Dan badly needs a new tool:
Vox Day is still claiming that women in the workforce are bad for the economy. I still stand by my assertations that: (A) women in the workforce give employers more choice in selecting employees and allow for better overall employee value (in the absence of government interference in selection of employees or amounts paid to them), and (B) any drops in wage rates are compensated by drops in prices and better selection due to the increased productivity of having a larger worker pool in a free market. However, just in case there was something I overlooked, I checked the empirical evidence.
I pulled up Google Gapminder (previously blogged on here) and checked some data. First I pulled up a graph of Income per Capita vs. Percent of women in the workforce. I set the date at its earliest time and “played” the graph (this let the graph move through time). There was no correlation. I did the same for economic growth vs percent of women in the workforce (hint, put the economic growth axis on a log scale). While playing it through was entertaining, it did NOT show any correlation either. Finally, I did the same for fertility rates vs women in the workforce. Here, actually I was expecting to see some correlation, but I STILL didn’t find any. In fact, after playing through a number of other factors, the only correlations I could find were by inference. Women in the workforce correlate fairly well with time (the little bubbles tended to go up in general) and they correlate very well with Geographic Region (the balls clumped fairly closely together according to color).
Now, Correlation does not equal causation. However, causation DOES lead to correlation – unless the cause is so weak that it can be masked by background noise. In any case, percent of women in the workforce is NOT a significant factor an empirical look at income per capita, economic growth or even fertility rates – contrary to what Vox would have you believe.
I began to put together a comprehensive rebuttal, until it became quite clear that one was unneccessary. Dan’s tools simply aren’t up to the task and his entire argument is structured on their faulty foundation. This is what comes from relying on pretty Internet graphs instead of looking up the actual data yourself.
To give the most blatant example, Dan states that the percent of women in the American workforce is not a significant factor in fertility rates. However, the statistics clearly show otherwise.
Percentage women in the workforce 1960: 37.7 percent
Percentage women in the workforce 2000: 60.2 percent
Delta: 61.8 percent increase
Births per 1000 women 1960: 118.0
Births per 1000 women 2000: 66.3
Delta: 56.2 percent decrease
Far from being background noise, that’s a tremendous correlation and further investigation will reveal the obvious causation, as no one will be surprised to learn that working women are less likely to marry, less likely to have children if they do marry and have fewer children if they do have children.
As for the theoretical meanderings which these errant facts are supposed to support, the less said the better. Spurious and shallow are two of the first adjectives that spring to mind. And again, my original statement was that women working are bad for wage rates, which creates a host of other social ills. It is bad for the economy in many ways, beneficial in others and I have not worked out the net.
Back to the blackboard, Dan. You’re certainly welcome to try again, but I suggest doing your research next time.