>incomes for purposes of analysis. ... there is no real political
>issue here about biases in the reporting of the distribution of
>income or wealth."
>Yes, but Doug had it right. Top-coding imparts a downward bias
>to measures of inequality. Moreover, if income *gains* accrue
Only if the unvarnished data is reported. I have never seen it so. By one means or another, analysts figure out a number for the upper x percentiles and plug it in. Within an income class, the average is all that matters. We're not interested in the income of any specific rich person in this context.
I have never seen any analysis which did not plug the gap. Accepting the top-coded value would be ridiculous.
>It is good to know that there are standard techniques for
>imputing incomes to those above the floor, but they're not always
>used, so there *is* a political issue here --
>Fairly early on in the inequality debate, maybe 1990 or so,
>Joseph Pechman had an article, perhaps in the Am. Econ. Review,
>in which he used a different data set from everyone else (IRS
>instead of Census data, I think), a data set without the
>top-coding, and found that the Gini coefficients or whatever had
>increased a good deal more than previous studies indicated. This
>helped solve a puzzle at that time: people knew that inequality
>was increasing substantially, but the rise wasn't showing up in
Clearly the procedures for filling in the gaps are imperfect and open to criticism.
>But ... there is an even more important issue, I think, which is
>that unrealized capital gains are not counted as income. In the
>present bubble economy, and really since the bull market began
>about 15 years ago, that matters a good deal. Aside from
Not counted in GDP, but it is counted (estimated, again) in the better income distribution and tax studies. Where it is not, the authors usually make clear what's included and what's missing. All the government research in this vein is quite scrupulous, excepting perhaps the Joint Economic Ctee staff since 1994 (a relatively minor player).
>Besides which, the whole concept of "income" is pretty flaky. If
>the bourgeoisie were to provide company housing, feed us in
>company canteens, and provide transportation to and from work,
>all of that would be business expenditures, not income. Because
Government studies routine employ the concept of "family economic income" which includes imputations for all sorts of things, such as the rent of an owner-occupied home.
Fringes are included in the national accounts. If there were more of what you're talking about, they would find some way to count that too.
There is ample ammunition in official government reports to lambaste the bourgeoisie to your heart's content. We do not lack for information, by and large.