In 1871, engineers in Newcastle went on strike for a nine-hour day. Toward the end of that strike, the main spokesman for the employers, Sir William Armstrong sent a letter to the editor of the Times of London in which he outlined his estimate of financial loss that would be incurred by the employers if they granted the strikers' request for a 54-hour week.
Sir William's calculation was based on the naive assumption that output would decline in proportion to the reduction in working time. But it also contained an error that doubled the estimated loss. Armstrong projected a drop in net revenues of 17%. Using comparable productivity estimates from 1880, a more reasonable estimate would have been 2%. So Armstrong's estimate was off by roughly an order of magnitude. So much for the financial accounting prowess of management. In 1997, I wrote a piece on contract costing for collective bargaining. I studied how labor unions did their cost estimates and discovered a significant instance of double counting in the unions' standard costing model. This time the error was to the unions' disadvantage. Unions flocked to my door to learn about the mistake and to seek my advice on how to remedy it. Oh wait... Come to think of it, they actually didn't. Which reminds me of an economist joke. Two economists were walking along together when one of them spotted a $20 bill lying on the sidewalk. "Look, a $20 bill!" he cried out. "Impossible," replied his colleague, "incentives matter. If there was a $20 bill on the sidewalk, somebody would have already picked it up by now." Back to my story. The double-counting problem is a persistent and elusive one in social accounting. A supplement to my accounting textbook from 1977 has a half-page reference to the problem, with regard to national income accounting, somewhere in the back of the book. It explains that the "GNP accounting system must take care that the prices of _intermediate goods_ (i.e., goods and services for further processing, manufacturing, or resale) are not included over and over among the transactions accounted for. Why? Because the value of final goods _includes_ the values of prior or component intermediate goods." So how good are GDP accountants at excluding intermediate goods? Lousy. Not just lousy, but systematically lousy. There is a way around discounting the prices of intermediate goods and it is called "externalities." If a company can get somebody else to pick up the tab for their "intermediate goods", then they magically become final consumption goods and the whole lot can be counted in GDP. Hey, presto! Here we have an idiot-proof recipe for perpetual economic growth. Promote the proliferation of "externalities" and GDP can gallop ahead year after year, lifting all boats and putting a chicken (or at least an oil-soaked pelican) in every pot! Happy days are here again. The thing to keep in mind is that the national income accounting system and the treasury department have built in incentives for goosing the GDP through the inclusion of unacknowledged intermediate goods and incentives for not noticing that the GDP is being goosed. Probably more than a few low-level analysts ended up out on the street after scribbling an alarmist memo or two about the problem. Here's what I see as the crux of the problem: both critics and promoters of economic growth see pollution and resource exhaustion as undesirable side effects of economic growth. Critics view the side-effects as inherent and irreducible. Promoters insist that the problems are remediable and even more growth is needed to fund solutions to those problems. But what if both sides have put the cart before the horse? What if it is the "economic growth" itself that is the side effect? Furthermore what if that presumed growth is actually just a accounting artifact of the improper handling of intermediate goods, disguised as externalities? That is what my encounters with double counting suggest to me. Double counting occurs on the boundaries between discrete accounting units. It also takes place at the boundaries between monetized and non-monetized values. In a dynamic economy, those boundaries are ever shifting but national income accounts have no capability of handling the most significant changes. Furthermore, there is an overt financial incentive to gerrymander those boundaries to boost economic growth. Military Keynsianism and supply-side economics are two instances that can be clearly demonstrated (in the testimony of their proponents) to be deliberate efforts at boundary shifting so that intermediate goods can be counted as final goods and thus count as growth and, in the process, enhance government revenues. -- Sandwichman _______________________________________________ Futurework mailing list [email protected] https://lists.uwaterloo.ca/mailman/listinfo/futurework
