Showing papers by "National Bureau of Economic Research published in 1977"

Robust regression using iteratively reweighted least-squares

Abstract: The rapid development of the theory of robust estimation (Huber, 1973) has created a need for computational procedures to produce robust estimates. We will review a number of different computational approaches for robust linear regression but focus on one—iteratively reweighted least-squares (IRLS). The weight functions that we discuss are a part of a semi-portable subroutine library called ROSEPACK (RObust Statistical Estimation PACKage) that has been developed by the authors and Virginia Klema at the Computer Research Center of the National Bureau of Economic Research, Inc. in Cambridge, Mass. with the support of the National Science Foundation. This library (Klema, 1976) makes it relatively simple to implement an IRLS regression package.

Family Migration Decisions

Abstract: This paper joins a few very recent attempts to analyze migration in the awareness of the family context. In contrast to most of them, my focus is exclusively on the family context. The paper defines family ties relevant to migration decisions and explains their effects on the probability of migration, on consequent changes in employment and earnings of family members, as well as on family integrity itself. Hopefully, the paper provides material for a missing chapter on family economics as well as an addition to the economics of labor supply arid of human capital formation.

An Economic Theory of Self-Control

Abstract: Although many economists, most notably Strotz, have discussed dynamic inconsistency and precommitment, none have dealt directly with the essence of the problem: self-control. This paper attempts to fill that gap by modeling man as an organization. The Strotz model is recast to include the control features missing in his formulation. The organizational analogy permits us to draw on the theory of agency. We thus relate the individual's control problems with those that exist in agency relationships.

Social Security and Retirement Decisions

Abstract: One of the most striking features of the postwar U.S. economy has been the rapid decrease in the labor force participation of the elderly at a time when the health of this group has been improving. In spite of this, previous research, based on retrospective interviews with the retired population, usually concludes that poor health accounts for the overwhelming majority of retirements. This paper suggests that nothing could be further from the truth. Using data from the Panel Study of Income Dynamics, we follow a cohort of white married males through their sixties to estimate a model of retirement behavior. Using several definitions of retirement suggested in the literature, we find that the two key policy parameters of the social security system—the income guarantee and the implicit tax on earnings—exert an enormous influence on retirement decisions. For example, our results suggest that a decrease in the implicit tax rate on earnings from one-half to one-third would reduce the annual probability of retirement by about fifty percent! Applying the coefficient estimates to time series data on the labor force participation of the elderly implies that the social security system has been the major factor in the explosion in earlier retirement.

Price Behavior in the Light of Balance of Payments Theories

Abstract: The purpose of this paper is to describe the behavior of that subset of prices and price indexes that is relevant to the theory of balance of payments adjustment. The theoretical writings on the balance of payments may be viewed at this juncture as falling into two main groups -- the "standard" theories and the more recent monetary theories. Each of these is examined to determine the assumptions and predictions made about particular kinds of prices, and the empirical evidence regarding these prices is then set out. Although some assessment of the theories -- solely from the price aspect -- is offered, the emphasis is on the price structure and price behavior that ought to be captured in a satisfactory theory of the mechanisms of international adjustment. For pragmatic reasons, attention is placed mainly on the theory relating to exchange rate changes rather than on the explanation of adjustment with fixed exchange rates.

Linear Regression Diagnostics

Abstract: This paper attempts to provide the user of linear multiple regression with a battery of diagnostic tools to determine which, if any, data points have high leverage or influence on the estimation process and how these possibly discrepant data points differ from the patterns set by the majority of the data The point of view taken is that when diagnostics indicate the presence of anomolous data, the choice is open as to whether these data are in fact unusual and helpful, or possibly harmful and thus in need of modifications or deletion The methodology developed depends on differences, derivatives, and decompositions of basic regression statistics There is also a discussion of how these techniques can be used with robust and ridge estimators An example is given showing the use of diagnostic methods in the estimation of a cross-country savings rate model

An Economic Study of U.S. Aircraft Hijacking, 1960-1976

Abstract: This study attempts to explain the general pattern of aircraft hijacking in the U.S. between 1361 and 1976, the reasons for the dramatic reduction in hijackings after 1972, and the costs and benefits of regulation instituted in 1973 that required mandatory preboarding searches of all passengers and carry-on luggage. The main findings of the paper can be summarized as follows: (1) Increases in the probability of apprehension, the conditional probability of incarceration and the sentence are associated with significant reductions in aircraft hijackings in the 1961 to 1976 time period. These findings are based on two methods of estimating the rate of hijackings , a quarterly time series and the time or flight intervals between successive hijackings, and alternative estimates of the deterrence variables. (2) Regression estimates from the sample period ending in 1972 were used to forecast the number of additional hijackings that would have taken place between 1973 and 1976 if (a) mandatory screening had not been instituted and (b) the probability of apprehension (once the hijacking is attempted) had remained constant and equal to its 1972 value. Under these assumptions, there would have been between 41 and 67 additional hijackings compared to the 11 that actually occurred in the 1973 to 1976 period. (3) Although the mandatory screening program is highly effective in terms of the number of hijackings prevented, its costs appear enormous. The estimated net increase in security costs due to the screening program (which does not include the time and inconvenience costs to persons searched) is $194.24 million over the 1973 to 1976 period. This, in turn, translates into a $3.24 to $9.25 million expenditure to deter a single hijacking. Put differently, if the dollar equivalent of the loss to an individual hijacked passenger were in the range of $76,718 to $219,221, then the costs of screening would just offset the expected hijacking losses.

Socioeconomic factors affecting the utilization of surgical operations.

Abstract: Between 1963 and 1970 public programs were introduced to reduce inequalities in access to medical care We examined differentials in surgical utilization among socioeconomic groups in 1970

Solving Systems of Non-Linear Equations by Broyden's Method with Projected Updates

Abstract: We introduce a modification of Broyden's method for finding a zero of n nonlinear equations in n unknowns when analytic derivatives are not available. The method retains the local Q-superlinear convergence of Broyden's method and has the additional property that if any or all of the equations are linear, it locates a zero of these equations in n+1 or fewer iterations. Limited computational experience suggests that our modification often improves upon Eroyden's method.

Middle-Age Job Mobility: Its Determinants and Consequences

Abstract: Our paper uses the wealth of information available in the NLS to expand on previous work in several ways. First, we investigate whether there is a meaningful distinction among types of job separations. Traditional analysis has categorized job separations as either employee-initiated (quits) or employer-initiated (layoffs). We question whether this dichotomy is correct. The National Longitudinal Survey data is especially useful for studying the relationship between wages and the probability of quitting. Most theoretical work on the determinants of job separation concludes that the probability of changing jobs is related to a reservation wage. The NLS data set allows us to test this relationship since it includes information on the individual's "hypothetical wage"-- that is, the wage required to induce the individual to accept another job. Given this information, we are able to compare the effects of different measures of the individual's price of time (e.g. the current wage and the reservation wage) on the probability of quitting. In addition, we analyze the role of human capital variables, job related characteristics and family background in the determination of job mobility. The analysis of the determinants of job separations in the cross-section naturally leads to an investigation of the relationship between previous separations and future separations. In particular, we consider whether such a relationship exists, and whether the nature of previous separations is a good predictor of the nature of future separations. Finally, we analyze the effects of job mobility on earnings and on job satisfaction. We distinguish between the immediate gains to mobility and the future gains to mobility, and also consider whether the nature of the separation is an important determinant of the consequences of job mobility.

The Private and Social Costs of Unemployment

Abstract: This short note emphasizes and illustrates two basic points: (1) The private costs of unemployment, i.e., the costs borne by the unemployed themselves, vary substantially and are often extremely low. This low private cost is an important cause of the permanently high unemployment rate in the United States. (2) The social costs of unemployment, i.e., the costs of unemployment to the nation as a whole regardless of how they are distributed, must be judged by considering the specific policy by which a worker would be reemployed. It is wrong to regard unemployment as either without cost (because the unemployed enjoy the opportunity for job search and leisure) or as having a cost equal to lost output. Examples are given to show that output may overstate or understate true social cost, depending on the options available for reemployment.

An Adaptive Nonlinear Least Square Algorithm

Abstract: NL2SOL is a modular program for solving the nonlinear least-squares problem that incorporates a number of novel features. It maintains a secant approximation S to the second-order part of the least-squares Hessian and adaptively decides when to use this approximation. We have found it very helpful to "size" S before updating it, something which looks much akin to Oren-Luenberger scaling. Rather than resorting to line searches or Levenberg-Marquardt modifications, we use the double-dogleg scheme of Dennis and Mei together with a special module for assessing the quality of the step thus computed. We discuss these and other ideas behind NLZSOL and briefly describe its evolution and current implementation.

An Economic Analysis of Children's Health and Intellectual Development

Abstract: The basic purpose of our research is to contribute to an understanding of the joint determination of children's cognitive development and their health. Although there is a large literature concerning the first of these issues, there has been little work on the latter. We also explore interrelationships between various aspects of children's physical health and their intellectual development and, in particular, attempt to answer the important question of whether poor health retards the cognitive development of children.

On Testing Theories of Financial Intermediary Portfolio Selection

Abstract: We generalize the stochastic specification to account for first-order vector autocorrelation in the system of portfolio demand equations and test for the validity of the "no autocorrelation" restrictions, in addition to the restricitions on the demand system implied by theory. When testing for the significance of individual parameters or for the empirical validity of more general theorectical restrictions, one must be aware of the the fact that test results may depend critically on the generality and reliability of the maintained hypothesis.

Altruism in Law and Economics

Abstract: A classic example of external benefits is the rescue of the person or property of strangers in high transaction cost settings. To illustrate, A sees a flowerpot about to fall on B's (a stranger's) head; if he shouts, B will be saved. A thus has in his power to confer a considerable benefit on B. The standard economic reaction to a situation in which there are substantial potential external benefits and high transaction costs is to propose legal intervention. In the example given, this would mean either giving A a right to a reward or punishing A if he fails to save B. Either method, we show, is costly and may result in misallocative effects. These objections to using the law to internalize the external benefits of rescue would be much less imposing were it not for altruism, a factor ignored in most discussion of externalities. Altruism may be an inexpensive substitute for costly legal methods of internalizing external benefits, though this depends on the degree of altruism, the costs of rescue, and the benefits to the rescuee. Although the general legal rule is not to reward the rescuer (nor to impose liability), the law recognizes the fragility of altruism and entitles the rescuer to a reward in certain instances. These include rewards to professional rescuers on land (normally a physician) and to rescuers at sea. In both instances the costs of rescue are likely to be sufficiently high to discourage rescue unless the rescuer anticipates compensation.

Rosetak Document 4: Rank Degeneracies and Least Square Problems

Abstract: In this paper we shall be concerned with the following problem. Let A be an m x n matrix with m being greater than or equal to n, and suppose that A is near (in a sense to be made precise later) a matrix B whose rank is less than n. Can one find a set of linearly independent columns of A that span a good approximation to the column space of B? The solution of this problem is important in a number of applications. In this paper we shall be chiefly interested in the case where the columns of A represent factors or carriers in a linear model which is to be fit to a vector of observations b. In some such applications, where the elements of A can be specified exactly (e.g. the analysis of variance), the presence of rank degeneracy in A can be dealt with by explicit mathematical formulas and causes no essential difficulties. In other applications, however, the presence of degeneracy is not at all obvious, and the failure to detect it can result in meaningless results or even the catastrophic failure of the numerical algorithms being used to solve the problem. The organization of this paper is the following. In the next section we shall give a precise definition of approximate degeneracy in terms of the singular value decomposition of A. In Section 3 we shall show that under certain conditions there is associated with A a subspace that is insensitive to how it is approximated by various choices of the columns of A, and in Section 4 we shall apply this result to the solution of the least squares problem. Sections 5, 6, and 7 will be concerned with algorithms for selecting a basis for the stable subspace from among the columns of A.

Export Prices and Exchange Rates

Abstract: The present paper is intended to make a modest contribution to an under-standing of one small but important link in this complicated chain of interacting factors. It is a link that has often been ignored because strong simplifying assumptions have until very recently usually been made about it. We refer to the relation of exchange rate changes, export prices, and domestic prices. During the last few years a number of attempts have been made to examine the extent to which exchange rate changes were "passed through"; that is, the extent to which a given depreciation in the U.S. dollar, for example, resulted in a corresponding decline in the price of U.S. exports in foreign currencies. However, the possibility that a change in the exchange rate might also alter the relationship between the export price and the domestic price of a given product, expressed in the same currency, has been almost completely ignored. The assumption made, implicitly by most past writers in the theory of international trade and more recently explicitly by advocates of the monetary approach to the balance of payments, has been that the "law of one price" applies to shipments destined for home markets and for foreign markets.

The Service Industries and U.S. Economic Growth Since World War II

Abstract: During the past 15 years employment and current dollar gross product continued to shift to the Service sector at about the same rate as in the early post-World War II period, while the Service sector's share of gross product in constant dollars remained relatively constant. Productivity (as measured in the National Income Accounts) continued to grow less rapidly than in Industry or Agriculture. The rate of growth of output per worker for the total economy was almost one percent per annum less than in 1948-65, but the shift to the Service sector contributed less than .1 percent per annum to the decrease in productivity growth. Real CDP grew almost as rapidly as in 1948-65, while employment growth accelerated due to a sharp increase in the population of working age. The expansion of service employment contributed substantially to the growth of female employment throughout the post-World War II period, but the increase in female labor force participation was not a significant factor in either the acceleration of employment or the slowdown of productivity growth in 1961-76. The growth of the Service sector also contributed to the growth of government employment. Apart from changes in industry mix, the expansion of government employment has been quite modest. Population projections to the end of this century indicate the likelihood of a marked decrease in the rate of growth of employment (and output per capita) 1990-2000 because of slow growth of working age population and the end of the transition to high female labor force participation.

An Agricultural Time Series-Cross Section Data Set

Abstract: The Agricultural Time Series-Cross Section (ATICS) dataset described in this Working Paper is based on the annual crop and livestock statistics collected by the United States Department of Agriculture. These statistics, scattered through a wide assortment of published and unpublished USDA bulletins and circulars, are extensive in their coverage of the agricultural sector, are highly disaggregated, and span a time period over one hundred years in length. Yet these rich sources have never been unified into a single compilation of data which is accessible, uniform, and machine readable. The ATICS dataset is an attempt to fill this gap.

The Gains and Losses from Industrial Concentration

Abstract: In essence, this paper will try to decompose the concentration-profits relationship into separate concentration-price arid concentration-cost relationships. By doing this, I hope to shed light on some of the allocative and distributive issues that, I suspect, give the subject its intrinsic interest, but which have not so far been confronted empirically: Does high concentration save or waste resources? Does it lead to higher prices? Who gains and loses from a social policy hostile to high concentration? Since the unique aspect of the paper is its focus on a concentration-cost relationship, most of the analytical effort is spent here. I review the theory underlying such a relationship, and develop and implement a model designed to estimate its importance. Subsequently, I try to estimate how much of the usual profit-concentration relationship is due to cost effects and how much to price effects. The main conclusion is that, while price effects are not absent, the cost effects so dominate them as to cast doubt on the efficacy of any general legal rule hostile to industrial concentration.

Industrial Demand for Energy

Abstract: In this study we examine the characteristics of industrial demand for energy, which accounts for more than one-fourth of annual energy consumption in the United States. Our research has been focused on four topics: 1. interfuel substitution in two-digit industries; 2. substitution among energy, capital and labor; 3. technical change in energy use; and 4. dynamic structure of energy demand. The results indicate that technical change has occurred through factor augmentation at unequal rates. Statistically significant labor-using and material-saving biases are found. There also appears to have been a small energy-saving bias, but it is not statistically significant.

A Model of Social Security and Retirement Decisions

Abstract: The purpose of the present paper is to focus on the potential inducement to retire earlier in the presence of social security and on the implied effects on lifetime savings. This problem is analyzed within the framework of a model of intertemporal utility maximization. The organization of this work is as follows. Section 1 introduces the topic. Section 2 presents the model of individual optimization and of the market equilibrium. Sections 3 through 5 present the comparative statistics analysis. Section 3 evaluates the effects on the equilibrium retirement age, section 4 modifies the benefits formula to depend on retirement age and section 5 examines the wealth-income ratio effect. Section 6 introduces the intergenerational transfer problem. Section 7 presents the general model underlying the previous sections.

Export and Domestic Prices Under Inflation and Exchange Rate Movements

Abstract: It is almost invariably taken for granted in theoretical descriptions of the international price mechanism and in the construction of trade models that a country's export price for a particular product is identical to its domestic price Any impact of foreign or domestic events on prices is expected to fall identically on the export and the domestic price for a good In contrast to these conventional assumptions, the few empirical studies of international prices have shown that there are fairly substantial and long-lasting divergences between export and domestic price changes for the same or closely related products If there can be divergences between export and domestic prices, a type of relative price mechanism may be at work: the depreciating country should find export prices rising relative to domestic prices of the same goods Since a producer can shift more easily from domestic to export sales of a product than from production of home goods to production of export goods we should expect the changes within commodities between domestic sales and exports to occur more rapidly Since the evidence is strong that there are divergences between export and domestic prices, we wish to trace through the effects of foreign price changes and exchange rate changes on export and domestic prices and see whether a mechanism of the hypothesized type exists In this paper we concentrate our attention on price movements, but offer some evidence that the response of exports to these price divergences is in the expected direction

Mitigating Demographic Risk Through Social Insurance

Abstract: A two-period lifetime overlapping generations growth model is used to evaluate the possibility that social insurance can effectively offset economic risks associated with uncertainty about the rate of population growth. Crude measures of the seriousness of this type of risk in the current United States situation are presented. Sufficient conditions on the structure of the economy for such intergenerational risk pooling to be mutually beneficial to all members of society are derived. Although it is logically possible to satisfy them1 we argue that they are unlikely to be realized empirically in an economy similar to that of the United States. Because of this failure, some more complex types of policy options are also discussed.

Fluctuation in Equilibrium Unemployment

Abstract: Macroeconomists tend to play down the role of fluctuations in product demand in their accounts of the movements of employment and unemployment. In the reigning mode of thought, shifts in demand account only for transitory departures of the unemployment rate from its natural level. The more persistent movements of unemployment, which account for the bulk of its variance around a long-term mean, are assigned to changes in the natural rate itself. Current thinking about the natural rate puts its emphasis entirely on the supply side of the labor market. The recent decline of the U.S. unemployment rate to below 6 percent is seen as the consequence of the declining importance of young people in the labor

The Choice of Diet for Young Children and its Relation to Children's Growth

Abstract: In this paper we analyze the choice of diet for young children in low income families in the United States and its relation to the children's growth. Our most important finding is that the education and income levels in low income households are generally sufficient for the provision of adequate diets for children in the household. This conclusion is based on empirical results which show that low income parents have pushed the growth of their children through choice of diet nearly as much as possible, and which also show that mother's education and family income are insignificant determinants of the nutrient intakes of children in low income households.

The Cumulative Unanticipated Change in Interest Rates: Evidence on the Misintermediation Hypothesis

Abstract: The term structure of interest rates is carefully analyzed over the period 1947-77 in order to construct a monthly series on cumulative unanticipated changes in long-term interest rates. This series is a sort of synthetic interest rate, changes in which over several months or years represent entirely unanticipated changes in interest rates. The behavior of this series is examined over recognized business fluctuations, and it is found to be actually more reliably pro-cyclic than the raw long-term interest rate, in spite of Kessel's finding that the market tends to correctly predict the direction of change of interest rates over phases. That the series is pro-cyclic supports the hypothesis we have put forward in another paper, that business fluctuations may be caused by "misintermediation", by which we mean the traditional mis-matching of asset and liability maturities on the part of financial intermediaries.

Money And Income, Causality Detection

Abstract: In this paper we intend to survey and suggest the theoretical framework of the important aspects of causality detection with the purpose of conveying to the reader the essential features and the different forms in which inferences may be drawn from given data. Section II presents the basic theorem characterizing the causality events and suggests two feedback detection methods which, like the one suggested by Pierce and Haugh (1977), are based on correlation analysis. In Section III we survey other well-known causality detection methods and try to relate and to compare them with the methods suggested in Section II. Section IV briefly reviews the theoretical controversy of the relationship between money and income and presents some empirical evidence based on the methods discussed in this paper. Conclusions are in Section V.

Taxation and Aggregate Factor Supply: Preliminary Estimates

Abstract: This paper extends the analysis of aggregate factor supply to a model which accounts simultaneously for the consumption/saving and labor/leisure choices. A translog utility maximization model is used to derive the set of consumption and leisure demand equations; these in turn are estimated on U.S. aggregate time series data. The empirical results are striking: we estimate (quite precisely) substantial own and cross price elasticities for current and future consumption and labor supply. The implied interest elasticity of saving is approximately 0.4.The results suggest that previous studies of labor supply and/or consumption which have ignored cross-price effects are mis-specified. We also strongly reject the hypothesis that implicit social security had no effect on factor supply.

Schooling and occupational earnings

Abstract: This paper investigates the relation between schooling and earnings across and within occupations. Across occupations earnings are positively related to mean education. Within occupations the variance in schooling levels is generally substantial, but within two thirds of the occupations no relation between schooling and earnings is observed, while in the remaining third the pattern of sensitivity varies considerably. The sensitivity of earnings to education is greater for white males than white females and substantially greater for whites than blacks. When the white sample is divided into age cohorts, the degree of sensitivity of earnings to schooling is found to be greater for younger cohorts than older ones, except for the youngest cohort in 1970. In the conclusion, a structural interpretation of the distribution of earnings is proposed to account for the findings.