## Understanding the CBO Minimum Wage Report: Why a Knowledge of Basic Statistics Matters

Image by Bettmann/CORBIS

The Congressional Budget Office released a report (pdf) arguing there would be minor job losses if a \$10.10 per hour minimum wage were passed. I’m not taking it too seriously: CBO estimates are usually pretty shoddy*. Kenneth Thomas and Mike Konczal provide detailed smackdowns (and Heidi Moore some critical context–spend some goddamn money), but what bothers is this bit from the report (boldface mine):

Once fully implemented in the second half of 2016, the \$10.10 option would reduce total employment by about 500,000 workers, or 0.3 percent, CBO projects. As with any such estimates, however, the actual losses could be smaller or larger; in CBO’s assessment, there is about a two-thirds chance that the effect would be in the range between a very slight reduction in employment and a reduction in employment of 1.0 million workers.

I’m guessing the range isn’t “about a two-thirds chance” but a sixty-eight percent chance (FORESHADOWING!). For those whose statistics are rusty, what the CBO is reporting is the mean (-500,000 jobs) and the standard error (± 500,000). Assuming a normal distribution, 68 percent of the time (if we had infinite U.S. economies to experiment with) somewhere between 0 and 1 million jobs will be lost. However, there’s a fourteen percent chance that 1 to 1.5 million jobs will be lost. Zoiks! But here’s the flip side of that coin: there is a fourteen percent chance that zero to 500,000 jobs will be created.

Oddly enough, that’s not in the CBO report. Go figure (pun intended).

This ends today’s episode of Why Everyone Should Learn Basic Statistics. Because this stuff has real world consequences. It is an immoral state of affairs:

If you will judge anything here in this struggle, you’re commanding that this city will respect the dignity of labor. So often we overlook the worth and significance of those who are not in professional jobs, or those who are not in the so-called big jobs. But let me say to you tonight, that whenever you are engaged in work that serves humanity, and is for the building of humanity, it has dignity, and it has worth. One day our society must come to see this….

…you are reminding the nation that it is a crime for people to live in this rich nation and receive starvation wages….

Do you know that most of the poor people in our country are working everyday? They are making wages so low that they can not begin to function in the mainstream of the economic life of our nation. These are facts which must be seen. And it is criminal to have people working on a full-time basis and a full-time job getting part-time income.

…But it seems that I can hear the God of the universe saying, “even though you’ve done all of that, I was hungry and you fed me not. I was naked and ye clothed me not. The children of my sons and daughters were in need of economic security, and you didn’t provide for them. So you cannot enter the kingdom of greatness.”

*I’ve never understood the impulse on parts of the center-left to defend the CBO even when it’s estimates are laughably wrong, simply because Republicans attack the CBO (usually for admittedly idiotic reasons).

This entry was posted in Economics, Statistics. Bookmark the permalink.

### 5 Responses to Understanding the CBO Minimum Wage Report: Why a Knowledge of Basic Statistics Matters

1. Robert L Bell says:

97.65 +/- 213.45 percent of reports do not understand confidence intervals.

Markov chains? Meh.

2. Robert L Bell says:

70 +/- 30 percent of reporters know how to spell “reporter.”

Stupid fat fingers, stupid skinny phone.