David Henderson writes in his “Reply to DeLong”:
In the 1981-82 recession, the fall in real wages helped end the recession.
I don’t see it in BLS series Nonfarm Business Sector: Real Compensation Per Hour.
Figure 1: Log nonfarm business sector compensation, deflated using nonfarm business sector deflated. NBER defined recession shaded gray; assumes last recession ends 09Q2. Source: BLS via St. Louis Fed FREDII and NBER.
if you look at unit labor costs you see a decline in 1983. 1983, 2001, and 2008 are the only years when we see declines in unit labor costs in the data.
I don’t want to scare anyone, but the company I worked for in 1981 instituted a 10% wage cut for the troops, and 20% wage cut for management (the idea being they could afford the bigger cut).
This they claimed would “save jobs”, and most of us generally believed it. They said the wage cut would be restored after the recession ended, which would be the cause-effect relationship there.
I can’t recall if they told us that this was the first time that happened since 1921.
It was announced in the summertime while everyone was at work. Don’t recall now if it was at 8:00 AM in the morning or 6:00PM in the evening.
Ian Dew-Becker: Interesting. But Dr. Henderson’s statement was about real wages.
This is wages for people who are actually working, no? The unemployed, who are not working, are not measured, correct? (Their wages are not ‘zero’; they are simply excluded from the survey.) Is that why wage increases have returned to trend since around last May? Nice to know wages are going up and are a good bit higher than before the recession!
And what’s special about non-farm employment? Are farm laborers somehow not worthy of measurement?
Ian: This makes perfect sense even without any fall in wages.
It’s likely that with any decline in employment, productivity will rise as the least productive workers are laid off. This explains a reduction in unit cost of labour for a given unit of output.
Whether this reduces unit labour costs compared with the non-recession scenario depends on employment (at least private sector employment) falling more than GDP, which I believe is correct in the three recessions that you mention. I wasn’t sure how far into 1983 employment continued to decline, but according to this chart, unemployment seems to have continued to rise throughout the year.
Labor’s Share is the real wage adjusted for changes in productivity. The recession began in the 3rd quarter of 1981. As you can see from the data below, the adjusted real wage is increasing leading up to the recession and peaks in the middle of the recession reaching 105.084 in the first quarter of 1982. The adjusted real wage declines in the subsequent quarters, declining to 103.977 as the recession ends in the 4th quarter of 1982. The adjusted real wage decline further in 1983 and 1984 with the beginning of the Reagan boom. You’re right. Henderson should have referred to the adjusted real wage not the real wage.
Major Sector Productivity and Costs Index
Series Id: PRS85006173Duration: index, 1992 = 100Measure: Labor ShareSector: Nonfarm Business
Year Qtr1 Qtr2 Qtr3 Qtr4 Annual
1981 101.862 102.935 102.137 103.112 102.515
1982 105.084 104.505 104.641 103.977 104.543
1983 103.222 101.359 99.971 99.971 101.081
1984 100.600 100.056 100.182 100.291 100.282
Although David’s statement may not be literally true, real wages certainly declined relative to trend by remaining stagnant for roughly five years.
I think it’s worth noting that real wages were declining from 2007-mid 2008 before prices fell well below expectations. After 2008 when prices fell dramatically real wages shot up. There is definitely a strong argument that real wages are too high now and a correction might be necessary (unless AD can be raised to justify/lower real wage rates).
Cameron: I agree. If we redefine the statement as real wages relative to some trend, he’s right. A linear trend over the entire 47Q1-09Q4 period will do the trick, although an HP filter will not.
Let’s not forget one of the greatest government work projects ever enacted: President Reagan’s Arms race and military build up. To get a good idea of its impact, one just has to look at the impact it had on the economy when the Cold War ended.
And later in the 1980’s to further increase employment, President Reagan then declared war on drugs, and that as well helped the employment situation in 85-86, as unemployment was plateauing around 7%.
They have a theory that cutting wages leads to employment growth.
Because they could find no evidence to support the theory they switched to claiming it was real wages.
So, yes real compensation did fall in the 1982 recovery when it is deflated by the CPI-U.
But it is the only recovery in the post WW II era when this happened. So their theory is correct if you count this one time it happened and ignore the other eleven post WW II recoveries when real wages rose in the recoveries. So they can claim that their theory is correct because it happened in one time out of twelve.
Menzie and Cameron,
If we consider real wages relative to trend, then we still have not helped Henderson make his argument. Real wages fell away from trend years before the double recession began, and rose more during the second dip than at any time since that little blip that looks like 1978. For the story to work in its clearest possible way, real wages should have been rising into the recession, then fall toward the end. In fact, just the opposite happened.
This is the typical foolishness that too many economists engage in. If the market sets the rates then falling or rising rates will cause employment to be optimal. If the government influences wages to be either higher or lower than the market you will have a decline in the economy. Higher wages will lead businesses to reduce labor while lower wages will could cause bankruptcies (as DeLong claims though it is doubtful this would be a serious problem).
This is the problem with those who desire interventionist policies. They think they know more than they can ever know. It is the destruction of hubris. The bottom line is wage and hour laws distort the market and destroy wealth. It is destruction in the name of compassion.
How can we make sense of the claim that declining real wages will help end a recession? Let’s go back to the underlying theory.
The Keynesian view is that, in a recession, the (aggregate) labor market is in disequilibrium–the labor demand curve has shifted to the left, but real wages have not fallen (or have not fallen enough), resulting in unemployment. So, if real wages begin to fall, employment will rise (we move down along the labor demand curve) and unemployment will fall (as we also move down along the labor supply curve).
But. The decline in real wages reduces the incomes, at least of those who already had jobs. As a consequence, they reduce their consumption spending, and the agregate (product) demand curve shifts to the left. Unless the (re)employment effect is strong enough, aggregate income will fall, a negative multiplier effect will kick in, the aggregate labor demand curve will shift (again) to the left, and we do not get the boost to employment we expected.
Can we put some (not entirely hypothetical) numbers on this? Let’s try.
Suppose employment is about 138 million and unemployment is about 15 million. Suppose the (aggregate) own-wage elasticity of demand for labor is -0.5 (Ehrenburg and Smith, Modern Labor Economics). Then a 5% decline in real wages will increase employment by 2.5%.
What effect does this have on aggregate labor income? Aggregate labor income will *fall* by about 2.5%. And we would therefore expect a decline in consumption spending, and a further leftward shift in the labor demand curve.
Declining real wages will lead to increased labor income only if the (aggrgeage) demand for labor is own-wage elastic, and almost all the empirical work with which I am familiar suggest that it is not. So it’s not clear what the mechanism is that would lead from declining real wages to icnreased aggregate demand and increased employment.
The problem is that, in the aggregate, the labor demand curve is not independent of the level of the real wage. Because the level of the real wage affects aggregate income, aggregate consumption, and aggregate demand.
Ending The Recession
From my perspective, the per capita real US economy and mean real disposable income have gone nowhere since about 1970, corresponding to America’s persistant current account deficit.
America has largely paid for its oil addicted profligancy through the sale of its industrial assets, hence manufacturing’s decline as a fraction of GDP and the rise of foreign controlled domestic manufacturing. In the absence of a huge mineral strike, or lucrative new technology, it is inevitable that profit margins and wages will have to drop in the US in order for domestic businesses to survive.
I regard discussions about policy manipulation of the economy as a side-show. Economic management’s primary function being the maintenance of the current oligarcy. Eventually, the dead-weight of all the markers in the game makes continued play impossible-
Would be useful to compare real compensation with an appropriate productivity number in a similar graph.
If you take into account house prices, wages and the mortgage rate, there was a big shift in the affordability of housing. In the 70s, it took 600 hours of work a year to buy a house. It went up dramatically when interest rates were raised to fight inflation, but then settled down around 800 hours where it has remained. If you assume about 2,000 hours of work a year and a 1/3 of one’s income for housing, that means a second income became necessary.
Of course, there was also a long term trend starting in the mid-60s in which wages shrank as a fraction of the GDP per capita.
I see Wall Street is leading the way in cutting real wages in order to reduce unemployment…
Joy unconfined as ‘capitalism works just perfectly’
Wall Street bonuses shoot up 17 percent in 2009
By DAVID B. CARUSO and MICHAEL GORMLEY, Associated Press Writers
Tue Feb 23, 6:24 pm ET
NEW YORK Wall Street bonuses climbed 17 percent in 2009 to $20.3 billion as many of the investment banks that were bailed out at taxpayer expense reported blowout profits.
The announcement Tuesday by New York Comptroller Thomas DiNapoli was likely to outrage many Americans who are barely getting by. And it happened on the same day that private economists reported a plunge in consumer confidence a blow to hopes that spending by shoppers would help speed up an economic recovery.
I’m sure you’ll discount this but I think the more relavant price indicator at that time was gold. If we believe the changes made to CPI methodology since the 70s make sense we know the CPI was a flawed measure at the time. That being the case should we use it to determine whether real wages rose or fell during the time? Considering the proximity to the end of Bretton Woods, I’d say that a lot of people at the time watched gold as an indicator of inflation. If that is true, maybe the price of gold can give us a clue as to what businesses were responding to.
Volcker became Fed chair in August of 79 when the price of gold was roughly $300. He pursued his monetarist experiment and gold moved to $800 by January 80. Gold started to fall early in 80 as it looked like Reagan would get elected and cut taxes. It fell all the way back to $300 by July of 82 as Volcker continued to concentrate on money supply while ignoring demand. Considering how gold was viewed at the time, the run up was inflation and the fall back was deflation. Volcker finally eased in August of 82 due to the Mexican debt crisis.
So if the fall in gold from $800 to $300 was seen as deflationary, it could certainly be perceived as a rise in real wages and the easing in August of 82 which caused gold to rise from $300 to $400 could have been perceived as a fall in real wages.
I’m not saying that is what Henderson meant. In fact, I’m almost certain it isn’t. But I think we forget how prominent gold was to people back then and the role it played in the economy. If the perception of inflation and deflation was more related to gold then (and I think it was) then the movements in gold are probably a better indication of what businesses were thinking and probably had a bigger effect on hiring and firing decisions than they do now.
GNP wrote:
Would be useful to compare real compensation with an appropriate productivity number in a similar graph.
GNP,
A very wise and insighful observation. High real wages come from high productivity. Higher employment comes from higher productivity (demand).
A very wise and insighful observation. High real wages come from high productivity. Higher employment comes from higher productivity (demand).
If that were true, the extremely high productivity of the last 3 quarters should soon translate into a burst of new jobs.
Somehow, that is not happening.
I think the people who are still employed in the private sector are resigned to the fate of doing 20% more work for the same pay.
The public sector, of course, is the reverse (20% more pay for the same work).
Higher productivity means fewer workers are necessary to meet our basic needs.