Companies come; companies go. And the combination is what keeps the economy hopping and jobs growing, according to recent studies that focus on regional economies

Many people would find a stable economic order very convenient. Businesses could develop long-range programs that would optimize their efficiency and, presumably, provide a predictable return to owners and stockholders. Those in the workforce could plan their work lives and personal lives better. Policymakers could develop programs to assist businesses, regulate economic activities, and monitor the distribution of wealth among citizens.

Despite the advantages of a predictable economic order, however, one disadvantage looms: stability in an economic structure is associated with economic decline. In market economies, businesses are interdependent players in a fluid, volatile system undergoing constant change. That change -- often provided by entrepreneurs -- reflects adaptation to new technologies, new consumer demands, and new forms of local, national, and international competition. The churning can have costs as companies disappear, individuals switch jobs, and policymakers try to keep up with the current economic order. But the cost of a system without change -- economic stagnation -- is even higher.

Recent regional studies analyze the processes of turbulence -- business births and deaths -- and growth. The patterns that are often masked in broad nationwide studies become readily apparent when regions within a country are compared. (In any one year there is much greater variation across regions in the United States than the entire country experiences in decades.) Specialized data sets developed by the Small Business Administration for 1976 through 1988 allow researchers to separate the births and deaths of independent companies from the births and deaths of branches. The following discussion emphasizes autonomous start-ups.

Interestingly, company births and deaths are highly correlated: regions tend to be either high or low in both start-ups and shutdowns. We have also found positive correlations throughout the 1980s between business turbulence and the percentage of net job gains in the same region two years later. For example, from 1984 through 1986 the annual independent-company birthrates in the Fort Myers, Fla.; Grand Junction, Colo.; and Traverse City, Mich.; regions were twice the national average; from 1988 through 1990 the average net job gains there were more than 4% a year, more than twice the national average. Similar patterns have been found in other advanced Western economies, such as Sweden, where it has been possible to separate independent companies from branches.

Turbulence does not always precede growth. Some regions with high turbulence do not have growth in later periods. The southeastern region of Colorado had an independent-business birthrate twice the national average from 1984 through 1986 but an average net job gain (.8%) less than half the national average from 1988 through 1990. It is rare, however, to find growing regions without turbulence in prior periods. Business turbulence is a necessary, but not sufficient, condition for growth in jobs.

The question, then, for those of us concerned about economic growth is, what types of regions have high company birthrates? With the support of the European Union and the Organization for Economic Cooperation and Development, research teams produced comparable analyses for eight countries -- France, Germany, Ireland, Italy, Japan, Sweden, the United Kingdom, and the United States. The findings were roughly similar across all the countries.

The most fertile regions had annual start-up rates two to three times higher than the least fertile regions. In general, regions with higher business birthrates in all economic sectors are those with increases in population, higher levels of personal income, greater household wealth, greater urbanization, higher levels of unemployment, a larger proportion of small businesses (independent companies and branches of other companies), and greater economic specialization. (The specialization index is high if more people work in one sector, such as manufacturing or wholesale, and low if workers are uniformly distributed across all sectors.) The last two factors (higher proportion of small establishments and economic specialization) have a particularly strong relationship to the birth of manufacturing companies.

Virtually no evidence in the U.S. regional comparisons suggests that lower production costs increase start-up rates. That may be because most new businesses are started by established residents who know the region and are prepared to respond to business opportunities "at home." There is evidence, however, that production costs can be important for some types of manufacturing branch start-ups as international businesses shift production from site to site for temporary cost advantages.

There is no question that higher company birthrates and economic growth occur together. But which comes first, business start-ups or economic growth? We found that company births have a small but statistically significant impact on growth, particularly in manufacturing. New companies appear to be the major mechanism by which competitive advantages associated with location, raw materials, or human potential are translated into shifts in the economy.

The implications for government policies are substantial. There is little systematic evidence that variations in government expenditures, in the presence of government programs, or in the costs of doing business have much impact -- positive or negative -- on the number of start-ups. Given the massive amount of human and financial resources devoted to new business start-ups, governments may not have the resources to make much difference, except perhaps to slow down the process. Careful studies in Peru found that when the regulatory burden on start-ups became too onerous, most start-ups and growth occurred without government approval or control.

Efforts to use public funds to subsidize businesses in trouble or even entire industries in decline may be more than just a waste of money. They may delay the transfer of critical resources -- facilities and equipment, money, workers, administrative and managerial talent -- to new businesses or growth industries. While such delays may reduce turbulence, they may also slow economic adaptation, putting the region at a competitive disadvantage.

Using public resources to reduce transition costs for both workers and businesses would be of benefit. Programs for retraining workers, for example, or for redeploying financial and physical assets make better use of public funds than subsidies to unprofitable businesses in declining industries do. Government efforts to reduce the costs associated with business births and deaths could also facilitate economic change and adaptation.

A critical decision for social policy is the extent to which public monies should subsidize entrepreneurial behavior in low-potential regions. In virtually every advanced country some regions -- usually rural and agricultural ones -- experience systematic economic decline and, usually, population loss. For example, the regions around Cheyenne, Wyo.; Bluefield, W. Va.; and Lawton, Okla.; had net job losses almost every year from 1984 through 1990. New business start-ups are often seen as one option for improving their economic situation. But in the absence of customers with money to buy products, the long-run viability of unsubsidized new businesses is hard to determine. Business turbulence without economic growth is expensive for everybody.

The recent studies described here are only a beginning. A lot more data are needed to understand the effect of business births and deaths on economic growth. Unfortunately, few countries have complete, reliable data on the birth and death of independent businesses or annual measures of economic well-being for their internal regions. Improved regional data will help us to learn much more about the processes that lead to economic growth -- and economic stagnation -- knowledge that, in turn, could inform and promote more efficient and effective government policies.

Paul Reynolds is the Coleman Foundation Chair in Entrepreneurial Studies at Marquette University, in Milwaukee. Research on the United States was completed with Wilber Maki of the University of Minnesota. Research on European Union countries was coordinated by David Storey of the University of Warwick, U.K.