In this week’s report we would like to pay attention to the debate of unemployment versus inflation, this from an alternative angle by using behavioural economics. The findings will be applied to the current macro-economic situation in the US.
Among classic economists, such as Keynes and Samuelson, there was a strong belief in the positive relationship between low unemployment and permanent high growth, highlighted by the Philips curve (Figure 1) which shows high levels of inflation only to be reached when unemployment is exceptionally low.
Figure 1: Illustration of Philips Curve of US in 1960.
Source: Wikipedia
Friedman was one of the first to neutralise this thesis by introducing the concept of ‘money illusion’, which refers to the tendency of economic agents to think in nominal opposed to real terms. This is a typical pattern among monetarists as they make the assumption in their theories that all economic agents act rationally.
This assumption has been attacked over the last 5-7 years as behavioural scientists are gaining importance in the debate. Proof of this is the Nobel Prize reward to Kahneman in 2002, for having integrated insights from psychological research into economic science, especially concerning human judgment and decision-making under uncertainty. (1)
However, the money illusion does play an important role in the mechanics between unemployment and inflation. With the help of behavioural finance we will quickly notice that money illusion is the missing link in the natural rate theory.(2)
In the footprints of Kahneman, G. Akerlof and R. Shiller have done extensive research on this subject and give further guidance on this issue in “Animal Spirits”. Once again it is like a place where tectonic plates come into contact with each other. In this debate it’s the Chicago School under the guidance of the great and late Milton Friedman on the one hand versus the introduction of irrational behaviour to text book economics by the behavioural scientist.
Of course, mainstream economists do not like this approach as it becomes fairly impossible to model or quantify economic relationships when elements like uncertainty are being added to their models. They think in terms of probability which can be mathematically formulated. However, uncertainty, like intuitive behaviour, can not be expressed in a mathematical equation. Here psychology and social group behaviour will lead the way.
The inflation-unemployment debate follows a similar rationale. In the current economic climate jobs are at risk and employers are cutting back on labour as excess capacity outpaces demand. However before an employer (contrary to what many socialists believe) will decide to lay off people, he will look at other options such as reducing productivity which basically comes down to cutting back on the number of working hours and even salary cuts.(3)
Certainly in a deflationary environment wage cuts could be more defendable as disposable income rises as opposed to falling prices. Unfortunately there is something fundamentally rigid between this relationship as we all consider wage cut s as being highly unfair. Of course this is an intuitive approach but statistical date underlines this thought.
“Downward money rigidity can be easily detected from data on wage changes. All one needs to do is look at how often wage changes of different magnitude occur. Suppose we see that wage changes bunch at exactly zero, and there are many more wage changes just above zero, than just below zero. Then we can conclude that employers stop to think before they give their workers wage cuts.”
Further insights are given by Truman Bewley who did a similar qualitative approach.(research conducted in 1999 based upon 334 interviews)
He came to a similar conclusion that in the employers opinion workers would view wage cuts as unfair. They would reduce their commitment to their jobs. Furthermore when the economy revived, they would still be angry and thus more likely to quit.
Employees would only accept exceptional circumstances, such as eminent bankruptcy, as a last effort to save their jobs. Mainstream economists have finally accepted this thesis of wage rigidity but still consider the impact as marginal.
Akerlof did research on this to what extent wage rigidity plays a role. He and his colleague Dickens made simulations on the impact on unemployment of going from 2% to zero inflation. In the benchmark simulation, a permanent reduction of inflation from 2% -> 0% increases the unemployment permanently by 1.5%.(4)
A back-of-the-envelope calculation shows why these results occurred and so robustly. “If workers resist wage cuts, when inflation is lower, their wages will be higher. In the benchmark simulation they would be higher 0.75% if inflation would be 0% rather than 2%. The 0.75% on wages translates into an increase of unemployment by 1.5%.” (5)
How do we know? Akerlof continues that there is a rule of thumb that comes from the Philips curve itself. It takes a 2% increase in unemployment to reduce inflation by 1%. Therefore to neutralise the 0.75% cost increase of the companies, unemployment must rise by 1.5%. This argument is made precisely, with the underlying equations shown, in both Akerlof (1996-200) and Akerlof and Dickens (2007).
If we apply these findings on the current economic data of the US we come to the following findings.
US Inflation dropped by a staggering 6.3% (this reflects a drop from its high at 5.6% in July 2008 to a deflation level of 0.7% late in April 2009). Let us now compare this with the US unemployment rate at the respectively dates. July 2008 unemployment was at 5.8%. April 2009 unemployment stood at 8.9%. As unemployment is a lagging indicator, this would mean the US unemployment rate could still potentially rise to 18.4%. This is under the assumption that the Fed does not succeed in fighting deflation.
Of course a number like this needs to be double checked with reality as well. Our own reference in this case is Japan during the 1990s. The December 1991 inflation was 2.7% at the time. This continued to drop over the next three years to a negative number of -0.3% at the beginning of 1995 and stayed there for the next five to six years. Simultaneously unemployment rose from 2% to 6 % over that period.
This is a much lower factor compared to the study of Akerlof, which is 1.33. For further reference we also checked the Philips curve of Germany, which showed an even lower factor of approximately 1.15.
Important to note in the comparisons of wage inflation versus unemployment among several countries one needs to take into account the influence of unions. We can conclude that the factor in countries where the unions have significant power, such as Japan and Germany, the factor will be rather low. A trivial reasoning goes for countries with low union influence such as the US.
In this case let us assume that the factor of Akerlof is overoptimistic and that the relationship is slightly higher than Japan, for example 1.5. In this case we still come out at a potential unemployment rate for the US at 15.25%. Even a similar factor of 1.33 as in Japan would leave us a number of 14.18% for the US.
The major conclusion of this exercise is that unemployment in the US has much further to go in the coming years under the assumption that the Fed does not succeed in re-inflating the economy. This has two major consequences. First of all we need to take into account that the US will face a prolonged period of below par growth. The second and more eminent consequence will be on the state of the banking industry. Remember that the US Government based their stress-test of the solvency of the US banking industry on the assumption of an US unemployment rate of 8.8% which is already lower than the current rate.
Going forward the unemployment rate has further to rise and as a consequence US banks will face another capitalisation round.
This also means that in the current environment the inflation hype is a bit too early to jump on the band wagon. Of course there will be huge inflation concerns 2-3 years from now, but at this very moment we are still in a deflationary environment.
The current rise of US Treasuries and inflation has more to do with the concerns the market has on the aftermath of the astronomical debt build up that now needs to be issued. To put it into perspective, the US government needs to issue $ 1.8 trillion over the next 10 months. Last year this was “only” $ 400 bln. Take into account that also the UK government needs to come to the market with close to a trillion. The EU governments and Japan will have similar amounts and soon we get a situation where everybody is rushing for the same fire exit. It is this which causes great concern and nervousness.
But at the same time, as unemployment rises, salaries remain contained and have a considerable deflationary impact.
This is a great dilemma for central banks who see their QE initiatives not working as the long end of the curve is running away from them..
1. "Nobel Laureates 2002". Nobelprize.org. http://nobelprize.org/nobel_prizes/lists/2002.html. Retrieved on 2008-04-25.
2. The natural rate theory describes this mechanics as mentioned above by arguing that Actual unemployment cannot fall below the NAIRU, and the inflation rate is likely to rise quickly (accelerate) in times of strong labor demands during periods of growth. It is sometimes referred to as the "natural rate of unemployment" as well, although this term describes an estimated unemployment rate derived from the market's actual performance while the NAIRU is calculated from the Philips Curve.
3. See studies of wage stickiness in Australia, Japan, Germany, Switzerland, USA an UK by Kahn, Card and Hyslop (1997)
4.Similar results were obtained from econometric estimations and in hundreds of other simulations, in which parameters were chosen randomly over a reasonable range.
5. George A. Akerlof and Robert J. Shiller, “Animal Spirits: How Human Psychology Drives the Economy, and Why It Matters for Global Capitalism”, Princeton University Press, 2009
No comments:
Post a Comment