To: Wyätt Gwyön who wrote (66726 ) 7/23/2006 11:23:47 AM From: CalculatedRisk Respond to of 110194 Clearly Lazear is using different stats than me. Apparently he is capturing some other compensation paid to high income earners. For some analysis (PhD economist knzn), check out the last 3 posts:knzn.blogspot.com Excerpt: Why are their results so different? It turns out they are using different deflators for compensation. Lazear apparently uses the product deflator for the nonfarm business sector to produce his real compensation series. Calculated Risk uses the “real hourly compensation” series from the Bureau of Labor Statistics, which is produced using a consumption deflator. (I can’t figure out exactly what series the BLS uses to deflate compensation. It may be some version of the CPI or a consumption deflator from the national accounts. In any case, it appears to follow consumer prices.) In other words, Lazear measures compensation in terms of what people are producing (the “product wage”), whereas Calculated Risk measures compensation in terms of what people can buy with their compensation (the “consumption wage”). Why do these two deflators produce such different results? They actually produce similar results until about 1970, and then they diverge. Something new has been happening in the last 35 years. To put it in general terms, many of the things Americans produce (computers, for example) have been getting cheaper, but most of the things Americans buy (health care, for example) have been getting more expensive. Much of the productivity growth we have experienced has been in the production of investment goods. That kind of productivity growth benefits the people who (indirectly) buy those investment goods, but it doesn’t help today’s workers very much (unless they’re also investors).