by Mustapha Hammida
August 2014
Job creation and job destruction are phenomena that are impressive and simultaneous in the Minnesota labor market. In earlier articles in this publication we have discussed their statewide distributions and their differences across industries.1 However, much remains unknown about how these distributions behave across regions of Minnesota. In this article we will present a regional analysis of the Minnesota Business Employment Dynamics (MBED). Specifically, we will investigate the significance of regional differences (1) in the distributions of job creation and job destruction and (2) in the type of establishments responsible for job creation and job destruction. The regional analysis will focus on the state six planning areas: Central, Northeast, Northwest, seven-county Metro, Southeast, and Southwest.2
The analysis is based on seasonally unadjusted MBED statistics for the third quarter of 2013. The MBED statistics, produced each quarter with the cooperation of the U.S. Bureau of Labor Statistics, are constructed3 from the Quarterly Census of Employment and Wages (QCEW) by linking employment data at the third month of each calendar quarter at each business or establishment. The QCEW is a list of all businesses subject to Minnesota Unemployment Insurance (UI) and covers more than 97 percent of Minnesota employment.
Comparing third-month employment levels between a particular quarter, referred to as current quarter, and the quarter preceding it helps identify different types of establishments. Two types of establishments that are exclusively responsible for job creation or gross job gains are called expanding establishments and opening establishments. Opposite to these establishments, two other types of establishments are responsible for job destruction or gross job losses and are referred to as contracting and closing establishments.
Statewide, Minnesota establishments created 128,055 jobs between June and September 2013 on a seasonally unadjusted basis. Slightly over half of this gross job gains, or 69,600 jobs, was gained by establishments in the Metro area. Establishments in Central and Northwest areas each accounted for approximately 10 percent of the total gross job gains or 13,212 jobs and 11,119 jobs respectively. The remaining three planning areas contributed to job creation roughly another 20 percent between them, or 9,456 jobs in the Southeast area, 8,416 jobs in the Southwest area and 6,527 jobs in the Northeast area (Table 1).4
| Gross Job Gains, Gross Job Losses, and Associated Labor Market Measures by Planning Region, Quarter Ending September 2013 | ||||||||
|---|---|---|---|---|---|---|---|---|
| Planning Regions | Gross job gains | Gross job losses | Job reallocation | Net job change | ||||
| Jobs | Rate (%) | Jobs | Rate (%) | Jobs | Rate (%) | Jobs | Rate (%) | |
| 7-County Metro | 69,600 | 4.3 | 80,376 | 5.0 | 149,976 | 9.3 | -10,776 | -0.7 |
| Central | 13,212 | 5.0 | 15,112 | 5.7 | 28,324 | 10.7 | -1,900 | -0.7 |
| Northwest | 11,119 | 5.1 | 16,373 | 7.4 | 27,492 | 12.5 | -5,254 | -2.3 |
| Southeast | 9,456 | 4.0 | 12,814 | 5.3 | 22,270 | 9.3 | -3,358 | -1.3 |
| Southwest | 8,416 | 4.7 | 11,286 | 6.4 | 19,702 | 11.1 | -2,870 | -1.7 |
| Northeast | 6,527 | 4.6 | 8,652 | 6.1 | 15,179 | 10.7 | -2,125 | -1.5 |
| Total* | 128,055 | 4.7 | 149,482 | 5.5 | 277,537 | 10.2 | -21,427 | -0.8 |
| Source: Minnesota Department of Employment and Economic Development, MBED Program
*The sum across the planning regions may not match exactly the state total because of establishments with no definite region. |
||||||||
While some establishments contribute to gross job gains of 128,055, simultaneously Minnesota establishments also shed 149,482 jobs during the third quarter of 2013, and these gross job losses were spread in all regions of the state. Interestingly, the regional shares in gross job losses were not much different than those in gross job gains. Indeed, we can distinguish the same three groups of areas. First, establishments in the Metro area produced a gross job loss of 80,376 jobs accounting for 54 percent of the total. Second, about 21 percent of gross job losses was shared almost equally between Northwest (16,373 jobs) and Central (15,112 jobs) planning areas. Lastly, nearly another 22 percent of gross job losses occurred at establishments in Southeast (12,814 jobs), in Southwest (11,286 jobs), and in Northeast (8,652 jobs).
Using gross job gains and gross job losses together adds to our understanding of regional job dynamics. First, taking the difference between gross job gains and gross job losses we get the familiar measure of net job change.5 As the seasonally unadjusted gross job losses were greater than gross job gains in each of the six planning areas between June and September 2013, a negative net employment change was the dominant, distinct feature of third quarter job dynamics. This particular feature may partly be the result of the seasonal reduction in employment between June and September in educational establishments in all regions of the state. A more general feature of regional job dynamics, however, is manifested through the significant levels of job creation and job destruction compared to the net job change in all planning areas.
On the other hand, taking the sum of gross job gains and gross job losses we obtain the job reallocation, a measure of job churn. Given the relative magnitude of regional gross job gains and gross job losses, all six planning areas experience a considerable amount of job churn relative to net employment change. For instance, in the Metro area approximately 150,000 jobs were either created or terminated between June and September 2013, yet they resulted in a negative net job change of only 11,000 jobs. While this relationship is also strongly evident in each area within Greater Minnesota, it is achieved with much lower levels of job churn compared to the Metro area. In fact, job reallocation in Greater Minnesota ranged between a high of 28,324 jobs in the Central area and a low of 15,179 jobs in the Northeast area.
Further dissecting the relationship between job reallocation and net job change, not all planning areas experienced the same intensity of job dynamics. The Metro and Central areas appeared to have a more bustling labor market as they recorded the strongest job churn of all planning areas. In these two areas a reduction in the levels of employment between June and September by one job involved approximately 15 jobs that were either created or eliminated during the three month period. By contrast, in the other four areas — Northeast, Northwest, Southeast, and Southwest — a reduction in their levels of employment between June and September by one job involved only between five to seven jobs churned.
Returning to the job churn by itself, we find the regional shares in total job reallocation follow a similar distribution as those of gross job gains and gross job losses discussed above. The Metro area witnessed an impressive 54 percent of the state job churn, while the five regions of Greater Minnesota collectively generated about two-fifths. The first one-fifth was approximately evenly split between Central and Northwest areas, and the second one-fifth approximately evenly shared among the other three areas: Southeast, Southwest, and Northeast.
The above discussion may at first glance indicate that the regional shares in gross job gains and gross job losses are unbalanced and thus, worrisome, i.e., most of job creation and job destruction occurs in the Metro area. Moreover, note that up to this point we have made no mention of the regional distribution of employment. Bringing the regional shares in employment into the picture reveals a healthy distribution of job creation and job destruction across the planning regions of the state (Chart 1). Interestingly enough, the regional shares of job creation and job destruction mimic the regional shares of the state employment. Thus, all planning areas fare equally well in their job dynamics relative to their employment levels and shares.
A more accurate analysis on the significance of regional differences in the distributions of job creation and job destruction can be achieved by studying the relative measures of job dynamics. We obtain these relative measures through dividing a particular job dynamic numerical measure by the average employment in the current and the previous quarters. Table 1 gives also the relative measures of job dynamics by planning region for the third quarter of 2013.
Statewide for the period between June and September 2013 gross job gains represented 4.7 percent of total employment, and gross job losses represented 5.5 percent of total employment. Intuitively, these rates tell us that 4.7 percent of jobs in September 2013 were not there in June 2013, and 5.5 percent of jobs that were there in June 2013 disappeared in September 2013. Comparing the extents of job creation and job destruction across the planning areas reveals the existence of a range of variability. Specifically, some major findings in Table 1 are:
The heterogeneity of job creation and job destruction between Minnesota planning areas is also evident in the job churn as a percent of employment. The lowest rate of job reallocation was 9.3 percent of total employment in two areas: Metro area and Southeast area. In other words, in these two areas 10 of 108 jobs were either created or terminated between June and September 2013. At the other end the highest rate of job reallocation was in the Northwest area at 12.5 percent of total employment. Expressing this percentage differently means, for comparison purposes, that about 14 jobs of the 108 mentioned above were created or terminated in the Northwest area. The second-highest rate of job churn was in the Southwest at 11.1 percent of total employment. The rate of job reallocation in the remaining two regions, Central and Northeast, was 10.7 percent. Thus, the seasonally unadjusted job dynamics for third quarter 2013 points to a possible regional clustering, pinning Northwest to Southwest, Central to Northeast, and Metro to Southeast.
According to the MBED statistics, job creation comes about from two types of establishments responsible for adding jobs: expanding establishments and opening establishments. For the purposes of the MBED, expanding establishments are businesses whose non-zero third month employment levels increase between previous quarter and current quarter. On the other hand, opening establishments are businesses whose third month employment levels in the previous and current quarters increase from a level of zero in the previous quarter. Together, jobs added at expanding and opening establishments form gross job gains or job creation.
Job creation between June and September 2013 was dominated by expanding establishments in all Minnesota planning regions (Chart 2). Aside from regional variation, expanding establishments generated almost nine-tenths of gross job gains. Opening establishments accounted for the remaining one-tenth. Further investigation of the breakdown of gross job gains between expanding and opening establishments across regions shows some considerable differences. While opening establishments accounted for 11.9 percent of gross job gains generated in the Metro area, they could generate only 6.7 percent of jobs created in the Southwest area. Looking at this from the other side of the coin, 93.3 percent of the jobs created in the Southwest area were at expanding establishments, 7,855 jobs of 8,416 gross jobs gained. The share of gross job gains coming from opening establishments in the other four areas varied between 7 and 10 percent.
A more comprehensive examination of the sources of job creation across regions can be accomplished by examining the regional rates of gross job gains by establishment type. Although expanding establishments dominated job creation in all areas, the job creation engines did not run at the same speed in all regions. Among expanding establishments, the fastest pace was in the Northwest area at 4.6 percent of total employment, while the slowest pace was in the Southeast at 3.7 percent of total employment (Table 2).
| Gross Job Gains and Losses by Establishment Type and Planning Region, Quarter Ending September 2013 | ||||||||
|---|---|---|---|---|---|---|---|---|
| Planning Regions | Gross job gains | Gross job losses | ||||||
| Expanding establishments | Opening establishments | Contracting establishments | Closing establishments | |||||
| Jobs | Rate (%) | Jobs | Rate (%) | Jobs | Rate (%) | Jobs | Rate (%) | |
| Central | 11,889 | 4.5 | 1,323 | 0.5 | 13,557 | 5.1 | 1,555 | 0.6 |
| Northeast | 6,017 | 4.2 | 510 | 0.4 | 7,692 | 5.4 | 960 | 0.7 |
| Northwest | 10,110 | 4.6 | 1,009 | 0.5 | 14,611 | 6.6 | 1,762 | 0.8 |
| Southeast | 8,781 | 3.7 | 675 | 0.3 | 11,985 | 5.0 | 829 | 0.3 |
| Southwest | 7,855 | 4.4 | 561 | 0.3 | 10,219 | 5.8 | 1,067 | 0.6 |
| 7-County Metro | 61,344 | 3.8 | 8,256 | 0.5 | 71,365 | 4.4 | 9,011 | 0.6 |
| Total* | 111,151 | 4.1 | 16,904 | 0.6 | 132,515 | 4.9 | 16,967 | 0.6 |
| Source: Minnesota Department of Employment and Economic Development, MBED Program
*The sum across the planning regions may not match exactly the state total because of establishments with no definite region. |
||||||||
Shifting gears to job creation at opening establishments, we detect a less impressive regional variation compared to expanding establishments. The rate of gross job gains at opening establishments as a percent of total employment had a range of only two-tenths of a percentage point. The highest rate was 0.5 percent in the Central, Metro, and Northwest areas, and the lowest rate was 0.3 percent in the Southeast and Southwest areas.
As with gross job gains, the MBED statistics break job destruction between two types of establishments responsible for terminating jobs: contracting establishments and closing establishments. Contracting establishments are businesses whose non-zero third month employment levels decrease from previous quarter to current quarter. On the other hand, closing establishments are businesses whose third month employment levels in the previous and current quarters decrease to a level of zero in the current quarter. Together, jobs terminated at contracting and closing establishments form gross job losses or job destruction.
Chart 3 gives the shares of contracting and closing establishments in gross job losses by region. Clearly, job destruction between June and September 2013 was dominated by contracting establishments in all Minnesota planning areas. At least close to nine-tenths of job destruction emanate from contracting establishments, leaving only about one-tenth to closing establishments. By planning area the largest share of contracting establishments in gross job losses was in the Southeast area at 93.5 percent, while the smallest share was in the Metro area at 88.8 percent. Excluding the Southeast area, the five regional shares of job destruction by source seem to bunch up over a range of barely 2 percentage points. For example, the share of closing establishments in job destruction varied from 9.5 percent to 11.2 percent. Similarly, the share of contracting establishments in job destruction varied from 90.5 percent to 88.8 percent.
This moderate concentration, however, appears to mask some significant regional differences in the extent of job destruction by source, in particular for contracting establishments. Consider, for instance, the Metro, Northeast, and Southwest areas. In these areas the share of contracting establishments in job destruction was almost identical, 88.8 percent in the Metro area, 88.9 percent in the Northeast, and 89.2 percent in the Southwest. However, the extent of job destruction as a percent of employment at contracting establishments was far from being identical between these three areas. Specifically, the rate of job destruction from contracting establishments was 6.6 percent in Southwest Minnesota, the highest rate of all Minnesota planning areas, 5.4 percent in Northeast Minnesota, and only 4.4 percent in the Metro area, the lowest rate of all Minnesota planning areas (Table 2).
Not to neglect the closing establishments, they too show variation in their capacity for terminating jobs, albeit at a more modest volatility, than the contracting establishments. Among closing establishments, the highest rate of gross job losses as a percentage of employment was 0.8 percent in the Northwest area while the lowest rate was 0.3 percent in the Southeast area. The second highest rate was in the Northeast area at 0.7 percent. The remaining three areas all had a rate of 0.6 percent of total employment.
In the broader context of the MBED statistics, the net employment change captures fluctuations in employment levels between the third months of two successive calendar quarters. These fluctuations summarize the net effect of the employment decisions at a wide range of establishments, some of which gain jobs while others lose jobs. So putting together the gross job gains from expanding and opening establishments with the gross job losses from contracting and closing establishments, we are able to discover the forces that shape the net employment change. We can achieve this by comparing the gross job flows, on one hand, between expanding and contracting establishments and, on the other hand, between opening and closing establishments.6
As established above from Table 1, all Minnesota planning areas experienced job decline between June and September 2013. This happened because job gains did not keep up with job losses between expanding and contracting establishments and between opening and closing establishments in all areas except Southeast Minnesota (Table 2). The negative employment change in Southeast Minnesota was mainly from the imbalance in job flows between expanding and contracting establishments. In addition, its job creation at opening establishments was just enough (0.3 percent) to offset its job destruction at closing establishments (0.3 percent).
With the exception of the Metro and Central areas, no two areas were alike in how their net employment change came about. Table 2 indicates that the Metro and Central areas had the lowest job decline, -0.7 percent, between June and September 2013. Both areas saw a drop of 0.6 percent in employment as a result of job flows between expanding and contracting establishments and another drop of 0.1 percent as a result of job flows between opening and closing establishments. Although these results may hint that Metro and Central areas have identical job dynamics, the two areas are far from being similar. In particular, the job flows between expanding and contracting establishments were more extensive in Central than in the Metro area.
The findings we reached in this regional analysis of MBED statistics bear many similarities to those we obtained in the earlier articles. First, gross job gains or job creation and gross job losses or job destruction are significant and occur simultaneously in all Minnesota regions. Second, in each region, most of the job creation was generated by expanding establishments, and likewise most of the job destruction was generated by contracting establishments. Third, heterogeneity across regions characterized all distributions of gross job gains and gross job losses by type of establishment. Finally, a word of caution in generalizing these results is warranted. It is likely that the extent and magnitude of these heterogeneities across regions are time-dependent. The results of this analysis are pertinent to the third quarter time period. Job dynamics from other quarters of the year might yield different results. Extending this analysis to each of the other quarters will improve our understanding of how regional job creation and destruction evolve over a calendar year.
1See my articles in the Review publication of July 2014, June 2012, and May 2012
3For a detailed discussion on how the BED statistics are developed.
4Regional totals do not add up to the gross total because some firms do not have a fixed physical location in the state.
5Some care needs to be exercised when using the net employment change from MBED. The net employment change from the MBED statistics represents not a monthly, but an aggregate of three monthly net changes spanning the four months contained within the third months of two successive calendar quarters. Let E denote employment and using abbreviation for months, net employment change between June and September is given by: ESep – EJun = (ESep – EAug) + ( EAug – EJul) + (EJul – EJun).
6Using simple algebra, net employment change = gross job gains from all establishments – gross job losses from all establishments
= (gross job gains from expanding + gross job gains from openings) - (gross job losses from contracting + gross job losses from closings)
= (gross job gains from expanding - gross job losses from contracting) + (gross job gains from openings - gross job losses from closings)
