Knowledge that Transforms

A DAM SMITH'S use of the manufacture of pins to illustrate the benefits to efficiency of the division of labor (1776, Book 1, chap. 1) is famous, but there are few references to the subsequent development of the pin trade in the literature. The purpose of this note is to outline the development of the pin trade since Adam smith's reference to it and to place his comments in a new perspective. Pins have changed very little during the two hundred years since Adam Smith wrote about the trade. This makes possible meaningful measurement of long-term changes in labor productivity, which are described later. First, the structure of the trade, which has changed, is outlined. In an appendix the techniques of production, which have also changed, are described.

PHE PERSISTENT presence of inflationary problems in most industrialized countries has lead to a renewed interest in the relationships between inflation and the tax system.' There exists a vast literature on the microeconomic aspects of this problem and especially on various accounting procedures used to compensate for distorting inflationary effects. These aspects of inflacounting will not be dealt with in this paper, at least not directly.2 This paper will begin with a typology of the effects of inflation on the tax system; next it will concentrate on some macroeconomic aspects of the impact of inflation on the tax system on the one hand and the impact of the tax system on inflation on the other. On a more general level this problem leads to the question of how far inflation-intentionally or unintentionally-acts as a means of taxation, i.e., as an alternative way to transfer resources from the private to the public sector of an economy.

We are indebted to many people for their encouragement and advice in undertaking this project. Above all, however, we must thank the editor of this journal, whose practical and scientifc judgment kept us on track through what turned out to be a long and at times difficult task. In addition, we must thank William Baumol, Gerald Faulhaber, Michael Intriligator, George Lee, Stephen Littlechild, Fritz Machlup, Thomas Marschak, Herve Moulin, Andrew Postlewaite, Juan Eugenio de la Rama, Robert Rosenthal, David Schmeidler, Anne Schotter, and Martin Shubik, all of whom have read the manuscript either in part or in whole and have given us extremely useful scientific and editorial advice. Professor Schotter's participation in the project was made possible by a grant from the Office of Naval Research Contract No. N00014-78-C-0598, whose support is gratefully acknowledged. All errors in the work that follows are those of the authors alone.

IN PRESENTING his estimates of la richesse territoriale de France, Antoine Lavoisier [100, 1791] pointed out that the object of such estimation is to furnish a veritable thermometer of national prosperity.' That certainly has provided the motivation behind much work on national income estimation and comparison. Lavoisier was an optimist, and he felt that there need be little room for dispute if the problem were correctly posed and pursued:

JN the United States, unemployment insurance benefit payments are financed predominantly by taxes on employers' payrolls. The base of this tax is, however, not the actual payroll but the taxable payroll. Because of the legal provisions according to which the taxable payroll is computed, precise analytical formulae for the taxable payroll can be derived only for certain postulated employment patterns. In a previous theoretical paper (Brechling, 1977a) the determination of the taxable payroll for various types of employment patterns was examined. Further, on the basis of some simplifying assumptions about the employment patterns and some approximations, a fairly general but approximate formula for the taxable payroll was derived. It incorporates all the essential features of the specific formulae. In subsequent research the validity and adequacy of this general formula were examined empirically. This paper contains some theoretical extensions of the previous work as well as the results of the empirical research. The empirical work is unconventional in that it does not consist of tests of hypotheses that arise from assumptions about the behavior of economic agents. Instead it is designed to discover whether a formula which has been derived theoretically on the basis of technical assumptions is, in fact, valid and, hence, useful for practical purposes. Three such practical purposes seem to be especially important: First, theoretical considerations (Brechling, 1977a and b) have led to the conclusion that the taxable payroll may be related positively to labor turnover. Such a relationship implies that there is a tax cost of labor turnover that is influenced by the parameters of the tax structure. Hence changes in these parameters, effected by policy makers, may well affect labor turnover in the economy. Thus it is important that the theoretical relationships between the taxable payroll and its determinants be examined empirically. Second, a correct representation of the taxable payroll is highly desirable in order to be able to model adequately the total tax inflows into the unemployment insurance funds. In many states these funds were depleted substantially in recent years so that some corrective action by policy makers has been called for. Such action is likely to be particularly effective and purposeful if it is based on an adequate model of benefit outflows and tax inflows. Third, for the assessment of the recently enacted employment tax credit, changes in employment levels are approximated by changes in the total taxable payroll of an employer. For an evaluation of the working of this tax credit, it seems essential that the taxable payroll be modelled adequately. For (at least) these three reasons, a theoretical and empirical examination of the determinants of the taxable payroll of the unemployment insurance tax is regarded as an important and relevant research task. This paper is structured as follows: In section lI the formula for the taxable payroll of a particular type of employee is presented and its main implications are highlighted. This section is based on the previous theoretical work (Brechling, 1977a and b) and, hence, is quite brief. In section III, the theoretical analysis is extended. In particular, expressions are derived for a mean taxable payroll for situations in which data for particular types of individual workers are not available. In essence, section III is devoted to dealing with an aggregation problem. In section IV the empirical relationships between taxable payrolls and their determinants are examined. The empirical results lend substantial support to the theoretical models. But some problems still remain to be solved. Received for publication March 27, 1978. Revision accepted for publication February 15, 1979. * University of Maryland. The research underlying this paper was made possible by a contract with the U.S. Department of Labor, ASPER (No. J-9-M-6-0103). Its contents do not necessarily represent the official opinion or policy -of the Department of Labor. The research was undertaken while the author was visiting the Public Research Institute at the Center for Naval Analyses in 1976/77. Kathy Classen Utgoff kindly read and commented on a previous draft of this paper. Her help is acknowledged with gratitude.

William A. Niskanen, Steve H. Hanke, Land Prices Substantially Underestimate the Value of Environmental Quality: A Rejoinder, The Review of Economics and Statistics, Vol. 62, No. 1 (Feb., 1980), p. 156