Socialist Security #1

Max Sawicky sawicky at
Fri Jul 24 14:53:27 PDT 1998

I had written:

>It is good to know that there are standard techniques for
>imputing incomes to those above the floor, but they're not
>used, so there *is* a political issue here --

Max challenges: "Really? Where?"

Take a look at Doug's comparison of CBO and Census figures for 1992, for the top 5%. According to the Census, their share of income was 18.6%; according to the CBO, it was 25.2%, more than *one-third* greater! And then there's the Pechman study vs. the others. And so on. But if "where" means "where" is the politics in this issue, I've already answered that: "there *is* a political issue here -- the usual problem of lying with statistics by using them selectively."

That the numbers differ has no necessary bearing on whether somebody is telling fibs, or whether the CBO has the more accurate answer. One would like to know whether the definitions of income in the two studies differed. Census tends to focus on money income and eschew the sort of economic adjustments to data that CBO and others pursue.

As for the rest of Max's post, my response is that we're talking about different things. He's talking about the potential availability of data and analyses thereof, while I'm talking about how the statistics are in fact used.

Census doesn't "use" these statistics; they report them. I'd be amazed if George Will ever cracked a Census Report. From what I've seen, he is not constrained by Any resort to facts.

He does this kind of thing for a living. Undoubtedly he's got intimate knowledge of and access to all kinds of sources, methods

Not really. By EPI standards my grasp of microdata is childlike. Access is no issue. Most data is public. Exceptions obtain where Privacy is an issue, but in such cases nobody outside of a few Government agencies has access.

handbooks, technical appendixes in which footnote 27 indicates that unrealized capital gains are excluded, studies reporting 7 alternate measures of this and 13 alternative measures of that. Sure, the data are there, the analyses are there, and someone with enough time, and knowledge of the sources, and expertise in deciphering them, can find out the truth.

Which is different from lying.

And then there are the people who get their information, catch as catch can, from the George Wills and Michael Parentis.

Not quite fair to Parenti, but I assume you're speaking loosely here.

Until 1994 when the loonies took over the asylum, government Reporting of data had been mostly professional and objective, Including distribution data. Exceptions have been analyses coming Out of the Council of Economic Advisors. As I said, there has always Been Ammunition for whomever could unearth it.

Biases creep into analysis, not data reporting. There has also been some omission of data since 1994, such as distributional analysis, from places where it could have been expected, but this again differs from juking the numbers.



More information about the lbo-talk mailing list