An excerpt from
Is a Volatile Economy Good for America?
Clair Brown, John Haltiwanger, and Julia Lane
Overview of the Book
It's better to be a dog in a peaceful time than be a man in a chaotic period.
The U.S. economy is both celebrated and reviled for its dynamism. New jobs are constantly created, new firms replace old, and the American economic model is the one adopted around the world. Yet its unceasing and turbulent change creates enormous angst about the loss of jobs, the loss of earnings, and the loss of competitiveness of American firms. At the same time that employment is at an all-time high, CNN reporter Lou Dobbs captures the national angst in his book, The Exporting of America.
What is the overall impact of this change on jobs, workers, and firms? Every week, in every part of the economy and in every corner of the country, some firms shut down and others start up, some jobs are created and others are destroyed, some workers are hired and others quit or are laid off. Giant Food, a fixture in the Washington, D.C., economy, is one example. It shut down its Maryland headquarters in May 2005 and laid off five hundred workers. The local news was full of stories about the demise of good middle-class jobs and how the local community would be hurt. But almost unnoticed in the very same week was a small report that MOM (My Organic Market) was creating fifty new jobs by opening a new store in western Maryland. Of course, the nature of the news industry is to report on visible and traumatic events, which tend to be job losses, worker layoffs, and plant closings. Yet maybe small startups like MOM will turn into many more jobs, although added slowly and over time. And maybe workers laid off at Giant will end up with better jobs. It’s also possible that MOM will fail and that workers at Giant will never land the kind of jobs that they had before.
Finding out what happens to jobs, workers, and firms—like Giant workers and workers at MOM’s new store—is what this book is about. While stories point out the successes and traumas for individual businesses and in individual lives, decisions need to be based on facts. This book does just that. It uses new information to go beyond anecdotes and establish some facts about economic turbulence and its impact on things that people, and their politicians, worry about: firm survival, worker job ladders and career paths, and the future of middle-class incomes.
The book focuses on five industries that are familiar to all Americans: semiconductors, software, financial services, retail food, and trucking. These industries have been affected by the same sets of forces that have affected all industries in the economy, but to different degrees. The semiconductor industry has experienced rapid technological change combined with restructuring caused by the rise of foundries and startup fabless companies—semiconductor companies that outsource the actual manufacturing—combined with the proliferation of product markets. In software, small startup firms also have played a critical role in the explosion of new products and applications, and these startups are closely linked to dominant firms like Microsoft. In financial services, the range of changes has also been staggering. Regulatory restructuring increased competition both within and between sub-industries and led to industry consolidation during a period of.massive technological change, including the introduction of the ATM, widespread use of call centers, and the introduction of online services. In food stores, changes in market structure have been enormous, with power retailers like Wal-Mart playing an increasingly large role. In trucking, deregulation has led to tremendous heterogeneity across firms as they pursue different business strategies and seek to serve different segments of the for-hire transportation market.
The five industries studied here include a wide range of different human resource practices. High-tech, high-skill industries such as semiconductors and software should have more skill development with lower turnover and higher wages than a low-skill service industry such as food stores and a low-skill infrastructure industry such as trucking, which differs again from a high-skill infrastructure service industry such as financial services.
The analysis in this book combines facts gleaned from studying millions of data points on millions of firms and workers, as well as from interviews with firms in each industry to answer some key questions.
- How much turbulence is there and why does it happen?
- What is the impact of economic turbulence on:
- Firm performance and survival: What is the relationship between workforce quality, turnover, and firm survival?
- Firm job ladders: What has happened to jobs within a firm? Is it still possible to land a good job that pays good initial earnings with good raises? What kinds of firms offer the best job ladders?
- Worker career paths: What impact has economic turbulence had on workers’ lifetime earnings and employment? How much impact does job loss have on a worker’s earnings?
- Wage distribution: What has happened to middle-, low-, and high-income jobs? Are there still “good” jobs? Do new firms pay more or less than the firms that fail?
The next sections provide a brief preview of the answers to these questions, which are discussed in much more detail in later chapters.
What is turbulence, why does it happen, and what is the impact?
Turbulence is the entire process of economic change: worker reallocation as workers change jobs and job reallocation from firms contracting and shutting down, to firms expanding and starting up. Chapter 2 spells this out in more detail, but the sheer amount of turbulence is staggering. In any given quarter, about one in four job matches either begins or ends, one in thirteen jobs is created or destroyed, and one in twenty establishments closes or is born. Why does it happen? Some turbulence reflects the natural selection processes, and some reflects the fundamental changes in the economy, like globalization, technological change, and deregulation.
Dynamic Selection of Workers
The refrain to a well-known song begins “Take this job and shove it,” and one of the most famous lines in television is “You’re fired.” Put more prosaically, turbulence can be caused by a shuffling of workers across jobs. Firms will hire workers, and workers will accept jobs, but then one or both sides will decide that the job match isn’t right. The worker then leaves and is replaced by someone else. In a lot of low-wage industries, like the retail food industry, this worker reallocation is quite high because the skills required are easily learned and it is easy to replace workers once they leave. In a lot of high-wage industries, like the semiconductor industry, worker reallocation is lower because the costs of replacement are high.
Wal-Mart has made headlines both because of its low prices and because of its low wages. Other firms, like Costco, have workers lining up to work for them. Different firms, even within the same industry, can have different levels of worker turnover simply because firms choose different personnel strategies. This means that different firms have different levels of wages and different amounts of worker turnover. An article in the Seattle Times pointed out the differences between Costco and Wal-Mart:
A cashier at Costco can make more than $40,000 annually within four years. The average store manager makes $107,000, with a crack at $40,000 in performance bonuses on top. The company also pays hourly workers annual bonuses from $4,000 to $7,000. No wonder they stick around: Turnover at Costco is less than a third the industry average.
Costco follows a high-wage, low-turnover strategy, while Sam’s Club, owned by Wal-Mart, has substantially higher turnover and lower wages. The net impact on overall economic turbulence can be substantial: Costco has become one of the ten largest retailers worldwide and has outstripped Sam’s Club in terms of employment, which has had the result of lowering worker turnover in the industry.
Dynamic Selection of Firms
PanAm, Montgomery Ward, Bethlehem Steel. There is a long list of firms that have gone out of business in recent years, with an equally long list of new ones. Turbulence can result from new, more productive firms replacing old, less productive ones, even within the same industry. This process, which Joseph Schumpeter called “creative destruction,” means that jobs get reallocated from one set of firms to another, and accounts for a large fraction of aggregate (industry) productivity growth. In a vivid example of this, some call centers can be closed by firms like Capital One and JPMorgan Chase in the very same city—Tampa Bay—at the very same time that firms like HSBC are opening them.
This turbulent selection process means that economic growth in the U.S. is unsteady and complex. There is much trial and error in companies searching for the “right” way of doing business—the right technology, the right market niche, and the right workforce. As a result, most turbulence occurs within industries: even though more than one in ten jobs are created and destroyed every year in the U.S. economy, only about 10 percent result in employment growing or shrinking across industries.
However, there are big differences across industries. In the software industry, for example, businesses enter and exit quite quickly, but entry and exit rates are much lower in the semiconductor industry.
CNN reporter Lou Dobbs’s book Exporting America (2004) paints a vivid picture of the third reason for economic turbulence: there are fundamental changes in the way in which goods and services are produced. He focuses on globalization, but others have lamented the impact of technological change and deregulation. Changes like these are much harder to measure in a systematic way, which is why our book focuses on an industry-by-industry analysis. As will become clear in chapter 3, globalization is a driving force in the software, semiconductor, and financial services industries with the relocation abroad of some design, manufacturing, and back office activities; technological change has been important in financial services industry, retail food, and semiconductors as is clear from the advent of ATM machines, scanning technology, and smaller, faster chips. And deregulation has had a major impact in the trucking and financial services industries.
A good way to understand how dramatic economic changes like these affect how business is done is to go and directly talk to firms in the industry. That is precisely what the researchers who contributed to this book did. They talked to dozens of firms in each industry using case study techniques that permitted them to describe very specifically the nature and type of external shocks in each industry.
The combination of this approach and the direct measurement of job and worker reallocation and firm entry and exit can lead to a very different view of the world than one gets by reading the newspaper. To take one example, Austin, Texas, has been featured as an example of the negative impact of globalization because semiconductor employment in that city dropped in four years by about sixteen thousand workers, and one-half of its major semiconductor factories closed. However, the facts do not show that Austin’s experience is representative of the industry. The data show that the number of jobs in the semiconductor industry has actually increased, and case study evidence suggests that Austin’s job loss was other cities’ job gain because the structure of the industry changed substantially.
Some things are known about the impact of economic turbulence, but much is not. To start with, not much is known about the relationship between economic turbulence and economic growth. Is the shuffling of jobs across firms and workers across jobs efficient? Does it contribute to economic growth? Chapter 4 begins to answer this question by showing the relationship between the reallocation of workers with varying levels of skill to different types of firms and workers' earnings and firms' performance.
Not much is known about the relationship between economic turbulence and either the job ladders provided by firms or the career paths of workers. Chapter 5 examines the impact of economic turbulence on job ladders by examining the impact of working for high-turnover and low-turnover firms, or for expanding and shrinking firms, on workers’ earnings and earnings growth. Chapter 6 looks at how much job change there is in different industries and examines the impact of job change on worker career paths.
Finally, not much is known about the impact of economic turbulence on the earnings distribution, particularly what has happened to low-income, middle-income, and high- income jobs. A popular concern is that “good” jobs have been lost because the old high-paying firms have been replaced by new firms that pay much less. Chapter 7 examines the evidence on this.
A Preview of the Rest of the Book
Chapter 2 provides an overview of the amount of economic turbulence in the economy,) and chapter 3 surveys the economic change sweeping the five industries. Chapters 4 through 7 are analytical; they discuss the impact of economic turbulence on firms, on firm job ladders, on worker career paths, and on the earnings distribution, respectively. Chapter 8 gives an idea of how the information in the book, and the sources that are used here, can be used in policy analysis. The data appendices provide the background material, including information about the new and rich databases linking outcomes for firms and workers, that underlies the analysis and discussion in this book.
the impact of economic turbulence on firms. Firms’ survival depends on how they organize themselves. Firms behave differently, and these differences matter for their performance and survival. Different firms organize themselves differently, have different levels of workforce quality and workforce turnover, and these differences have significant effects on firm performance. High-productivity businesses have a large share of high-skill workers, with either general skills or experience, and also have low turnover, or churning, of workers. All of these factors independently affect firm survival—businesses with high productivity, low churning, and high skill (especially general skills) are more likely to survive. There are substantial differences across industries—one size does not fit all. For example, low worker turnover is especially important in the semiconductor industry, and having a highly skilled workforce is especially important in the trucking industry. New businesses have a disproportionately important impact in changing production methods, which deserves an important role in the study of entrepreneurship.
There is no “one size fits all” lesson even within an industry, and firms need to be examined within the context of their industry. For example, the popular press makes much of the importance of “small business.” Yet small businesses are very different things across industries, and their importance has changed. In the semiconductor industry, for example, industry restructuring has meant that the industry has moved more to fabless semiconductor establishments. These fabless design firms, which are small, highly skill intensive, and volatile, are changing the dynamics of firm performance in that industry. Similarly, in the retail food and trucking industries, large, national chains operate very differently from small, local entities. And large software producers, like Microsoft, are different from small, agile software producers that typically target small market niches.
the impact of economic turbulence on job ladders. Economic turbulence, as measured by firm growth, substantially affects the number and type of jobs offered by firms. The data confirm what one would expect: across industries, large, growing firms provide some of the best job ladders (initial earnings and earnings growth), and small shrinking firms tend to provide the worst job ladders (and few in number). In general, firms with growing employment offer better jobs than shrinking firms, except in trucking, where large, shrinking firms, often unionized, offer some of the best job ladders. Although large firms are the largest supplier of long-term job ladders, the importance of small and growing firms in providing excellent job ladders in semiconductors, financial services, and trucking indicates that these firms may be a growing source of good job ladders over time.
Economic turbulence often results in low initial earnings but higher than average wage growth at a firm, particularly in semiconductors and financial services. The combination of high turnover with high wage growth for workers that stay suggests that firms are sorting workers and workers are deciding whether to stay, within an "up or out" wage-setting system. Another way of thinking of this is that the job market within a firm is like a tournament: workers compete for “good jobs,” and those who are selected do well, and those who don’t move on to another job.
economic turbulence and career paths. The data show that there are three common career paths (individual earnings trajectories over time) that are observed in each industry. The data are consistent with the popular opinion that loyalists experience better career paths than job changers for all education groups of prime-aged (twenty-five- to fifty-four-year-old) men and women. However, the data also show that, over time, most workers who stay in the labor market are able to improve their career paths through changing jobs until they finally find a relatively good job ladder in a firm, despite not being able to catch up to the loyalists.
Career paths are very different across industries. Just as firm performance varies across industries, career paths for each education-gender group vary greatly across the five industries, with retail food at the bottom and software and semiconductors at the top. In general, workers improved their career paths by moving into the software, semiconductor, financial services, and trucking industries, and by moving out of retail food. Several underlying economic forces might explain these different outcomes in career paths. One force is individually based: loyalists in good jobs may have superior knowledge or other unobserved characteristics that make them more valuable to their employer than the job changers, who may need to increase their skills through experience before landing a better job. Another force is firm based: firms with good job ladders may operate in nonclearing or rationed labor markets with a wage premium—that is, with a wage higher than the market average, and have a queue of fairly homogeneous and qualified workers waiting for job openings. A third force is market based: firms and workers must learn which workers are good matches for which jobs; in effect, workers must change jobs in order to find a “good match.” Most likely it is a combination of these factors.
Many workers continually improve their career paths by finding better job opportunities with other firms, although the typical spell between jobs can be as long as twelve to eighteen months. Although the recent economic downturn has highlighted the costs associated with lost jobs, the long-run evidence is that the consequences of such losses, while important for some workers, are not substantial for most.
economic turbulence and job quality. The popular perception that jobs are vanishing is not correct. There are more jobs in each of the five industries than at the start of the period—and this is not an increase in “bad” jobs. The proportion of low-income workers declined in all five industries. The decline is larger in semiconductors, software, and financial services and smaller in retail food and trucking.
The proportion of high-income jobs increased substantially in high-skilled industries, contrary to fears about the impact of globalization on high-skilled workers. The percentage of high-income workers increased substantially in software, financial services, and especially semiconductors. However, in the retail food and trucking industries, there are fewer high-income workers at the end of the period.
A rising tide has lifted all boats in the high-skilled, high-tech industries. Workers in financial services, semiconductors, and software have seen increases in earnings across the board. By contrast, the retail food and trucking industries have experienced an increase in the middle group with fewer workers in the top and bottom groups.
In sum, although turbulence is very often equated with negative factors, this is only half the story. People see workers getting fired, and jobs being lost as firms shut down because that makes news. The other side of the story, workers getting hired and firms starting up and expanding, is not as visible. This book has been written because new data now exist that can measure many dimensions of turbulence: the reallocation of jobs and workers into as well as out of jobs; the entry and expansion as well as the contraction and exit of businesses. The following chapters explain the basic results in much more detail.