Thursday, December 08, 2005

Economic growth

Gene Sperling, formerly Bill Clinton's economic advisor (presiding over record economic growth), has written The Pro-Growth Progressive : An Economic Strategy for Shared Prosperity.

He lead a fascinating and free-wheeling discussion at TPMCafe's Book Club, and he has an interesting guest post at ThinkProgress. He points out that 43% of the public think we're still in a recession when it actually ended a couple years back.

Why do people think this? Why do 63% think the economy is bad, very bad, or terrible?

Here's part of Sperling's answer:
Real hourly earnings are down in the four years since the last recession—from $16.41 in November 2001 to $16.29 in October this year. [Bureau of Labor Statistics (BLS). Adjusted to October 2005 dollars]

Real weekly wages are down in the four years since the last recession—from $557.44 in November 2001 to $550.60 this October. This is the first time on record real weekly wage growth have been negative this long after a recession. [BLS]

Since the 2003 tax cut, real hourly wages have fallen 2.2% and real weekly wages have fallen 1.6%. [BLS]

Real median household income has fallen each year Bush has been in office and by nearly $1,700 since the recession 2001 [U.S. Census Bureau, Income Poverty, and Health Insurance Coverage in the United States: 2004, Aug. 2005, Table A-1.]
The end of a recession is defined in terms of sustained GDP growth. That's aggregated across all of the population, and is easily converted to an average (mean production).

One easy way to lie with statistics is to abuse the difference between mean and median. The mean is what we usually think of as the average, the total number of events (in the jargon) over the total number of things that the events happened to. So we have the total number of hits a baseball player has (each is an event) divided by the total at-bats. We could also look at the average number of hits a team produces by counting the total number of hits any player on the team got and dividing by the number of players who had an at-bat.

But anyone who watches baseball knows that most of a team's hits are produced by the really good players. Consider my boys, the Chicago Cubs. Of the 1506 hits generated by the 38 players who had an at-bat. That's an average of 39.6 hits per player.
There are 12 players with more than 40 hits (one with 39), which is about a third. Half the players have fewer than 8.5 hits, half have more, making 8.5 a statistic called the median.

For highly skewed distributions like hits per player or income per person, the median is often a more accurate representation of the state of the population than the average we're all used to.

The mean is handy when you're dealing with values that are distributed in more conventional ways. If we look at the ages of the players on the 2005 Cubs, the median is 27, the mean is 27.84211, almost identical.

I mentioned that income per person is highly skewed. That means that the mean (GDP per person) is a poor representation of the state of the economy for the average person. That's why Sperling refers to the median household income, not the average income.

When Bill Gates walks into a room, he raises the average wealth in the room, but not the median. If you make the richest third of the country richer, but the poorest half poorer, you lower the median, even of the average rises.

Good economic policy not only raises average income, it raises income across the board. There's a classic line that a recession is when your neighbor loses his house, a depression is when you lose yours. That's a median argument, not an average. And most people know someone struggling to make ends meet (possibly themselves), and that's why people think there's a recession. Looking only at averages ignores the fact that for every Derrek Lee, there are a bunch of Roberto Novoas.