Friday, October 27, 2006

Living In an Immaterial World

A myriad of obligations deny me from watching Real Time with Bill Maher when it airs on Friday night. Additional distractions have prevented me from commenting on last Friday's episode and I feel I should post before the next installment.

The episode featured Congressman Barney Frank, Jason Alexander, Stephen Moore. Bill Maher and two others spent part of the show lamenting that the average wage was going down while the DOW grew. I'll give one guess which two were on Maher's side (and Moore got a masters in economics from Mason). Moore brought up good points but it mostly consisted of pointing out Maher was rich, too. What he should've pointed out is that the data didn't matter in the first place.

Wage data can be immaterial in two basic ways and which way depends on how much time the analysis covers: did wages drop base on last year's wages or last decade's? If it's last year's wages, then it is just a tick in the market. The lastest adjustment does not allow for sweeping conclusions about the nature of the economy. (See earlier post here.) Imagine a great but developing relationship and then your date says something stupid. Do you end it based on that? Of course not; you are still getting to know one another and mistakes will be made. The economy is similarly in a constant state of adaptation. There are rigidities and there are errors. The lastest tick means little.

If the data is last decade's it avoids the minor change problem but it eases to a different, more subtle problem. As people enter the bottom of the work force (immigrants, new adults) and people leave from the top (retiries), there's downward pressure on average wages even if everyone's getting richer. The better measure is income mobility over long spans of time. (See Steve Horwitz's myths page; the third item has an excellent discussion of the topic.)

So even if we set aside other problems with real wage data (inflation's over estimated, there's no inclusion of product quality, there's no inclusion of job quality), the data doesn't matter.

No comments: