Job Creation and Destruction

Citation

Booth, A.L. (1998), "Job Creation and Destruction", Personnel Review, Vol. 27 No. 2, pp. 158-161. https://doi.org/10.1108/pr.1998.27.2.158.2

This important and well‐written book gathers together the authors’ empirical work charting patterns of job creation and job destruction in the US manufacturing sector over the period 1972 to 1988. The book is based on meticulous statistical analysis of quarterly and annual plant‐level data from the Longitudinal Research Database. It is thought‐provoking but easy to follow, and will be read by a wide audience, including not only economists but also policymakers and non‐economists. The layout is accessible, and the introduction provides a “road‐map” delineating a short easy route for any readers who wish to absorb the main findings without a detailed reading of all chapters of the book. The more technical statistical material is confined to appendices. The main points of the book can be understood by simply reading the summary of results in bullet form at the start of each chapter. Mandatory reading for all policy makers is Chapter 2 (setting out the statistical framework used in the remainder of the book), the summaries given in bullet point form at the start of each chapter, plus Chapter 7 (setting out the economic and policy implications of the stylized facts about job creation and job destruction).

The stylized facts in the book are based on the application to the data of some simple definitions of employment change. Over a given time period, gross job creation represents employment gains summed over all plants expanding or starting up, while gross job destruction represents employment losses summed over all plants contracting or shutting down. Gross job reallocation is the sum of job creation and job destruction. As employment opportunities move across locations, workers also make similar shifts (but may in addition change status for reasons unrelated to job reallocation). How are worker and job reallocation related? Gross worker reallocation is defined as the number of people changing employment place or state over a given period. It therefore follows that job reallocation is the maximum amount of worker reallocation, since the job reallocation measure may count workers twice if they move from a shrinking plant (job destruction) to an expanding plant (job creation) within a given period. Minimum worker reallocation is equal to the larger of job creation or job destruction, and thus is a lower bound on the amount of worker reallocation accommodating job reallocation. Most studies to date have been based on worker flows; the approach followed in this book is to study job flows. Work on combining the two remains to be done.

Some of the most interesting facts emerging from the statistical analysis on which the book is based are as follows. In US manufacturing, plant‐level job creation and job destruction rates are large. Over a typical interval of a year, one in ten US manufacturing jobs disappear, while a comparable number of new jobs are created elsewhere. Such job creation and destruction reflects persistent plant‐level employment changes, since most jobs that are lost fail to reappear at the same location within two years. About one quarter of job destruction occurs at plants that close down. Job destruction rates are strongly countercyclical, while job creation is procyclical: job destruction rates increase sharply in recessions, but job creation rates slow down relatively less.

How do plant‐level job reallocation rates vary across observable characteristics? Annual job reallocation rates vary considerably across industries, and are lower in higher‐wage plants (perhaps because higher‐paid workers are characterized by higher levels of human capital). Job reallocation rates decline with plant age, plant size and firm size, and typically new and existing jobs with larger employers last longer. The authors argue convincingly (p. 57) that the “[c]onventional wisdom about the job‐creating prowess of small businesses rests on statistical fallacies and misleading interpretations of the data”. Policy makers should take note.

There are several vital implications of this book for academic economists. First, an argument of the book is that observable characteristics (such as industry, region, wages, employer size and age, capital and energy intensity, and foreign competition) account for little of the observed plant‐level heterogeneity in job creation and destruction rates. The authors suggest a number of factors that may account for such plant‐level heterogeneity, such as uncertainty about the development and implementation of new products and techniques, and heterogeneity in entrepreneurial and managerial ability. But they also note (p. 153) that “the current state of economic science provides little knowledge about the various factors or the precise reasons why they generate such heterogeneity in outcomes”. This heterogeneity also has implications for policy, since it is hard to target and evaluate industrial or commercial policies when plants exhibit such idiosyncrasies. However, it is clear that high levels of job destruction mean that workers need to be sufficiently flexible and well‐trained to adapt to changes in the location and requirements of different jobs; this necessitates an educational system providing workers with basic skills and problem‐solving abilities.

Second, the fact that, for US manufacturing, the job reallocation rate is countercyclical (since job destruction varies more than job creation over the business cycle) has implications for the modelling and interpretation of business cycles. It is suggested that “the prevailing view of business cycles, which is based on the notion that most or all of firms respond similarly to the cycle, is not supported by the evidence on job flows in the manufacturing sector” (p. 93). Instead, it is argued that a richer view of business cycles needs to be developed, highlighting their connection with the restructuring process.

As the authors make clear, there are a number of areas in which the statistical analysis could be extended. First, the book focuses only on US manufacturing, comprising just 19 per cent of total employment in the USA in 1988.

Second, many questions raised by the authors’ analysis might be answered using data linking job flows with workers flows, employer attributes and worker characteristics. Such data are available separately for the USA, but not as yet linked. The matching together of these data would facilitate the inclusion into the analysis of voluntary quits by workers, and would allow workers to be distinguished by, for example, part‐time, full‐time and overtime status (not done in the current book).

While the book does not attempt any formal economic modelling to explain the stylized facts of job creation and destruction, it makes a number of useful suggestions as to what the underlying economic rationale might be. In addition, it pinpoints areas in which the existing economic theory is unable to explain the stylized facts. This is of interest to a number of areas of economics, including labour economics, industrial organization, and macroeconomics. The book also discusses the policy implications of its major findings, and advances warnings about what policy makers ought to avoid in their efforts to ameliorate the social consequences of job reallocation. The book is highly recommended.