where it was in December, but with nobody offering a plan to deal with the situation (except perhaps the extremist parties). Even though the other European governments get rid of a leftist party that they obviously dislike, they have racked up even larger financial losses as a result of the Greek deposit flight. Destroying the Greek economy via inducing a bank run lowered the extremely slim chances that official creditors would ever be paid back.

Even at this late date, economic theory tells us little about what will happen. The euro is still a fiat currency, and so it was always possible to avoid this outcome. All the ECB needed to do was to buy Greek bonds, which is what it did to stabilise other peripheral economies. If the interest rate on Greek bonds was negative, it would be hard to see how Greek debt could be "unsustainable." The decision to induce a Greek bank run was driven entirely by partisan political reasons, and possibly to act as a "horrible example" for other peripheral countries. (The logic behind that justification is puzzling, as it seems to imply European policy makers should study a little bit more of their own history. Creating a situation where one side knows that negotiation is useless does not raise the odds of benign outcomes.)

With respect to the euro area, the best theoretical approach may be to look at the history of economic thought, so that we can understand the rather deficient mental processes behind policy setting. This is the approach of Bill Mitchell, whose book Eurozone Dystopia: Groupthink and Denial on a Grand Scale I have just started reading. Economic theory can tell us how the euro zone could be fixed, but that is just wishful thinking given the nature of European groupthink.

Whither The American Labour Market?

The "output gap" is an extremely important concept that is used to explain inflation dynamics within mainstream economics, and variations are used even by heterodox economists. (I think the basic idea is reasonable, but I have some doubts about how it is calculated in practice. I will discuss this in later articles.)

The output gap is supposed to give a summary measure of whether the economy is growing faster than "potential" and causes inflation to rise due to overheating. Older methods of calculating the output gap using the Hodrick-Prescott filter were embarrassingly bad, and so methods using a "production function" have arisen to replace them. The key driver of these methods is some variant of the "unemployment gap," which is the deviation of the unemployment rate from some "neutral" value. (This appears to be a cosmetic difference from old Phillips Curve theories about inflation, but that is another story.)

The well-known problem for U.S. economy-watchers is that the unemployment rate has been falling due to the ongoing collapse of the Participation Rate (chart below), which hit a new low in June. Although some of the fall in the Participation Rate was the result of demographics (older age cohorts have a lower participation rate), I am in the camp that this decline was the result of weak demand. (I discussed the Participation Rate controversy in this article.)

Since the Unemployment Rate as a stand-alone indicator appears dubious, we need to look at other methods of judging labour market slack. I will discuss two methods here - creating a composite labour market indicator, and qualitative analysis. (There are other approaches, including the highly popular Kalman filter. I will discuss the use of the Kalman Filter in such a context in later articles, as it is a fairly big topic.)

One approach to measure labour market slack is to create a composite indicator based upon a variety of labour market data series, The Labor Market Conditions Indicator (LMCI) was created by Fed researchers to fill this gap. Although I think the concept is reasonable, the problem is that the indicator is too new. A composite indicator is the result of "data mining", and will always do a good job of explaining the back history (as otherwise researchers would not publish the results). A composite indicator needs to be used in real time for a cycle or two in order to see how well it acts out-of-sample. (Given the tendency towards longer economic cycles, that might be in 2030 or so.)

The second method might be described as "chartblogging": looking at a wide number of data series, and trying to tell a story with the data. The chart above shows what I view as the best summary measure of the labour market (for this cycle, at least) - the employment-to-population ratio. (Other indicators offer a story that are consistent with it, but I do not want to have a long discussion here.) Although it had been rising at a decent clip since the end of 2013, it is now flat (at 59.3%) year-to-date in 2015. Even if we make allowances for demographics, this measure is nowhere near "full employment" or even "overheating" levels. (The consensus focus on the number of jobs created per month according to the Nonfarm Payrolls Report is just another example of groupthink - those monthly job numbers are revised heavily, and are not adjusted for the size of the labour force.)

Although I think the latter method is actually the most reasonable, it is obviously subjective, and does not meet the current fetish for quantitative models in economics. Therefore, there is no solid way of determining whether or not the current labour market is close to overheating, other than by waiting for another five years for data that we can analyse using non-causal methods.

This means that even though there is a fairly broad theoretical consensus that labour market slack matters for the trend in inflation (and hence Fed policy over the medium term), we have no reliable method of calculating that slack.