"One Market Under God, and Heaven Help Us All"

[kelley]

At 01:38 PM 7/28/00 -0400, you wrote:
>kelley wrote:
>
>>haven't read this stuff since i did some benchmarking community research 
>>on education and job training about 5 years ago.  but, as i recall, the 
>>issue is that the wages of those with a high school degree have dropped 
>>dramatically because of deunionization.  the college premium is relative 
>>to what you would earn without it, with only a highschool or less than 
>>highschool degree.  the college premium looks greater but it is being 
>>compared to wages of non college degreed that have been steadily declining.
>>
>>same deal with women's wage gains:  much of the gain is explained by the 
>>fact that men's wages decreased (also attributed to the decline in 
>>unionized jobs that paid fairly well)
>>
>>in other words, it's not like women are getting paid more per se (tho 
>>disagregating the data tells a better story than i'm doing here).  it's 
>>that men's wages have decreased over all.
>
>Not exactly. The overall gender gap did narrow, in part because men's real 
>wages fell, but women's wages also rose. Ditto the education premium - 
>non-college folks took a hit, but folks with college and more also saw 
>wage increases. Here are some figures, computed from spreadsheets on the 
>EPI site (and what's with this 1997 ending date, Max? it's 2000!):

that would be why the keyword above is "much of the gain can be 
explained".  you know, those lovely regression equations in which the goal 
is to figure out how much a particular variable might be 
quasi-causal.  which is to say:  social life is complex and the statazoids 
like to be able to assign a number that says, "x amount of the variation in 
Y can be attributed to d; x amount can be attributed to e, etc"

kelley

(it is preordained that i will now have a nightmare about being locked in 
the 'fuser lab at school with a menacing dos prompt blinking at me, getting 
angrier and angrier with my incompetence.)