by Steve Hine
December 2018
PDF of article
Wrap up: Mighty improvements in unemployment, but we're undergoing a long period of slowing labor force growth.
With change in the gubernatorial administration upon us, it's an opportune time to look back over the past eight years of the Dayton administration, assess our employment conditions, and see how these conditions compare to previous administrations. We do this not because we believe any elected state official deserves all the blame or all the credit for changing conditions well beyond their control. Quite the contrary. This comparison highlights the fact that governors of both parties have seen broader cyclical trends impact the state's economy. To ignore these impacts and attribute good or bad performance to the office holder overstates their influence over these trends at a state level.
Nonetheless, gubernatorial administrations are frequently used as time references in discussing the state of our economy, and it is useful to compare each to predecessor administrations. In a time when partisan preferences often override evidence in our personal gubernatorial assessments, it is especially valuable to provide an objective comparison for the record.
Availability of data constrains our analysis to a few recent gubernatorial administrations. Total non-farm employment is available back to 1950, but a dramatically changing economy over the past few decades, and corresponding changes in the classification system used to describe the components of our economy, most notably the adoption of the North American Industry Classification System (NAICS) in 2000, make a detailed assessment more challenging than is intended. Another limitation is due to state-level estimation of unemployment and other labor force indicators starting in 1976.
But comparing the behavior of such high-level aggregates as total non-farm employment and the unemployment rate still serves our purposes. We compare the behavior of total employment beginning with Elmer Andersen who took office in January 1961, and we can compare unemployment rates beginning with Rudy Perpich's first term when he replaced Wendell Anderson in December 1976.
Table 1 provides high-level indicators of employment and unemployment conditions across administrations. Average employment changes range from the 6,677 per month during Perpich's first abbreviated (two-year) administration to the 894 average loss during Al Quie's subsequent tenure. Those of us who remember those times 40 years ago recall that the driving force behind these disparate outcomes was the fact that Quie had the misfortune to be in office during two national recessions during his one four-year term. Similarly, the only other administration to have shed jobs over its time in office was Pawlenty's, and here again this was in no small part due to the Great Recession coming midway through his second term.
| Table 1. Employment Gains and Recessions by Gubernatorial Administration | ||||
|---|---|---|---|---|
| Term Start | Governor | Average Monthly Employment Change | Average Monthly Employment Growth Rate | Share of Months in Recession |
| January 1961 | E. Andersen | 1,615 | 0.17% | 8% |
| March 1963 | Rolvaag | 4,020 | 0.37% | 0% |
| January 1967 | LeVander | 2,620 | 0.21% | 25% |
| January 1971 | W. Anderson | 3,158 | 0.23% | 24% |
| January 1977 | Perpich (1st term) | 6,677 | 0.42% | 0% |
| January 1979 | Quie | -894 | -0.05% | 50% |
| January 1983 | Perpich (2nd term) | 4,672 | 0.25% | 6% |
| January 1991 | Carlson | 4,671 | 0.20% | 3% |
| January 1999 | Ventura | 1,504 | 0.06% | 19% |
| January 2003 | Pawlenty | -104 | -0.00% | 19% |
| January 2011 | Dayton | 3,473 | 0.12% | 0% |
Indeed, the timing of our national recessions has much to do with the relative performances of our gubernatorial administrations. Of the 11 listed, only three spanned a time when there were no recessions, including Rolvaag's one term, Perpich's short (two-year) first term, and now Governor Dayton's two terms. In addition, both Elmer Andersen and Arne Carlson took office during a recession, both of which ended within a short (two months and three months following inauguration, respectively) time after, and both administrations were recession-free from then on. Perpich's second eight-year term ended in recession, the relatively mild nine-month downturn that began in July 1990 and spilled over into Carlson's term. The remaining governors (LeVander, Anderson, Ventura, and Pawlenty) each served a term that included one recession.
To see how this timing matters, we note that the administrations with the five highest average monthly job gains, including Dayton's, are also those with the lowest share of months in recession, and the three slowest job growth administrations are those with relatively high shares in recession. An interesting illustration of this dependence of average job growth on the timing of recessions is a comparison between Perpich (2nd term) and Carlson. They shared a mild nine-month recession between them, six months for Rudy and three for Arne, but otherwise each served eight-year terms without another downturn. And as a result, their respective average monthly job gains differed by one job.
But as we consider Dayton's record, the close correlation between being recession-free, or nearly so, and seeing high monthly job gains slips somewhat. Comparing Dayton's job growth rate to Perpich's and Carlson's, we see that Dayton's job growth is 1,200 per month lower than during both of those earlier administrations. Does this imply that we have been underperforming for some reason relative to these earlier administrations? Table 2 reveals this is not the case; rather, the reason we have seen job growth slip is that over the last two administrations, the rate at which our labor force has grown has fallen, as baby boomers began to hit retirement age. In fact, relative to the Perpich (second term)/Carlson era, labor force growth is lower during the Dayton years by 1,284 per month, just about enough to account for the shortfall in employment growth. In other words, despite being the only recent governor to hold office through a full recession-free two terms, his average employment growth lagged previous office holders simply because demographic trends reduced the number of warm bodies available to take jobs that might otherwise have been available.
This is further illustrated in Table 2. The average rate of unemployment during each gubernatorial term depends greatly on where the rate stood at the beginning of each term, or in other words what was happening to the rate in the prior term. For example, Ventura experienced the lowest average unemployment rate of the governors listed, but this is largely because it was so low by the time he took office following the long expansion of the 1990s. What is more revealing about the behavior of unemployment over the course of any administration is how it changed from beginning to end.
| Table 2. Labor Force Growth and Unemployment by Gubernatorial Administration | ||||
|---|---|---|---|---|
| Term Start | Governor | Average Monthly Labor Force Growth | Average Unemployment Rate | Change in Unemployment Rate During Term |
| January 1977 | Perpich (1st term) | 4,968 | 4.6% | -2.1 |
| January 1979 | Quie | 3,409 | 5.9% | +5.1 |
| January 1983 | Perpich (2nd term) | 2,580 | 5.5% | -3.8 |
| January 1991 | Carlson | 3,245 | 4.1% | -2.6 |
| January 1999 | Ventura | 2,776 | 3.5% | +1.9 |
| January 2003 | Pawlenty | 840 | 5.3% | +2.6 |
| January 2011 | Dayton | 1,629 | 4.4% | -4.1 |
By this measure, Dayton's time in office looks slightly better than those of both Perpich and Carlson, with a 4.1 percentage point decline, from 6.9 percent to the current 2.8 percent (although we should note that the 2.1 point decline in Perpich's first term was accomplished in two years rather than eight). Not surprisingly, Quie and Pawlenty, two administrations in office during job declines, also saw the largest increases in the unemployment rate during their tenure. Ventura also held the post during an unemployment rate increase, but the 2.5 percent rate when he took office is our lowest on record and gave it essentially nowhere to go but up. So by virtue of the long expansion during his term, Dayton has seen the unemployment rate improve as much as, or more than, any of his recent predecessors. He leaves office with conditions, at least by these very top-level indicators, in much the same shape as Arne Carlson left them for Jesse Ventura.
Of course, a full evaluation and comparison of gubernatorial administrations is not our goal here. These data, being as limited as they are, don't provide for the kind of detailed analysis that may attribute credit or blame for changing conditions. In fact, we can see from these data that relative performance across administrations depends on factors such as whether the nation experiences a recession while in office, or on long-term demographic trends and an aging population. These are not factors under the influence or control of any party or administration.
Mark Dayton has had the good fortune to serve during a time free from national recession, a fortune not enjoyed by many of his predecessors; but he has also served during a period of slowing growth in our labor force. As a consequence, the last eight years have seen dramatic improvements in unemployment, with slower but still impressive rates of job growth. Let the record show that during Mark Dayton's time in office, improvements in our economy align closely with those experienced under Perpich and Carlson, or in other words, pretty nicely.
