Jim Devine wrote, >But again, as I said above, >poetry in fine in and of itself. But social science, though it shares a lot >with poetry (using metaphors, etc.), is also about trying to understand the >objective world. Poetry tries to understand two worlds, the world of language and the world -- Jim's "objective world" -- to which that language refers. What I'm about to demonstrate may not look like poetry to someone who thinks poetry is "fine in and of itself." But take my word for it, the techniques I use are derived from the study of literary texts. Judge for yourself whether they have any relevance for the criticism of social scientific texts. The problem comes from a survey of the literature on "Payroll Taxation and Employment" written by an economist in the Canadian Ministry of Industry, Joni Baran. The evidence consists of several ungrammatical constructions -- bad poetry, one might say. In the paper's executive summary "the evidence concludes". In the conclusion, "most of the earlier work surveyed by Dahlby and by Hamermesh is unanimous in implying". A rose is a rose is a rose, but "evidence" can't conclude anything. Nor can "most" be said to be "unanimous", even "in implying". Something is not quite right with the words and further investigation is warranted. Guarding the non-economist's entry into the main text of the paper, however, is an eight page "Economic Primer" that "provides a brief primer of the theoretical underpinnings of the labour market and familiarizes the reader with some of the terminology used throughout the remainder of the paper." It probably -- albeit unintentionally -- insures that non-specialists will venture no further into the 45 page paper. It thus upholds the definitive status of the executive summary. Here are three clues from the paper (the paper can be downloaded from the Industry Canada website: http://strategis.ic.qc.ca/ssg/ra0127e.html): >From the executive summary: "The consensus of the surveyed literature is that increases in employer payroll taxes tend to have negative short-term effects on employment. Most of the evidence, however, concludes that adverse employment effects do not persist in the long run." (p. i) >From the conclusions: "Second, while most of the earlier work surveyed by Dahlby and by Hamermesh is unanimous in implying that any adverse effects on employment fully dissipate over the long run, some of the more recent empirical work suggests that increases in payroll taxes can have persistent long-term effects on employment. Moreover, there is evidence that real wage resistance is relatively strong in Canada (Tyrvainen, and Wilton and Prescott), thereby delaying and potentially hindering the long-run adjustment process." (p. 39) >From the literature review: Exhibit 1. "Dahlby concluded that the collective results of the theoretical predictions and the econometric studies implied that over 80 percent of the employer payroll tax burden in Ontario is borne by labour in the long run. Therefore, payroll taxes do not directly inhibit job creation at least in the long run. "Payroll taxes may, however, have an indirect effect on employment in the long run. Taxes can distort labour market decisions. Dahlby refers to the Keil and Symons model. While the direct effect of a payroll tax increase on wages has only a short-run effect on employment, the short-run increase in unemployment reduces the real wage rate and decreases the size of the labour force. This short-term reduction in the labour force can lead to the discouraged worker effect which can have long-term implications on employment. "Dahlby also referred to Coe's 1990 analysis which indicated that increases in employer payroll tax rates in Canada caused the natural rate of unemployment to increase by 1.5 percentage points from 1971 to the late 1970s and by another one percentage point by 1990. Research performed in other countries also tends to suggest that increasing payroll taxes may have a long-term impact on unemployment." (pp.18-19) Exhibit 2. "Hamermesh's survey of the empirical literature led him to deduce that the lack of consensus regarding the incidence of payroll taxes and its long-run impact on employment, makes it impossible to draw any firm conclusions. Therefore, Hamermesh maintained that economists must choose between the robust estimates of supply and demand elasticities provided by the theoretical model or the mixed conclusions of the empirical evidence." (p. 21) What the poetic analysis of the text allows the reader to do is stride past the technical mumbo-jumbo with "ten league boots" and quickly grasp the central contradiction of the author's predicament (in both senses of a "dangerous situation" and of "logical predication"). The author's predicament is dangerous indeed, she somehow must temper her criticism of government policy with the mitigating suggestion that even though payroll tax policy may be unfair to low income earners, it is -- or at least was -- consistent with the views of a consensus of economists that any adverse effects are only short lived. The problem with this "consensus" is that it runs counter to at least some of the evidence. Baran's ungrammatical sentence in the executive summary could be made grammatical with only a slight revision: "[In spite of] the evidence [to the contrary], however, [most economists] conclude that adverse employment effects do not persist in the long run." It may be objected that the same textual contradictions could be found without resort to a poetic analysis. I agree. All the poetic analysis does is act as a kind of dye marker that helps me focus immediately on the ungrammatical bits as prime suspects. Instead of proceeding linearly through the paper, my understanding of conventional textual structure leads me directly from the executive summary to the corresponding analysis in the conclusion and from there to the relevent sections in the reviews of Dahlby's and Hamermesh's work. In other words, it saves time. Ben Franklin said, "Time is money". Wallace Stevens said, "Money is a kind of poetry". So, it's only logical: If Time=Money and Money=Poetry then Poetry=Time. Regards, Tom Walker ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ knoW Ware Communications Vancouver, B.C., CANADA [EMAIL PROTECTED] (604) 688-8296 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ The TimeWork Web: http://mindlink.net/knowware/worksite.htm