Real wages, the recession that was, and the recession that wasn't

After the tech bubble burst, the United States went into recession, but Canada did not. This isn't to say that we escaped entirely unscathed; there were some nervous moments, and the Bank of Canada saw fit to reduce its overnight rate target from 5.75% to 2.25% between January and November 2001. But if you look at the employment numbers, you only see a slowdown in the rate of growth:

But here's something I hadn't noticed before. If you look at real wages - both the average weekly earnings and the average hourly earnings series - then 2001 looks a lot worse, and the most recent recession doesn't look so bad:

While there was less in the way of job losses in 2001-02, real wages fell sharply. In contrast, the sharp fall in employment in 2008-09 didn't seem to slow down the trend in real wage growth we've been seeing since 2003. More people lost their jobs this time around, but those who kept their jobs saw real wage gains.

To what extent do those effects cancel out? Not entirely: here is a plot of the y/y growth rates of real total wages and salaries:

The job losses of the 2008-09 recession were strong enough to outweigh the gains being made by those who remained working. And workers' fall in buying power was worse than in 2001-02, but by a margin much less than what the employment numbers would suggest.

So why did real wages fall in 2001-02 and not in 2008-09? My first reaction was that since the 2008-09 recession was more severe and since the Bank of Canada had reached the lower bound, disinflation boosted real wages. In 2001-02, inflation stayed higher, this keeping real wages down.

But that's not what happened:

In 2008-09, the growth rates of prices and of nominal wages slowed, but nominal wage growth generally stayed just above that of prices. Bit in 2001-02, nominal wages fell sharply - more sharply than in 2008-09 - while inflation stayed near the Bank's target. So the fall in real wages was generated by a fall in nominal wages, not an increase in prices.

Several questions suggest themselves:

1. Why did nominal wages fall so fast in 2001-02?
2. Would the recent recession have been less severe if we had been able to engineer a similar fall in nominal wages?
3. Why didn't it happen?
4. Isn't it supposed to be hard to reduce nominal wages?

And probably many more. Answers - and answers to even more interesting questions - are welcome in the comments.

Comments

[Nick Rowe]

Interesting. Part of the puzzle may be gas/oil/commodity prices. Comparing total CPI to core CPI, we see that total exceeded core through most of 1999, 2000, and 2001. While total was less than core through most of 2009 and 2010. (And core stayed closer to the 2% target than total in both episodes.)

http://www.bankofcanada.ca/en/cpi.html

1. Is there maybe a composition effect in nominal wage changes? Hi-tech workers had higher wages, and/or higher wage inflation, so the tech wreck changed the composition?

2. Ah! the big macro question. Do wage cuts stimulate AD? If the BoC has room to cut interest rates, then of course they do. Disinflation leads the BoC to loosen monetary policy to keep inflation at 2%, so AD increases. But at the ZLB, the answer is less obvious.

3. Dunno.
4. Yes.

[Mazi Kazemi]

I would be inclined to say that the nature of the 2001 recession was decidedly different than that of our most recent downturn. Namely, when it was discovered that much of the tech industry was built on air and that many of these "generate fans, then generate revenue" business strategies did not hold up in the long run, the entire industry was reexamined. For those who lost their jobs, they may have experience structural unemployment, as many of those tech sector jobs vanished, since I would guess the industry shrunk (at least for a little while). For those who remained employed, their nominal wages should have dropped. Demand for the companies' products (whatever those were) fell, and perhaps it was realized that these workers didn't have as a high a marginal product of labor (since many of them weren't adding anything at all!!!). That's just a thought.

[Paul Friesen]

Could the fall in nominal wages after the tech bubble have been due to lots of high-paid tech workers losing their jobs? The losses could have been made up by other sectors, so you don't see the effect on over-all employment.

[Wendy]

I was in the dot com sector in 2000-2001 and noticed exactly what Mazi is saying. Programmers and other internet specialists were in short supply in 2000 and so wages and benefits rose rapidly. Meanwhile, tens of thousands were hitting the 1 year ITI diploma programs in web programming. When the crash happened, there was an over supply of talent. My co-workers who were programmers saw their wages decline by about 25%-40% when they found new jobs in 2001-2002. What I'm not sure about is whether the dot com crash refugees were a large enough group to create the effect Stephen's graph shows.

[K]

Does employee stock option income show up as "wages"?

[Determinant]

Many tech sector companies were very new, speculative and "entrepreneurial". Workers in these firms took on a significant market risk when they signed on that the business model might not work out. In this sort environment, wages are more adjustable. Stock options are the poster-child for this result.

Conversely, wages and total compensation at more established companies may in fact be stickier. More financial return history provides more information for workers to bargain for a stickier wage structure.

Speaking of hot air, I had a summer job in 2004 at an aviation company. A good number of the engineers had formerly worked for Nortel and had been laid off during its long death-throes. They commuted three hours from Ottawa to where the company was. I lived in town and was told after I was hired that I got the job partly because I was local. I could work extra time to get a project done, go home, get a full night's sleep and come back bright and chipper the next morning. Their expectation was in fact justified because that situation actually happened.

[Wonks Anonymous]

Scott Sumner would not be surprised by the negative correlation of real wages and unemployment.

[Nick Rowe]

Wonks: I think you mean positive?

[RSJ]

Yes, if an employee is granted a stock option, the difference between the exercise price and the market price is considered considered compensation and shows up as wages. I was once granted options for $0.25 a share and exercised for $40/share (ahh, good times).

But once the option is exercised, any subsequent dividends or capital gains from sale is considered capital income, but almost everyone sells the stock as soon as the option is exercised.

I'm not sure if there would be enough of that to account for the swing, but it would account for some of the swing. The 2000 stock market crash was much bigger than the 2007 crash, in terms of dollars of stock wealth erased.

One way to detect this would be to look at something like the Employer Cost Index instead of wages, which is probably what should be looked at anyways.

Does Canada have an ECI? If so, they probably charge access for it :P

[reason]

I KNOW. IT is the answer. The combination of the internet boom and Y2K (and in Europe the introduction of the Euro) meant that IT salaries skyrocketed - often through short term contracts and bonuses. Then it went back to normal. I work in the branch. For a couple of years I earnt twice what I earned before and after.

[reason]

I see other people know this also.

[Simon Howey]

Interesting points. It would be really interesting if there was sufficient variation between countries in compensation packages (% based on fixed income and % based on things that can vary from year to year such as bonuses) to try and see if it has an effect on employment during recessions. It also raises the question if maybe we should be encouraging bonuses/stock options/ and other non-traditional compensation schemes that can be easily reduced by employers as a macro stabilization tool?

[Charles Peterson]

Back in the tech boom era, options were a big part of pay packages offered by dotcom startups. The goal of many startups was to get to the buyout phase, then everyone in the company would immediately become rich. So, you'd suck up subsistence living for awhile (I was there...) in the hope of getting rich. Unfortunately, stock options immediately count as "income" you might have to pay taxes on, and in many cases they did turn out to be worthless before they could ever be struck. The value of the stock price might have been determined by a few angels buying a percent, for capital that was almost immediately vaporized, so there was nothing actually there except hot air.

Funny that the final deep plunge occurred just after January 2001.