I'm too sick to do "real" work (except the stuff that _really_ needs to be done) and I'm stuck in the office because my wife has my car, so here's the continuing saga of the NAIRU/MERU... In response to my question of whether or not we can "say _definitely_ that unemployment in the US is lower now than it was in 1933," Tom Walker writes: >>I agree with much of what Jim says. And I'm all for having "some idea rather than no idea at all." But, having some idea is not the same as being able to say _definitely_ that unemployment in the US is lower now than it was in 1933. My own "definite" sense that unemployment was higher in 1933 comes from the mass of anecdotal evidence, not from comparison of the U rates (and I'll bet Jim's does, too).<< It's not _just_ anecdotal: looking at the social indicators like food consumption per capita, these fell as U rose. (BTW, this indicates the foolishness and perhaps mendacity of people like Michael Darby or Robert Lucas, who aver that the 1930s unemploy- ment was frictional, voluntary.) >>I frankly wouldn't know where to begin to compare the differences in data collection methods, definitions of unemployment, level of participation in market vs. subsistence economy etc. No, I have to correct myself, I *would* know where to begin -- by listing all of the substantive social-historical differences I could find and then trying to find anecdotal evidence that might allow me to interpret the data in such a way that I could make a reasonably confident comparison. At the end of such a process, I might well want to present the results in a table comparing the (now highly qualified and possibly 'adjusted') "rates" of unemployment. For me, that would be more of a rhetorical practice (presenting information in a way that might be intelligible to my audience) than a scientific one. The scientific practice would involve making the distinctions between methods of data collection, etc., etc.<< I don't see why the manifest flaws in the measurement of the U rate means that the use of the U rate in time series (or even in the more-difficult between-country comparisons) is merely "rhetorical" (which seems like a negative term to me). Economic historians have their ways to deal with such things, e.g., as when Christina Romer does calculations of _current_ U rate using methods that fit the data-availability limitations of 1933, so as to get a better feel for historical comparisons (rather than futilely trying to measure 1933 U using the methods that are adapted to the data availability of 1997). Even then, of course it's true that we've seen a decline in the farm sector and similar alternatives to wage labor _plus_ the rise of welfare-state institutions (until recently), which provide new alternatives, usually temporary ones, to wage labor. But I would see that as modifying the _impact_ of U rather than U itself. Further, such structural changes usually take place very slowly, so that for a short period (like, say, 4 years) we can assume them constant, until someone can point up what's wrong with such an assumption. I would totally agree that the measured U rate is an inevitably imperfect measure of Marx's reserve army of labor or even of some neoclassical concept of U. But why do people like Dr. Harvey Brenner find correlations between the U rate and such key variables as the suicide rate, the crime rate, the rate of violent crimes against spouses, etc.? That (and the negative correlation between U and worker incomes) suggests that the U rate is measuring _something_. It is measuring an important component of the objective conditions that people face, the conditions that generate the anecdotal evidence that Tom refers to. >>I don't object at all to comparing "rates of unemployment" provided the numbers are embedded in a discussion of how the measurement has been arrived at and what it does and doesn't reveal.<< This is true of _all_ serious empirical work. As every historian knows, anecdotal evidence _also_ needs to be treated skeptically. >>What I object to is the comparison of rates in the abstract. And, IN ABSTRACTION, there is no comparing the 1933 rate of unemployment and the 1997 one. ... Such comparisons are no more meaningful than would be a "literary" evaluation that simply counts the number of words in a book.<< Rather than being an attack on the U rate _per se_, this seems to be an attack on abstraction in general, or rather an abstract attack on abstractions. But the point of abstractions is not the concepts themselves, but the constant back-and-forth (the dialectic, if you will) between the abstract theory and the imperfectly-known and -measured empirical world. (NB: the U rate is _definitely_ an abstraction, a number calculated with a certain theory in mind, consciously or unconsciously. It is NOT the same thing as the multidimensional relationships which we think of as "unemployment.") I totally agree when bill mitchell writes:>> it is foolish to get into a lather denying the verticality of the PC at some unemployment rate. i agree with tom that we have some work to do on what we think about or what constitutes employment and hence unemployment. this will be forced upon us by environmental concerns b/c there is not a hope in hell of everyone being able to work using today's technology to produce today's mix of goods and services without the environment collapsing. but that is another story and tom sort of waylaid the debate on the nairu by putting this in.<< Of course, we need to change the mix of goods! Luckily or rather unluckily, there's a lot of work to be done cleaning up the natural environment. >>back to vertical pcs. i think there is some positive level of un always required. it might be called the level of informational un. (sort of frictional). clearly information takes time to travel and be absorbed. changing agg. demand will not really affect this.<< Wouldn't you add structural U, where immobilities in labor markets prevent people from moving to where the jobs are or gaining the skills need by the existing vacancies? >>so then if say we are at that rate......capitalist class struggle aside....it defines full capacity (in the sense that capital is not the constraint but labour supply is) and the system cannot crank out any more physical output and thus cannot quantity adjust......then what....prices have to adjust. and once you accept that then you are on a vertical pc. the question is at what unemployment rate it occurs.<< Exactly! rather than criticizing the _concept_ of the NAIRU or MERU, we need to attack the idea that the MERU is _high_! >>i say low and i say that although there might be spikes of vertical or steep segments along the way, the govt can use AD policies to drive the economy towards that level (environmental considerations aside).<< Depending on one's time frame, this is true. With structural U, if AD rises it would be awhile before the "hysteresis effect" works to convert the structurally unemployed into the employed without bottleneck-induced cost increases that encourage inflation. >>we are not giving up anything to admit that such a regime shift - from a persistent or long memory segment of the PC where AD policy can have permanent quantity effects [the horizontal short-run P curve I referred to -- JD] to a vertical segment where it can't --- exists.<< >>if you put this into a class struggle context then you have to admit that probably the cappos won't like the un rate to be anywhere near the informational rate [or even as low as the structural- frictional rate -- JD]. and in that context you get the nairu concept becoming the description of the rate at which the cappos stop using margin push to get a higher profit share. of-course, equally it is the rate where the unions or workers generally stop using wage push to increase their real wages.<< BTW, this is Carlin & Soskice's theory of the MERU. >>it moves depending on institutions and other things economic. jim has done a fine job of describing this. << thanks. >>but it isn't the informational rate. the MERU is probably only a temporary constraint on AD expansion....<< That's why labor-market and incomes policies are needed in addition to high AD. For example, >>all quantity changes (unemployment flows) then get absorbed in or out by the public sector who doesn't produce k mart shit (environmental damage material) but sets to work undoing the wrongs of the capitalist system with its now expanded workforce. so the nairu concept can be declared irrelevant in this model if the public sector acts appropriately.<< Amen, brother! bill, you and I should do research together. On incomes policies: >>my latest work ... shows that the incomes policies can work. in australia we have just concluded 13 years of them and the end came as a ideological obsession not b/c they couldn't continue working. they do damage the workers in one way as jim notes but they unambiguously provide the economy with more room to move and create employment. yet this is in a world where the govt will not act as a buffer stock employer.<< So we agree again! This is getting like Plato's REPUBLIC, where after the first part (where there are a bunch of disagreements), it settles down into "I couldn't have said it better, Socrates!," "far out, Socrates!," and "right on, Soccy!" as more and more wine is consumed. Then we get to the name game:>>... better to invent our own terms [e.g., MERU rather than NAIRU] b/c they have different conceptual underpinnings even if they are describing at the superficial exchange relations level the same thing.... i don't think we should be frightened of using the concept NAIRU although i argued in a 1987 paper that we should follow Kuhn and Lakatos and invent our own terminology to separate us from them....<< I don't think this is a big deal, but if one has a concept that describes extremely similar behavior to one that the neoclassicals use, why not admit the connection? That is, even though Marx had a completely different theory of what's behind the phenomena of supply & demand than the classicals and neoclassicals, he still used the words "supply & demand." Right before leaving the country, Paul Phillips wrote: >> Jim's comments, to me, seemed to equate the long term to the short term. What I would really like is to see what he really meant in distinguishing between long and short term.<< I would define the "short term" analytically, as that period in which the horizontal "conflict inflation" PC is unchanged. And I _don't_ equate the short term with the long; in the latter, the amount of "conflict inflation" changes due to conditions in the labor-power markets. However, I think that for U < MERU, conflict inflation tends to rise faster than it falls when U > MERU. So one can approximate that by saying that the P curve is L-shaped in the medium term, which does conflate the long and short terms. Maybe I can go somewhere on campus to lie down... in pen-l solidarity, Jim Devine [EMAIL PROTECTED] [EMAIL PROTECTED] Econ. Dept., Loyola Marymount Univ. 7900 Loyola Blvd., Los Angeles, CA 90045-8410 USA 310/338-2948 (daytime, during workweek); FAX: 310/338-1950 "It takes a busload of faith to get by." -- Lou Reed.