Murphy on Sticky Wages

Robert P. Murphy in his recent response to Krugman on Mises and the Great Depression refers us to the following post to refute the idea that “sticky wages” really are a problem for free market economics:

Murphy, Robert P. 2010. “Do Sticky Wages Weaken the Case for Markets?,” Mises Daily, June 7
http://mises.org/daily/4353/Do-Sticky-Wages-Weaken-the-Case-for-Markets

This post deserves to be read by everyone who wants to see how far Austrian economists (and, incidentally, even mainstream New Keynesians like Mankiw) are from understanding real-world capitalism.

Keynes was clear that even if wages and prices were perfectly flexible this would still be no reliable and automatic cure for involuntary unemployment, and that there could still be failures of aggregate demand (Davidson 1992). Keynes, then, would not have been a New Keynesian like Mankiw.

But the issue here is wage stickiness.

First, Murphy does not even dispute that wage stickiness is a real phenomenon in modern market economies (what is now called a “stylised fact”), but instead wishes to put the blame for it mostly on modern governments.

I will return below to why this is wrong, but Murphy makes an astonishing claim that is utterly untrue:

“Now we have to ask, why do workers hold out for so long without jobs, insisting on wages that no one is willing to pay? After all, the other goods and services in the economy see their prices fall in a speedy fashion even though the sellers of these items depend on them for their livelihood. So if a street vendor knows enough to slash his hot dog prices when demand collapses, why don’t hairdressers accept pay cuts when the same happens to their industry?

What was that? Murphy is saying that real world “prices fall in a speedy fashion”?

Has Murphy ever noticed that in virtually all recessions since the late 1930s throughout the developed world, even recessions tend to be inflationary? Deflation has virtually disappeared.

Relative price rigidity is a fact of life. Even in the terrible disaster that was the Great Depression it was a clear phenomenon.

The reason has been known since the work of Gardiner Means: most prices in modern capitalist economies are relatively inflexible mark-up/administered prices (for the empirical evidence, see Appendix 2 below).

Flexible price setting as required in standard economics textbooks – where prices are set by supply and demand dynamics – is largely shunned by the private sector itself.

A great part of the economy of each modern capitalist nation consists of mark-up pricing sectors, but even though (of course) flexprice markets do exist they are a considerably smaller part of the economy than most people think (perhaps less than 30% in many nations).

And, even though many nations saw price deflation in the Great Depression, even in the 1930s mark-up prices were significant and relatively inflexible as compared with other markets: Gardiner Means, for example, discovered that the administered pricing sector of the US economy had seen price declines of only about 10% during the depression, whereas the more competitive or flexprice sectors had seen price falls of about 40 to 60% (Means 1975).

So even Robert Murphy’s initial assumption that we live in a world of highly flexible prices cannot be taken seriously, and utterly collapses.

But to return to the issue of sticky wages. The main cause of sticky wages is not government intervention.

The empirical evidence that has accumulated over the years shows that people in general object to having their nominal wages cut. But even managers and capitalists often dislike pay cuts. Recent studies suggest that employers avoid pay cuts because they diminish workers’ morale, and then falling morale reduces productivity, amongst many other reasons (Bewley 1999). The evidence of Bewley shows that even employers are often averse to wage cuts during recessions. A nice summary of Bewley’s work on wages is available here.

But moving on, even Murphy’s explanation of wage rigidity during the depression in the US is flawed:

“…why don’t hairdressers accept pay cuts when the same happens to their industry? In the case of the Great Depression, the answer is simple enough: the federal government didn’t allow wages to fall. After the 1929 crash, Herbert Hoover gathered the nation's leading businessmen for a conference in Washington and urged them to allow profits and dividends to take the hit, but to spare workers’ paychecks. Rather than cut wages, businesses were supposed to implement spread-the-work schemes where workers would cut back their hours.”

First, Hoover’s “high wage” policy was not an alien, evil government intervention imposed on unwilling and hostile business people: it was a policy that many largely agreed with, as I note here. Already in the 1920s and 1930s many employers had grown averse to wage cuts.

Secondly, Rose (2010) presents some (admittedly ambiguous) evidence that Hoover’s “high wage” policy actually did not have much effect on the timing of US wage cuts during the depression.

Of course, that rigid wages in the face of falling prices squeezed profits and induced business pessimism and bankruptcy is undoubtedly true, but the solution was not wage cuts.

Why? Keynes showed why in Chapter 19 of the General Theory (Keynes 1964 [1936]: 257–271), and one important reason is that, even when nominal wages fall, this induces debt deflationary effects (especially when levels of private debt are very high), just as Irving Fisher had argued (Fisher 1933; see also Dimand 1997 and Dimand 2011).

Above all, if price deflation is quite uneven across sectors and product markets (as in many countries during the depression), the burden of fixed nominal debt will soar, inducing difficulty in servicing debt to many debtors and actual bankruptcy to others.

A further means by which demand for final goods and services is contracted is that a greater transfer of wealth from debtors to creditors occurs and creditors often have a lower marginal propensity to consume.

And eventually severe debt deflation will cause distress and bankruptcy to creditors and the financial system. So the smooth and rapid market clearing as postulated by Austrians and other mainstream neoclassicals as the cure for high involuntary unemployment simply could not and did not happen.

Murphy cites the recession of 1920–1921 as an example of how (alleged) wage and price flexibility cleared markets and led to a (supposedly) quick recovery. But the recession of 1920–1921 is highly anomalous, as I have shown elsewhere (see Appendix 1 below), and a demand-side explanation is more convincing as an explanation of recovery in 1921. There was no financial crisis, no banking failures, the deflation was probably expected, some actual positive supply shocks, and private debt levels were considerably lower than in 1929.

Murphy is also wrong that the Federal Reserve “jacked rates up to record high levels,” if he is talking about the recovery in 1921:

Discount Rate of the Federal Reserve Bank of New York
Date | Rate
1920
May | 6%
June | 7%
Dec. | 7%
1921
Jan. | 7%
Apr. | 7%
May. | 6.5%
Jun. | 6%
Jul. | 5.5%
Sep. | 5%
Nov. | 4.5%

1922
Jan. | 4.5%
Jun. | 4%.
http://fraser.stlouisfed.org/download-page/page.pdf?pid=38&id=1477

As we can see, the Fed lowered interest rates from May 1921, and a recovery began in August 1921, after looser monetary policy had been adopted.

Then from November 1921 to June 1922, the Fed engaged in unprecedented open market operations to aid the recovery process.

The Austrian obsession with the recession of 1920–1921 is strange, given its anomalous nature. Why don’t they look at the US recessions of the 1870s and 1890s, for example?

The US had no central bank in these years, small government, a gold standard, either no or very limited unemployment relief at best, and (presumably) a greater degree of price and wage flexibility (although even in this period wage stickiness existed: see Sundstrom 1990; Sundstrom 1992; Hanes 1992; Hanes 1993). Yet serious economic problems occurred in both the 1870s and 1890s.

Take the 1890s. The graph below shows US unemployment in the 1890s according to three estimates. For various reasons, Lebergott’s estimates may be the better ones for unemployment rates in this period.

High unemployment continued for years after the shock of 1892–1893.

So what is the Austrian explanation for this? If wages and prices were not sticky, then there was still no rapid recovery and quickly self-adjusting labour market.

If wages were sticky, then not even late Gold Standard capitalism escaped the “stylised fact” of relatively rigid wages.

Either way the Austrian view is damned.

Appendix 1: The Recession of 1920–1921
“The US Recession of 1920–1921: Some Austrian Myths,” October 23, 2010.

“There was no US Recovery in 1921 under Austrian Trade Cycle Theory!,” June 25, 2011.

“The Depression of 1920–1921: An Austrian Myth,” December 9, 2011.

“A Video on the US Recession of 1920-1921: Debunking the Libertarian Narrative,” February 5, 2012.

“The Recovery from the US Recession of 1920–1921 and Open Market Operations,” October 4, 2012.

“Rothbard on the Recession of 1920–1921,” October 6, 2012.

Appendix 2: Empirical Evidence on Administered Prices
“Downward’s Pricing Theory in Post-Keynesian Economics: Chapter 8,” January 23, 2014.

“Mark-up Pricing in South Africa,” January 20, 2014.

“Mark-up Pricing in Sweden,” January 9, 2014.

“Mark-up Pricing in Canada,” January 7, 2014.

“Some More Empirical Evidence on Full Cost Pricing,” December 10, 2013.

“Mark-up Pricing in New Zealand,” November 30, 2013.

“Mark-up Pricing in Australia,” November 30, 2013.

“Mark-up Pricing in Japan,” November 29, 2013.

“Mark-up Prices in Iceland,” November 25, 2013.

“Mark-up Pricing in Norway,” November 23, 2013.

“Mark-up Pricing in Ireland,” November 22, 2013.

“Two Marketing Studies on US Administered Prices,” November 16, 2013.

“Hall and Hitch on Marginal Cost and Price,” November 4, 2013.

“Administered Pricing in the United Kingdom,” October 19, 2013.

“Administered Prices in the Eurozone: Some Empirical Data,” October 16, 2013.

“Gardiner Means on Administered Prices,” June 20, 2013.

“Early Literature on Administered Pricing,” May 8, 2013.

BIBLIOGRAPHY
Bewley, T. F. 1999. Why Wages Don’t Fall During a Recession. Harvard University Press, Cambridge, MA.

Davidson, P. 1992. “Would Keynes be a New Keynesian?,” Eastern Economic Journal 18.4: 449–463.

Dimand, Robert W. 1997. “Debt-Deflation Theory,” in D. Glasner and T. F. Cooley (eds), Business Cycles and Depressions: An Encyclopedia, Garland Pub., New York. 140–141.

Dimand, Robert W. 2011. “Lessons from the 1929 Crash and the 1930s Debt Deflation: What Bernanke and King Learned, and what they could have learned,” in Claude Gnos and Louis-Philippe Rochon (eds.), Credit, Money and Macroeconomic Policy: A Post-Keynesian Approach. Edward Elgar, Cheltenham. 33–44.

Fisher, Irving. 1933. “The Debt-Deflation Theory of Great Depressions,” Econometrica 1.4: 337–357.

Keynes, J. M. 1964 [1936]. The General Theory of Employment, Interest, and Money. Harvest/HBJ Book, New York and London.

Means, Gardiner C. 1975. “Simultaneous Inflation and Unemployment: A Challenge to Theory and Policy,” in Gardiner C. Means et al., The Roots of Inflation: The International Crisis. Wilton House Publications, London.

Sundstrom, William A. 1990. “Was There a Golden Age of Flexible Wages? Evidence from Ohio Manufacturing, 1892–1910,” The Journal of Economic History 50.2: 309–320.

Sundstrom, William A. 1992. “Rigid Wages or Small Equilibrium Adjustments? Evidence from the Contraction of 1893,” Explorations in Economic History 29.4: 430–455.

Mosler versus Murphy Debate Video

The video of the Warren Mosler versus Robert Murphy debate is below.

Watch live streaming video from clsit at livestream.com

N.B. the video does not seem to be working! If anyone knows the correct link, you can post it below.

Robert Murphy Takes Issue with my Reading of Empirical Data on the US Stimulus

For the interested reader, you can find his posts here:

“Just the Facts, Ma’am: “Testing” Keynesian Theory,” 29 April, 2013.

“Believing Is Seeing, Part II,” 29 April.

The unrealistic assumption that he appears to be making in his criticisms of me is this: that real output and private investment must have started expanding immediately after the stimulus began.

But it is obvious that government spending takes time to induce changes in private investment.

Why should there be an immediate and instant movement? I know of no Keynesian economist who has ever thought that there should be instant effects on private investment from stimulus.

With reference to the graph of US private-sector investment Murphy posts, I do not find it surprising that private investment continued to contract in 2008 and early in 2009. The shocks to business confidence were very severe indeed in 2008: probably worse than in any other recession after 1945.

But private sector investment did turn around in mid-2009: after 6 months or so of stimulus spending, which stabilised demand for products. There is a clear trend of rising private sector investment with rising government spending for years after mid-2009.

If private sector investment had continued to contract for years after the stimulus, then Murphy would have empirical evidence to support his anti-Keynesian, Austrian case. But that is not what the data show.

Murphy on Keynesianism and Great Recession

His views on the recent crisis in the US are given here.




The problems with his video are as follows:

(1) Romer’s estimates of the level of unemployment with stimulus were wrong, but given that you cannot precisely predict the future or do so with objective probability scores, this is hardly much of a point.

Romer and others underestimated the severity of the recession. This does not disprove the fundamentals of Keynesian theory.

The whole underlying argument of Murphy’s talk is that the stimulus “failed.” If this were so, then why did the US recession end in 2009 and positive real output growth resume? Failure would have been a continuing recession even in the face of fiscal expansion.

(2) It is fascinating how modern Austrians are now driven to extremes so absurd that they appear to deny that government increases in national income can raise output and employment.

In fact, not even Hayek or other die-hard Austrians have denied this. Take this statement by the extreme Rothbardian Huerta de Soto:

“What should be done if, under certain circumstances, it appears politically ‘impossible’ to take the measures necessary to make labor markets flexible, abandon protectionism and promote the readjustment which is the prerequisite of any recovery? This is an extremely intriguing question of economic policy, and its answer must depend on a correct evaluation of the severity of each particular set of circumstances. Although theory suggests that any policy which consists of an artificial increase in consumption, in public spending and in credit expansion is counterproductive, no one denies that, in the short run, it is possible to absorb any volume of unemployment by simply raising public spending or credit expansion, albeit at the cost of interrupting the readjustment process and aggravating the eventual recession.

Nonetheless Hayek himself admitted that, under certain circumstances, a situation might become so desperate that politically the only remaining option would be to intervene again, which is like giving a drink to a man with a hangover. In 1939 Hayek made the following related comments:

it has, of course, never been denied that employment can be rapidly increased, and a position of ‘full employment’ achieved in the shortest possible time by means of monetary expansion. ... All that has been contended is that the kind of full employment which can be created in this way is inherently unstable, and that to create employment by these means is to perpetuate fluctuations. There may be desperate situations in which it may indeed be necessary to increase employment at all costs, even if it be only for a short period—perhaps the situation in which Dr. Brüning found himself in Germany in 1932 was such a situation in which desperate means would have been justified. But the economist should not conceal the fact that to aim at the maximum of employment which can be achieved in the short run by means of monetary policy is essentially the policy of the desperado who has nothing to lose and everything to gain from a short breathing space.

Now let us suppose politicians ignore the economist’s recommendations and circumstances do not permit the liberalization of the economy, and therefore unemployment becomes widespread, the readjustment is never completed and the economy enters a phase of cumulative contraction. Furthermore let us suppose it is politically impossible to take any appropriate measure and the situation even threatens to end in a revolution. What type of monetary expansion would be the least disturbing from an economic standpoint? In this case the policy with the least damaging effects, though it would still exert some very harmful ones on the economic system, would be the adoption of a program of public works which would give work to the unemployed at relatively reduced wages, so workers could later move on quickly to other more profitable and comfortable activities once circumstances improved. At any rate it would be important to refrain from the direct granting of loans to companies from the productive stages furthest from consumption. Thus a policy of government aid to the unemployed, in exchange for the actual completion of works of social value at low pay (in order to avoid providing an incentive for workers to remain chronically unemployed) would be the least debilitating under the extreme conditions described above.” (Huerta de Soto 2012: 452–456).

But these days neo-Austrianism appears to deny even this.

(3) What is important to note here is that Murphy attacks New Keynesianism, and has a poor understanding of Post Keynesianism.

(4) By 18.00 onwards Murphy starts to sound like a New Classical and requires Ricardian equivalence and rational expectations for his idea that people cut their spending as the stimulus was implemented in anticipation of future tax increases. But Ricardian equivalence and rational expectations are not even supported by Austrian economics, and one can only marvel at the breath-taking hypocrisy here.

(5) Regarding predictions and the real world: the trouble with libertarians is that they are so focussed on the US, and rarely look at the rest of the world. In numerous other countries, governments used countercyclical fiscal policy and ended their recessions and kept unemployment low: e.g., Australia, Norway, South Korea, and Germany. Of course we hear not a word about any of this from Austrians like Murphy.

(6) Regarding the success of Keynesianism in the 1930s, Murphy is ignorant. There were nations in the 1930s that successfully used fiscal stimulus:

“Keynesian Stimulus in New Zealand: 1936–1938,” September 23, 2011.

“Takahashi Korekiyo and Fiscal Stimulus in Japan in the 1930s,” August 27, 2011.

“Fiscal Stimulus in Germany 1933–1936,” September 3, 2011.

(7) Murphy’s analysis of the Eurozone is flawed.

Murphy’s main point appears to be that Europe has not cut enough from government spending. He grudgingly admits that in Greece, Spain and Ireland total spending has been cut, and economic problems have persisted. Strangely, he ignores the countries that pursued much more extreme austerity: Latvia, Estonia and Lithuania. What happened there is confirmation of the disastrous consequences of fiscal contraction.

Murphy attacks bank bailouts in Europe. At this point, the bizarre nature of Murphy’s Austrian analysis reveals itself. Throughout his talk, Murphy is trying to suggest that austerity and liquidationism will cause recovery better than recoveries caused by fiscal expansion: yet allowing the financial sector to collapse is precisely a measure that would induce the worst depression imaginable.

Liquidationism by definition means the worst types of recessions/depressions possible.