What is Oklahoma’s business climate? That is an old question, and one that is usually answered with one anecdote or another. Unfortunately, using anecdotes to create public policy often leads to a “shotgun” approach to legislation. A lot of tiny problems get addressed, but no one is able to put together a view of the bigger picture.

A large part of the problem—until now—has been a lack of detailed empirical data on Oklahoma’s business establishments. But using a powerful new database of Oklahoma’s businesses called the National Establishment Time-Series (NETS) Database,1 the most comprehensive establishment-level census available, we have undertaken a study of job creation in Oklahoma.

In last month’s issue of Perspective we looked at one avenue of job creation: the in- and out-migration of business establishments. This month we will examine the expansion and contraction of Oklahoma establishments.

Every year in Oklahoma, existing establishments, based on their economic condition, add or eliminate jobs. Understanding this dynamic process relating to the creation of jobs from existing establishments is vital to ensuring that public policy helps rather than hinders job creation.

Oklahoma Jobs Gained and Lost

Each year between 1993 and 2008, Oklahoma establishments, on average, created 64,699 jobs from expansions while destroying 47,676 jobs from contractions—leaving an average annual job gain of 17,023.2 Over the entire 15-year time-period examined in this study, net establishment expansions created 272,377 jobs.3

Chart 1 and Table 1 show that Oklahoma’s year-to-year fluctuations in employment are influenced by the net job creation from expansions and contractions. While the two are 76 percent positively correlated—which means they generally move together in the same direction—note that expansions and contractions reinforce the employment trend rather than define it. For instance, net job creation from expansions and contractions only drops into negative territory on one occasion (in 2002) while total employment goes negative five times (1995, 2002, 2003, 2007 and 2008).

More specifically, Chart 2 and Table 1 shed light on whether expansions or contractions are driving the overall trend. Job losses from contractions are generally very stable from year to year, with the exception of 2002, when contractions spiked leading to the one year of overall decline from net expansions and contractions. Otherwise, the rate of job growth from expansions determines how many net new jobs are created in this category.

Overall, this analysis shows that expansions of existing establishments have yielded reliable job growth but still don’t fully explain the year-to-year job fluctuations, especially negative net change in total job creation.

Oklahoma’s Ranking

Table 2 shows the net change in jobs created by expansions and contractions between 1993 and 2008 as a percent of employment in 1993. Oklahoma’s 272,377 jobs created due to establishment expansion equates to a gain of 17.7 percent of Oklahoma’s 1993 workforce. Oklahoma’s performance ranks a respectable 19th in the country. In stark contrast, the state with the largest job growth due to expansions was Arizona (31.7 percent).

Two neighboring states rank higher than Oklahoma: Colorado (24.6 percent, ranked 4th) and Texas (20 percent, ranked 13th). The other four neighboring states performed more poorly: Arkansas (15.9 percent, ranked 24th), Kansas (14.7 percent, ranked 29th), Missouri (14.9 percent, ranked 28th), and New Mexico (15.4 percent, ranked 26th).

Conclusion

While establishment expansions have been a reliable and substantial source of new jobs for Oklahoma, we have not fully answered the question of which of the three sources of job creation—in-/out-migration; expansions/contractions; and births/deaths (which we’ll examine next month)—explains the year-to-year fluctuations in total jobs.

When comparing Oklahoma nationally and regionally, job creation from expansions and contractions is about average, so there is plenty of room to rev up the job creation engine in this regard. More specifically, our analysis shows that focusing on establishment expansions, rather than worrying about minimizing contractions, is most productive since expansions determine how many news jobs are created in this category.

Since adding jobs to an existing workforce often involves significant physical investment, such as a bigger manufacturing facility, one policy prescription would be to remove or streamline regulations that may overly burden such expansions. Otherwise, stalled job expansions could just as easily become job out-migration.

Endnotes

1. NETS is based on the far-reaching Dun & Bradstreet Marketing Information file that has tracked more than 41.7 million establishments nationwide between 1989 and 2008. The file tracks businesses via an assigned “DUNS number,” which is the business equivalent of a personal Social Security number. The NETS database is the most comprehensive establishment-level census available. The firm Walls & Associates performs the conversion of the Dun & Bradstreet Marketing Information file into a time-series database that is useful for economic research purposes. The file is proprietary to Walls & Associates, which licenses the database to researchers across the country—including the U.S. Department of Commerce’s Census Bureau and the Bureau of Economic Analysis.

The NETS database is based on establishments, which means that one organization can have numerous establishments in various locations, e.g., Starbucks. Additionally, different establishments can occupy the same location. For example, an organization at a single location could represent two different activities—such as a single organization with both a distribution and retail establishment under the same roof. This structure provides an unprecedented level of geographic and industry classification.

Though this study will use the term “Oklahoma jobs,” it does not mean that those employed are all Oklahoma residents. Since jobs are reported on a per establishment basis, there is no information on the residency of the workforce. Therefore, someone living in Texas but working for an Oklahoma establishment would be included in the “Oklahoma employment” number.

2. This study draws on data for national comparisons from www.YourEconomy.org (YE), which is based on the national NETS database. YE is a project of the Edward Lowe Foundation. For technical reasons, the YE website uses the NETS database for the years 1993 to 2008.

3. Some establishment contractions may, in fact, be a form of out-migration of jobs from one region or state to another. For example, a company may decide to consolidate several far-flung establishments under one roof.

If the new consolidated establishment is not located in Oklahoma, then there is no way to determine whether an establishment contraction was really a form of out-migration.

Economists J. Scott Moody (M.A., George Mason University) and Wendy P. Warcholik (Ph.D., George Mason University) are OCPA research fellows.