> I was really disappointed in Phillips for that
> article. I think he's a smart guy and usually
> very interesting, but that critique is populist
> crankery. The fellow who does the Shadow
> Government Stats, on whom Phillips relies
> heavily, really doesn't seem to understand how
> the employment stats are collected or reported.
> And the CPI issue is a lot more complicated
> than Phillips/Williams allows. How do you
> account for housing costs when people don't
> move every year? And when the sale of a house
> might involve a capital gain or loss? The
> statisticians say they're trying to separate
> the investment and consumption aspects of
> housing, which is in some sense the right
> thing to do. How to realize that ambition is
> pretty complicated. Also, accounting for
> quality improvements in the price indexes is
> really hard, conceptually and practically. My
>< impression, from having talked to official
> geekdom for more than 20 years, is that they're
> dedicated public servants trying to do the best
> job they can with limited resources.
Official geekdom responded directly to Shadowstats:
<http://www.econbrowser.com/archives/2008/09/shadowstats_deb.html>
At the risk of being too orthodox, Robert Solow also has some objections worth heeding:
"Even the better parts of Phillips's book fail to give a serious account of what has happened in the capital market and the channels through which it affects the modern economy. And those parts are interspersed with passages of surpassing ignorance. One chapter is largely given over to an attack on the integrity--not the accuracy but the integrity--of the consumer price index. This would be scurrilous if it were not so silly. Phillips does not seem to understand what a consumer price index is. There is a graph showing that between 2000 and 2006 house prices (which are not included in the CPI) rose much faster than 'owners equivalent rent' (which is). This is supposed to demonstrate that the Bureau of Labor Statistics was deliberately suppressing the bad news that ongoing inflation was really much faster than reported.
"But asset prices do not belong in a consumer price index. Neither do house prices or land prices or stock prices. If I buy a house today, it would be senseless to count its whole cost as part of my consumption expenditure _this_ year. In fact, at about the same time there was much talk, inside and outside the government, of 'asset-price inflation' as something distinct from 'goods-price inflation' and what should and might be done about it. The fact that house prices were rising faster than rents was quite analogous to a rise in the familiar price-earnings ratio for common stocks.
"This is not the only such category mistake in the book. Phillips criticizes National Income and Product Accounts because they do not prominently display figures on the sharply rising volume of private debt. But this lapse is owed to the fact that the National Income and Product Accounts are about, well, national income and product. They are more like an income statement, and certainly not like a balance sheet. There is no reason, of course, why the Department of Commerce should not inform interested citizens about the standing volume of debt, but there is also no reason why the absence of those figures from the accounts should be treated as a dark conspiracy against the public. The debt figures are readily available in other government documents."
<http://www.tnr.com/story_print.html?id=5adeed59-4ca0-443a-bc34-828405d7a5f9>
I still heed Phillips, but the best of his arguments do not require Shadowstats. For example, his take on the election:
<http://mailman.lbo-talk.org/pipermail/lbo-talk/Week-of-Mon-20081103/018617.html>
Shane