Real wages and productivity revisited

Andrew Jackson says:

I’ve argued for years, with much of the left, that average worker pay has lagged productivity growth mainly because of the increased bargaining power of capital vis a vis labour due to “globalization”, attacks on unions etc etc.

There's another explanation. It turns out that the way real wages are measured - using consumer prices or producer prices - matters a great deal here. Movements in the labour terms of trade - the ratio of producer to consumer prices - can erode or accentuate wage gains associated with increasing productivity. I've done this exercise for the US, so now let's look at the Canadian experience.

The data I'm using here are the from Cansim's Table 383-0008 on labour productivity in the business sector. Here is a graph of labour productivity and real hourly wages calculated using the business sector output deflator and the CPI:

If you use the output deflator, you get a real wage series that tracks labour productivity not too badly - although perhaps not as well as in the US. But it's also pretty clear that the real buying power has not kept up with productivity.

Looking more closely, something appears to happen around 1996. Real buying power stayed almost constant between 1981 and 1996, but it picks up in the latter part of the sample. Let's break this down into two subsamples. First up is the period 1981-1996:

And from 1996 on:

Something appears to have happened to the labour terms of trade sometime in the mid-1990's:

In the 15 years after 1981, the labour terms of trade fell by about one per cent a year, and then stabilised - with a certain amount of cyclical variation - sometime around 1993-1995.

So what is driving these movements? Since the CPI includes imports and the output deflator does not, movements in the the terms of trade are as good a place to start as any. Especially since it was around 1993-1995 that we started seeing a long string of big trade surpluses. I'll return to this point sometime.

Comments

[spencer]

the CPI and the gdp deflator have very different weight because they
are measuring two different things. One of the main reasons the deflator
shows smaller increases then the cpi is because capital equipment has
a big weight in the deflator but not in the cpi.

If you are going to use this approach use the deflater for personal
consumption expenditures rather then the gdp deflator. but also note
that the pce deflator excludes housing--about a quarter of the cpi.

[Stephen Gordon]

The original point was that the standard theory of the firm would use output prices to calculate the nominal wage that sets the real wage equal to its marginal product, but workers use consumer prices to calculate buying power. It would appear that real wages - calculated according to how the firm would measure them - are in fact tracking productivity. The problem is that consumer prices were rising more rapidly than output prices were.

But I will take a look at alternate measures for consumer prices the next time I go through this; thanks.

[Stephen Gordon]

Not in an open economy. Imports don't show up in the output deflator (imports aren't produced here), but they do show up in the CPI. And there may also be compositional things going on with the relative prices of investment and consumption goods. Offhand, I'm inclined to think that second point is not particularly important: increases in the cost of capital goods woule eventually show up in consumer prices. But in Canada, the role of imports in the CPI would be pretty important.

[Bruce Webb]

I have done enough damage on some threads elsewhere so let me just put this in the form of a question.

In the US we have a labor market where by the measure of Unemployment and real Social Security receipts (a nice proxy for wages) held up well through 2006 even as labor productivity ended up being reported at 1.6% for Q4. It still seems to be holding up, the BLS page today giving us a 4.4% unemployment rate. Now we have Q1 GDP coming in at 1.3% which as spencer suggests elsewhere means productivity right around zero. Which if true would mean that at least in the stort term the delinkage between productivity and real wages observed in the decade before 2004 has reversed itself. Welcome news if true, but seems lacking a mechanism.

So if the sharp slowdown in GDP and by extention productivity didn't hit wage rates (that should have shown in Social Security receipts) and didn't hit labor hours (ditto plus it should hit the unemployment rate) who took the hit? As yet it certainly hasn't hit Q1 earnings. Is the shock simply still rolling through the system like a tsunami in open water? If so whose beach it going to take the hit? This puzzle seems to be missing some pieces.