In fact, the model cannot even predict US unemployment today from our own historical data. If the model were on target, we would have negative unemployment by now, thanks to the steady decline in union membership.
Nor does the model explain why, in Canada, unemployment has risen even as union membership has declined.
Second, only three provinces showed “enough variation in the card check rules to allow for the reasonable estimation of any direct effects.” (p. 20) Due to this lack of variation, Ms. Layne-Farrar opted for an indirect approach to assessing the impact of EFCA on the economy. Thus, we have to interpret shadows not just on the wall of our own cave but on the walls of adjacent caves.
Third, decisions about labor are ultimately decisions about people, not econometrics. And people – like paid researchers – have motives. Canadians in 1976 had different motives for joining unions than Americans in 2009 might have. For one thing, in 1976 all Canadians had health insurance.
Thus, Ms. Layne-Farrar has performed a brilliant analysis on a fatally-flawed data set. This places her in the position of the orthopedic surgeon who has performed impeccable surgery on the wrong knee. And her conclusions, like those knees, cannot bear the weight they are now asked to carry.
Ms. Layne-Farrar is undaunted. She writes, “On the basis of these results, I conclude that a card check system which increases union membership would also lead to a considerably high unemployment rate.” (p. 23)
At this point, a reality check is in order. There are 153,000,000 jobs in the US economy and 12.1% of workers are unionized. Thus, the unionized workforce right now is about 18,513,000 members strong.
A 19% increase (based on British Columbia’s experience) in union membership would add 3,500,000 new union members and bring union density up to 14.4% of the workforce – 22,000,000 union members in total. The increase would equal 2.3% of the workforce (less than the 5% - 10% in the forecasts).
Here I make two assumptions. First, I assume that the average wage of these workers now is about $35,000 a year – 70% of the median wage countrywide. I selected that number arbitrarily, thinking that it’s likely the people who will be most interested in joining unions are people of the lower income strata now – people averaging about $17.50 an hour with limited benefits.
I further assumed that those 3.5 million people averaging $17.50 an hour would gain an average 15% wage increase (from the literature review) by joining a union. Their pay would go from $35,000 a year to about $40,250 a year.
Thus, 3.5 million people would get raises averaging $5,250 a year. If that happens, the direct salary cost nationwide would be $18.4 billion a year – substantially less than the forecasted 0.14% reduction in national output.
Coincidentally, that is the same amount Wall Street paid out in bonuses last year, and no one was making doomsday predictions for the economy before that money was paid. Indeed, something approaching doomsday had already happened, thanks to the negligence and greed of the very people who received the bonuses.
Of course, as Ms. Layne-Farrar points out, pay increases involve costs that go well beyond direct salary costs. But that is as true on Wall Street as it is on your street.
Let’s frame this reality check another way. Ms. Layne-Farrar says that if businesses are NOT able to:
-- Refuse to recognize the union without a secret vote;