ARE FIXED EFFECTS FIXED? Persistence in Plant Level Productivity
May 1996
Working Paper Number:
CES-96-03
Abstract
Estimates of production functions suffer from an omitted variable problem; plant quality is an omitted variable that is likely to be correlated with variable inputs. One approach is to capture differences in plant qualities through plant specific intercepts, i.e., to estimate a fixed effects model. For this technique to work, it is necessary that differences in plant quality are more or less fixed; if the "fixed effects" erode over time, such a procedure becomes problematic, especially when working with long panels. In this paper, a standard fixed effects model, extended to allow for serial correlation in the error term, is applied to a 16-year panel of textile plants. This parametric approach strongly accepts the hypothesis of fixed effects. They account for about one-third of the variation in productivity. A simple non-parametric approach, however, concludes that differences in plant qualities erode over time, that is plant qualities f-mix. Monte Carlo results demonstrate that this discrepancy comes from the parametric approach imposing an overly restrictive functional form on the data; if there were fixed effects of the magnitude measured, one would reject the hypothesis of f-mixing. For textiles, at least, the functional form of a fixed effects model appears to generate misleading conclusions. A more flexible functional form is estimated. The "fixed" effects actually have a half life of approximately 10 to 20 years, and they account for about one-half the variation in productivity.
Document Tags and Keywords
Keywords
Keywords are automatically generated using KeyBERT, a powerful and innovative
keyword extraction tool that utilizes BERT embeddings to ensure high-quality and contextually relevant
keywords.
By analyzing the content of working papers, KeyBERT identifies terms and phrases that capture the essence of the
text, highlighting the most significant topics and trends. This approach not only enhances searchability but
provides connections that go beyond potentially domain-specific author-defined keywords.
:
estimation,
production,
economist,
econometric,
macroeconomic,
estimating,
estimates production,
labor productivity,
industry productivity,
productivity differences,
productivity measures,
measures productivity,
factor productivity,
recession,
regression,
factory,
estimates productivity,
econometrician,
plant productivity,
productivity plants,
regressing,
plants industries,
productivity dispersion,
textile,
productivity dynamics
Tags
Tags are automatically generated using a pretrained language model from spaCy, which excels at
several tasks, including entity tagging.
The model is able to label words and phrases by part-of-speech,
including "organizations." By filtering for frequent words and phrases labeled as "organizations", papers are
identified to contain references to specific institutions, datasets, and other organizations.
:
Census of Manufactures,
Annual Survey of Manufactures,
Standard Industrial Classification,
Longitudinal Research Database,
Center for Economic Studies,
Ordinary Least Squares,
Columbia University,
Cobb-Douglas,
Harvard University,
New England County Metropolitan,
Review of Economics and Statistics,
Journal of Political Economy,
American Economic Review,
MIT Press,
Princeton University Press,
Princeton University,
Cambridge University Press
Similar Working Papers
Similarity between working papers are determined by an unsupervised neural
network model
know as Doc2Vec.
Doc2Vec is a model that represents entire documents as fixed-length vectors, allowing for the
capture of semantic meaning in a way that relates to the context of words within the document. The model learns to
associate a unique vector with each document while simultaneously learning word vectors, enabling tasks such as
document classification, clustering, and similarity detection by preserving the order and structure of words. The
document vectors are compared using cosine similarity/distance to determine the most similar working papers.
Papers identified with 🔥 are in the top 20% of similarity.
The 10 most similar working papers to the working paper 'ARE FIXED EFFECTS FIXED? Persistence in Plant Level Productivity' are listed below in order of similarity.
-
Working PaperWhittling Away At Productivity Dispersion🔥
March 1995
Working Paper Number:
CES-95-05
In any time period, in any industry, plant productivity levels differ widely and this dispersion is persistent. This paper explores the sources of this dispersion and their relative magnitudes in the textile industry. Plants that are measured as being more productive but pay higher wages are not necessarily more profitable; wage dispersion can account for approximately 15 percent of productivity dispersion. A plant that is highly productive today may not be as productive tomorrow. I develop a new method for measuring ex-ante dispersion and the percentage of dispersion "explained" by mean reversion. Mean reversion accounts for as much as one half the observed productivity dispersion. A portion of the dispersion, however, appears to reflect real quality differences between plants; plants that are measured as being more productive expand faster and are less likely to exit.View Full Paper PDF
-
Working PaperTechnology Locks, Creative Destruction And Non-Convergence In Productivity Levels
April 1995
Working Paper Number:
CES-95-06
This paper presents a simple solution to a new model that seeks to explain the distribution of plants across productivity levels within an industry, and empirically confirms some key predictions using the U.S. textile industry. In the model, plants are locked into a given productivity level, until they exit or retool. Convex costs of adjustment captures the fact that more productive plants expand faster. Provided there is technical change, productivity levels do not converge; the model achieves persistent dispersion in productivity levels within the context of a distortion free competitive equilibrium. The equilibrium, however, is rather turbulent; plants continually come on line with the cutting edge technology, gradually expand and finally exit or retool when they cease to recover their variable costs. The more productive plants create jobs, while the less productive destroy them. The model establishes a close link between productivity growth and dispersion in productivity levels; more rapid productivity growth leads to more widespread dispersion. This prediction is empirically confirmed. Additionally, the model provides an explanation for S-shaped diffusion.View Full Paper PDF
-
Working PaperProductivity Dispersion and Plant Selection in the Ready-Mix Concrete Industry
September 2011
Working Paper Number:
CES-11-25
This paper presents a quantitative model of productivity dispersion to explain why inefficient producers are slowly selected out of the ready-mix concrete industry. Measured productivity dispersion between the 10th and 90th percentile falls from a 4 to 1 difference using OLS, to a 2 to 1 difference using a control function. Due to volatile productivity and high sunk entry costs, a dynamic oligopoly model shows that to rationalize small gaps in exit rates between high and low productivity plants, a plant in the top quintile must produce 1.5 times more than a plant in the bottom quintile.View Full Paper PDF
-
Working PaperWhittling Away At Productivity Dispersion Futher Notes: Persistent Dispersion or Measurement Error?
November 1996
Working Paper Number:
CES-96-11
This note considers several hypotheses regarding measurement error as a source of observed cross-sectional dispersion in plant-level productivity in the US textile industry. The hypotheses that reporting error and/or price rigidity in either materials and/or output account for a substantial portion of the observed dispersion in productivity are consistent with the data. Similarly, the hypothesis that transitory product niches or fashion effects lead to differential markups and consequently dispersion in observed productivity is consistent with the data. The hypothesis that transfer pricing problems lead to persistent differences in plant-level productivity, in contrast, does not appear to be consistent with the data. Finally, the hypothesis that some plants have permanent product niches that lead to dispersion in observed productivity does not appear to be consistent with data. In order to avoid imposing a strong functional form on the data, this note follows a non-parametric methodology developed in the early paper.View Full Paper PDF
-
Working PaperEstimating Capital Efficiency Schedules Within Production Functions
May 1992
Working Paper Number:
CES-92-04
The appropriate method for aggregating capital goods across vintages to produce a single capital stock measure has long been a contentious issue, and the literature covering this topic is quite extensive. This paper presents a methodology that estimates efficiency schedules within a production function, allowing the data to reveal how the efficiency of capital goods evolve as they age. Specifically we insert a parameterized investment stream into the position of a capital variable in a production function, and then estimate the parameters of the production function simultaneously with the parameters of the investment stream. Plant level panel data for a select group of steel plants employing a common technology are used to estimate the model. Our primary finding is that when using a simple Cobb Douglas production function, the estimated efficiency schedules appear to follow a geometric pattern, which is consistent with the estimates of economic depreciation of Hulten and Wykoff (1981). Results from more flexible functional forms produced much less precise and unreliable estimates.View Full Paper PDF
-
Working PaperPlant Turnover and Demand Fluctuations in the Ready-Mix Concrete Industry
March 2006
Working Paper Number:
CES-06-08
Fluctuations in demand cause some plants to exit a market and other to enter. Would eliminating these 'uctuations reduce plant turnover? A structural model of entry and exit in concentrated markets is estimated for the ready-mix concrete industry, using plant level data from the U.S. Census. The Nested Pseudo-Likelihood algorithm is used to 'nd parameters which rationalize behavior of 'rms involved in repeated competition. Due to high sunk costs, turnover rates would only be reduced by 3% by eliminating demand 'uctuations at the county level, saving around 20 million dollars a year in scrapped capital. However, demand 'uctuations blunt 'rms'incentive to invest, reducing the number of large plants by more than 50%.View Full Paper PDF
-
Working PaperProductivity Races I: Are Some Productivuty Measures Better Than Others?
January 1997
Working Paper Number:
CES-97-02
In this study we construct twelve different measures of productivity at the plant level and test which measures of productivity are most closely associated with direct measures of economic performance. We first examine how closely correlated these measures are with various measures of profits. We then evaluate the extent to which each productivity measure is associated with lower rates of plant closure and faster plant growth (growth in employment, output, and capital). All measures of productivity considered are credible in the sense that highly productive plants, regardless of measure, are clearly more profitable, less likely to close, and grow faster. Nevertheless, labor productivity and measures of total factor productivity that are based on regression estimates of production functions are better predictors of plant growth and survival than factor share-based measures of total factor productivity (TFP). Measures of productivity that are based on several years of data appear to outperform measures of productivity that are based solely on data from the most recent year.View Full Paper PDF
-
Working PaperThe Demand for Human Capital: A Microeconomic Approach
December 2001
Working Paper Number:
CES-01-16
We propose a model for explaining the demand for human capital based on a CES production function with human capital as an explicit argument in the function. The resulting factor demand model is tested with data on roughly 6,000 plants from the Census Bureau's Longitudinal Research Database. The results show strong complementarity between physical and human capital. Moreover, the complementarity is greater in high than in low technology industries. The results also show that physical capital of more recent vintage is associated with a higher demand for human capital. While the age of a plant as a reflection of learning-by-doing is positively related to the accumulation of human capital, this relation is more pronounced in low technology industries.View Full Paper PDF
-
Working PaperThe Structure Of Production Technology Productivity And Aggregation Effects
August 1991
Working Paper Number:
CES-91-05
This is a sequel to an earlier paper by the author, Dhrymes (1990). Using the LRD sample, that paper examined the adequacy of the functional form specifications commonly employed in the literature of US Manufacturing production relations. The "universe" of the investigation was the three digit product group; the basic unit of observation was the plant; the sample consisted of all "large" plants, defined by the criterion that they employ 250 or more workers. The study encompassed three digit product groups in industries 35, 36 and 38, over the period 1972-1986, and reached one major conclusion: if one were to judge the adequacy of a given specification by the parametric compatibility of the estimates of the same parameters, as derived from the various implications of each specification, then the three most popular (production function) specifications, Cobb-Douglas, CES and Translog all fell very wide of the mark. The current paper focuses the investigation on two digit industries (but retains the plant as the basic unit of observation), i.e., our sample consists of all "large" manufacturing plants, in each of Industry 35, 36 and 38, over the period 1972-1986. It first replicates the approach of the earlier paper; the results are basically of the same genre, and for that reason are not reported herein. Second, it examines the extent to which increasing returns to scale characterize production at the two digit level; it is established that returns to scale at the mean, in the case of the translog production function are almost identical to those obtained with the Cobb-Douglas function.1 Finally, it examines the robustness and characteristics of measures of productivity, obtained in the context of an econometric formulation and those obtained by the method of what may be thought of as the "Solow Residual" and generally designated as Total Factor Productivity (TFP). The major finding here is that while there are some differences in productivity behavior as established by these two procedures, by far more important is the aggregation sensitivity of productivity measures. Thus, in the context of a pooled sample, introduction of time effects (generally thought to refer to productivity shifts) are of very marginal consequence. On the other hand, the introduction of four digit industry effects is of appreciable consequence, and this phenomenon is universal, i.e., it is present in industry 35, 36 as well as 38. The suggestion that aggregate productivity behavior may be largely, or partly, an aggregation phenomenon is certainly not a part of the established literature. Another persistent phenomenon uncovered is the extent to which productivity measures for individual plants are volatile, while two digit aggregate measures appear to be stable. These findings clearly calls for further investigation.View Full Paper PDF
-
Working PaperThe Survival of Industrial Plants
October 2002
Working Paper Number:
CES-02-25
The study seeks to explain the attrition rate of new manufacturing plants in the United States in terms of three vectors of variables. The first explains how survival of the fittest proceeds through learning by firms (plants) about their own relative efficiency. The second explains how efficiency systematically changes over time and what augments or diminishes it. The third captures the opportunity cost of resources employed in a plant. The model is tested using maximum-likelihood probit analysis with very large samples for successive census years in the 1967-97 period. One sample consists of an unbalanced panel of about three-fourths of a million plants of single and multi-unit firms, or alternatively of about 300,000 plants if only the most reliable data are considered. The second is restricted to the plants of multi-unit firms in the same time span and consists of an unbalanced panel of more than 100,000 plants. The empirical analysis strongly confirms the predictions of the model.View Full Paper PDF
