Printer Friendly

Jobless Recovery: A Time Series Look at the United States.

Introduction

The phrase "jobless recovery" became popular in the U.S. during the 2000 recession, when it took seven straight quarters of gross domestic product (GDP) growth to result in decreases in the unemployment rate. With the 2008 recession, the phrase found new life in the beginning stages of recovery. The term is actually first found in print in The New York Times during the depression era of the 1930's. (1) One definition of the concept reads "an economic recovery, following a recession, where the economy as a whole improves, but the unemployment rate remains high or continues to increase over a prolonged period of time". (2) In this paper we study jobless recovery as a relative term looking at the relationship of GDP growth and unemployment rates over time. We consider a post-recessionary period as a jobless recovery if the speed at which the rate of unemployment declines is statistically and significantly slower than prior recessions. The statistical evidence found in this paper lends credence to the applicability of the phrase when describing changes in the U.S. labor dynamics over time. We find that as U.S. GDP growth recovers after a recession, the size of decreases in the unemployment rate have lessened over time.

Others have written papers dealing with jobless recovery but without isolating recoveries or looking at the long time series vector auto-regression (VAR) relationship between actual unemployment rates and GDP growth. Groshen and Potter (2003) look at differences in the recovery from the 2001 recession compared to the recovery from the 1990-91 recession and find the summary evidence points to possible structural changes. They find the data suggest an increase in permanent job losses over temporary layoffs and inter-industry relocation of jobs may have created the 2001 jobless recovery. Chinn et al. (2014) also suggest jobless recovery is due in part to structural change. They use employment data going back to 1950 in a non-linear smooth transition error-correction model to ascertain whether the sluggish employment growth is part of a trend or due to specific business cycle factors. They find that U.S. employment in 2012 is about 1.2 million lower than predictions based on the historical co-movement of employment and GDP.

Aaronson et al. (2004) focus on comparisons to the 2001 recession looking at the effects of self-employment on jobless recovery. They use panel data of U.S. state unemployment rates from the Current Population Survey (CPS) back to 1979 to estimate predicted self-employment rates for the periods following the 2001 recession until the fourth quarter of 2004. They find the predicted estimates are well below the actual value for that time frame. They conclude that the joblessness of the recovery after the 2001 recession may be attributed to the temporary nature of the self-employed jobs. If these jobs are indeed temporary in nature, then not until the labor markets recover and real wages rise will there be a shift back from self-employment to employment, keeping unemployment rates persistently high. Jaimovich and Siu (2012) construct a search-and-matching model to show how the disappearance of middle-skill jobs contributes to jobless recovery. They link the disappearance to technological routinization with job polarization concentrated in downturns.

Faberman (2017) addresses jobless recovery using the Business Employment Dynamics (BED) data from the Bureau of Labor Statistics (BLS) covering the period from 1990-2006. He uses flow data to study job creation and job destruction as defined by Davis et al. (2006), again focusing mainly on the recoveries from the 1990-91 and 2001 recessions. Faberman extends his data back to 1947 using the BED and the previous estimates to create generalized method of moments (GMM) predicted estimates of job creation and job destruction. Much in line with the structural breaks found here, Faberman observes the magnitude of job flows began to steadily decline in the 1960's and the volatility of job flows dropped sharply in the mid 1980's. He attributes the jobless recovery from the 2001 recession to a persistent decline in the job creation rate and the recovery from the 1990-91 recession to an increase in the job destruction rate. He links them both to the reduction in volatility and increased persistence of job flows in the presence of aggregate shocks as seen in the Great Moderation period (Kim and Nelson 1999; McConnell and Perez-Quiros 2000).

The novelty of this paper is that it addresses the phenomenon of jobless recovery directly by isolating the recovery portion of the unemployment cycle using a long time series of available quarterly data. This approach allows us to quantitatively substantiate the statistical significance of the comparative joblessness of recoveries over time. Using VAR models with two endogenous structural breaks and controlling for changes in GDP, we show that following a recession, decreases in the unemployment rate significantly slowed over time. Specifically, the coefficients on the two structural breaks are positive and successively larger. Thus, holding GDP growth constant yields weaker recovery in the unemployment rate over time, i.e. jobless recoveries.

The first structural break supports the notion that the Great Moderation is part of a trend that goes as far back as the 1950s with a possible deviation in the 1970s (Blanchard and Simon 2001). The second structural break coincides with the period most often cited as the beginning of the Great Moderation in the 1980s. These breaks suggest that the most recent changes in unemployment dynamics are linked to the documented changes in output and productivity dynamics. A large body of work examines both statistically and theoretically the various macroeconomic shifts that occurred throughout the Great Moderation in the U.S. We utilize the evidence presented here to discuss a number of the hypotheses put forward for jobless recoveries in the Great Moderation era.

Data

We first present a preliminary look at both unemployment and GDP behavior over time. The unemployment data come from the BLS (2014) and the real GDP data (chained 2005 dollars) are from the Bureau of Economic Analysis (BEA 2014). Both series are seasonally-adjusted, quarterly, and span 1948 to 2012.

The mean unemployment rate over the time period is 5.81%, and the mean quarterly real GDP growth rate is 0.78%. We split the data into three periods: 1) before the fourth quarter of 1959 (1959Q4); 2) after 1959Q4 but before the fourth quarter of 1984 (1984Q4); and 3) after 1984Q4. The divisions stem from endogenous structural break points found in the following section. Both breaks indicate a statistically significant slowdown in the rate at which the unemployment rate falls post-recession.

The average unemployment rate before 1959Q4 was 4.56%, 6.05% after the break until 1984Q4, and 6.11% thereafter. The average negative change in the unemployment rate was -0.41% before 1959Q4, -0.25% after 1959Q4 until 1984Q4, and -0.17% after 1984Q4. These numbers demonstrate that the average negative movement in unemployment is smaller in absolute terms after each structural break.

These observations lend support to the idea that recovery from high unemployment will take longer after the structural breaks, but do not address the relationship between GDP and unemployment. The simple averages, of course, could coincide with differences in the length and strength of growth periods. Before 1959Q4, the average period of uninterrupted growth was 4.3 quarters, 6.4 quarters after the break until 1984Q4, and 17.0 quarters of growth afterward. Thus, the reduction in the unemployment rate recovery does not coincide with an increase in the frequency of recessions. The overall average growth rates during the three time periods were 0.91%, 0.87% and 0.65%, and the average growth rates during expansions in the three time periods were 1.53%, 1.21% and 0.78%. These figures also indicate a decrease in quarterly growth rates over all and during recoveries that coincide with the structural breaks. There is no link, however, between GDP and unemployment rates in these summary statistics. The idea of a jobless recovery is that unemployment rates are slow to recover despite GDP growth. If slower recoveries in unemployment rates are due merely to slowdowns in GDP growth, then we do not have a story of jobless recovery.

Calculating changes in employment per billion dollars of GDP growth during each time period allows us to look at the relationship between growth and at least one determining factor in changes in the unemployment rate. The economy grew in real terms by about $962 billion between 1948 and 1959Q4, with an increase in employment of 7.7 million, or 8,000 jobs for every billion dollars of GDP growth. Between 1959Q4 and 1984Q4 the economy observed 10,600 jobs for every billion dollars of growth. From 1984Q4 through 2012Q4 the data show 5300 jobs added for every billion dollars in growth. While no causal relationship can be discerned from this simple exercise, it is quite striking that GDP growth after the structural break in 1984Q4 was about half as effective in increasing the number of employed than before the break.

The prima facie evidence certainly suggests a case for increasingly jobless recoveries in the U.S. However, the basic statistical presentation here does not account for the non-linearity and asymmetry in unemployment rate movements. The higher the unemployment rate above long-run equilibrium, the greater the potential for rapid declines. Moreover, the changes to the simple averages could also be explained if a single outlier recession, the first oil shock being the most obvious candidate, behaved dramatically differently from the long-run. The issue is important because if jobless recoveries are events associated with characteristics specific to certain recessions, that suggests the causes are then linked to those underlying features of specific recessions. On the other hand, if the trend is toward jobless recoveries in all recessions over time, it points to toward broader, systemic changes in the labor market.

Analysis

We first identify structural breaks in the data on recovery periods. We isolate the downward portion of the cyclical movements in the unemployment rate by splitting the sample in order to account for the asymmetry in unemployment dynamics. We split the data into periods of negative changes in the unemployment rate, outflows, and periods of non-negative changes in the unemployment rate, inflows. (3) Below we split the data using changes in GDP as an alternative to changes in the unemployment rate itself and the results hold strongly.

In order to avoid biasing the regression results, we allowed the data to dictate the proper points for our structural breaks. We rely on the Quandt-Andrews (QA) test for an unknown breakpoint. With jobless recovery as the focus of this paper, we apply the QA test to the portion of our VAR model with changes in unemployment rate, DU R, as the dependent variable. We are looking for trend breaks in changes in unemployment rates, controlling for the level of the unemployment rate and GDP growth.

We begin with no structural breaks in the specification and apply the QA test to the entire (whole date range) split sample to find the best place for a single structural break. If the result of the QA test is significant we consider that date as a potential structural break. (4) We then test that date for significance in the single regression equation from our VAR with DU R as the dependent variable. If the structural break is significant in the regression, we then split the data around the break and repeat the test until we can no longer find statistically significant structural breaks. For the outflows subsample we find two structural breaks, both significant at the 1% level. We find one structural break in the fourth quarter of 1959 and one in the fourth quarter of 1984, both indicating increasingly jobless recoveries. The results are robust to using Bai and Perron (1998) sequential testing, allowing error distributions to differ across breaks. (5)

The core approach has its roots in Evans (1989) who describes U.S. unemployment dynamics through the use of VARs. Evans' paper addresses the behavior of the labor market from 1950 to 1985, asserting a structural break in 1974, which was chosen because it had the smallest standard error in his regressions. Evans' specification incorporated the unemployment rate, changes in growth and a structural break. Evans found a positive and significant coefficient for his structural break variable as well as for a time trend variable.

Evans' model is not adequate to study changes in the speed of recovery because it does not discern between the phases of the business cycle (Moosa et al. 2004). We analyze our times series data using VARs and, unlike Evans two variable system, we use both levels and differences of unemployment rates in our VAR with real changes in GDR Because our focus is jobless recovery, using differences in the unemployment rate is required. However, the unemployment level is an important factor in the size of the difference in unemployment rates. Larger declines in unemployment rates are more likely early in a recovery.

Below are the main specifications for our VAR regressions in this paper for both structural breaks and a time trend.

Structural Breaks

DU [R.sup.OF.sub.t] = [alpha] + [[beta].sub.1]U [R.sub.t-1] + [[beta].sub.2] DU [R.sub.t-2] + [[beta].sub.3][DGDP.sub.t-1] + [[beta].sub.4]SB[594.sub.t] + [[beta].sub.5]SB[844.sub.t], + [[epsilon].sub.t]; (1)

U [R.sup.OF.sub.t] = [theta] + [[gamma].sub.1]U [R.sub.t-2] + [[gamma].sub.3]DU [R.sub.t-2] + [[gamma].sub.3][DGDP.sub.t-1] + [[gamma].sub.4]SB[594.sub.t] + [[gamma].sub.5]SB[844.sub.t], + [[epsilon].sub.t]; (2)

[DGDP.sup.OF.sub.t] = [psi] + [[phi].sub.1]U [R.sub.t-1] + [[phi].sub.2]DU [R.sub.t-2] + [[phi].sub.3][DGDP.sub.t-1] + [[phi].sub.4]SB[594.sub.t] + [[phi].sub.5]SB[844.sub.t], + [[epsilon].sub.t]; (3)

Time Trend

DU [R.sup.OF.sub.t] = [alpha] + [[beta].sub.1]U [R.sub.t-1] + [[beta].sub.2]DU [R.sub.t-2] + [[beta].sub.3][DGDP.sub.t- 1] + [[beta].sub.4][QUARTER.sub.t] + [[epsilon].sub.t]; (4)

U [R.sup.OF.sub.t] = [theta] + [[gamma].sub.1]U [R.sub.t-1] + [[gamma].sub.2]DU [R.sub.t-2] + [[gamma].sub.3][DGDP.sub.t-1] + [[gamma].sub.4][QUARTER.sub.t] + [[epsilon].sub.t]; (5)

[DGDP.sup.OF.sub.t] = [psi] + [[phi].sub.1]U [R.sub.t-1] + [[phi].sub.2]DU [R.sub.t-2] + [[phi].sub.3][DGDP.sub.t-1] + [[phi].sub.4][QUARTER.sub.t] + [[epsilon].sub.t]; (6)

DU [R.sup.OF] is the change in the unemployment rate, U [R.sup.OF] is the unemployment rate and [DGDP.sup.Of] is logged differences in GDP for the outflows subsample. S5594, S[beta]844 and QU ART E R represent the structural break in the fourth quarter of 1959, the structural break in the fourth quarter of 1984, and the quarterly time trend. (6)

Table 1 presents the results for our main specifications. While coefficients on the lagged variables are biased, we see the expected signs on our coefficients. U[R.sub.t-1] represents a one-period lag of the unemployment rate, and has the expected negative sign. The higher the unemployment rate, the more likely there will be a negative change in the unemployment rate. [DGDP.sub.t-1] represents a one-period lag of log differences in real GDR and has the expected negative sign on the coefficient. Stronger GDP growth increases the likelihood of declining unemployment rates. DU [R.sub.t-2] represents a two-period lag of the difference in the unemployment rate, and has the expected positive sign. The larger the change in the unemployment rate from two quarters past, the larger will be this period's change in the unemployment rate.

Most notable are the results for the structural break and time trend variables. We find positive signs with high levels of statistical significance in the VARs for unemployment rates. Therefore, controlling for changes in GDP and the level of unemployment, during periods of outflows changes in the unemployment rate have become smaller in absolute value over time. It follows that recovery from elevated levels of unemployment following a recession will be slower as time goes on. All of the above provide statistically significant evidence supporting the trends observed in the previous section. (7) As a robustness check, we also split the data into two samples using changes in real GDP. We repeat the same exercise explained above using the QA tests to find unknown breakpoints for our expansion subsample based on changes in GDP rather than unemployment rates. We still find two structural breaks, however, in slightly different quarters. The first break in 1960Q1 is significant at the 1% level, but the second break in 1981Q3 is only significant at the 10% level. Even though the second structural break was weaker, the fact that the unknown structural breaks are found in such close proximity to the previous regressions adds support to the idea that something occurred in these time frames causing slower recovery of unemployment after a recession.

Table 2 presents the results. If we use the 1959 and 1984 structural breaks in this expansion subsample, both breaks are significant at the 1% threshold with the same signs as the outflows subsample regardless of the dependent variable. The findings are nearly the same as the outflows sample. (8)

Robustness

The first robustness test uses a more traditional approach. It includes only changes in the unemployment rates and GDP growth rates, but not levels. This specification therefore ignores the potential for non-linearity in the rate of recovery, specifically the convex decline in the unemployment rate following a recession. Again we repeat our exercise of finding unknown breakpoints using this specification.

We present a specification with four lags, but the results are robust from one to eight lags. We still find two breaks, but the specific quarters change slightly. The first break point moves from the first to the second quarter of 1960. However, when disregarding the level of the unemployment rate and focusing only on changes in the unemployment rate, the second structural break shifts to the first quarter of 1995. We put more weight on the main specification, however, due the importance of capturing the convexity of recoveries in unemployment rates as previously discussed. The 1990s were generally a period of lower unemployment rates, and one would expect smaller changes in the unemployment rate, controlling for GDP growth. (9) Table 3 shows the results are indeed robust to the more conventional time series specification. On both structural breaks and the time trend there is still a positive and significant coefficient indicating that negatives changes in the unemployment rate have become smaller over time.

We also want to know if these results are still significant using alternative measures of GDP growth. We replace GDP growth with estimated quarterly total factor productivity (TFP) growth rates from the San Francisco Federal Reserve Bank. Convincingly, the structural breaks are in the exact same place for the outflows subsample as they were when using GDP growth. Table 4 shows the results are indeed robust to this VAR specification, and are in fact stronger. Increases in productivity result in successively smaller changes in the unemployment rate at the structural breaks. Fernald (2014) calculates a number of different productivity measures along with this TFP, which he defines as output growth less the contribution of labor and capital. We also test the VAR using Fernald's measure of utilization of capital and labor and utilization-adjusted TFP. The results are robust to both specifications and support the conclusion that the U.S. is indeed experiencing jobless recoveries when compared to historical data.

Forecasting

Here we use the main specification to illustrate the economic impact of the results. We perform an out-of-sample forecast and repeat the process of finding unknown breakpoints with QA tests, but with the data cutoff before 2000. We then use the actual growth rates to predict the recovery period following the 2001 recession, which resulted in a peak unemployment rate of 6.3% in June 2003.

Figure 1 shows the predicted values of the forecast model are very close to the actual values. In fact, the average forecast error for the recovery is 0.08%. In order to understand the scale of the impact of slowing recovery times, Fig. 2 uses the whole data range for the outflows subsample. This forecast starts at the 7.85% unemployment rate at the end of 2012. Using an uninterrupted annual growth rate of 2.0%, the difference in the time for the unemployment rate to return to the historical long-run average of about 5.5% before and after the structural break in 1959Q4 is about four quarters. (10) The difference in the time for the unemployment rate to return to a level of about 5.5% before and after the structural break in 1984Q4 is eight quarters. The model predicts that before the structural break in 1959Q4 the unemployment rate would return to 5.5% in the first quarter of 2014. The model predicts that it will now take until the first quarter of 2016, a two year difference. (11) The implications for longer unemployment recovery range from political with voting implications to various cost implications in terms of lost tax revenues, extended unemployment benefits, decreases in consumption and countless other examples. It is essential to find causes and policy prescriptions for this negative trend. This paper contributes to that process.

Potential Contributing Factors

This section discusses and empirically assesses potential explanations for jobless recovery. First, we investigate whether the changes in the relationship between unemployment rates and GDP, represented by our structural breaks, are driven by changes in employment or changes in unemployment. While both changes in unemployment and changes in employment are affected by hiring and separation rates, employment is less affected by fluctuations in discouraged workers and labor force participation. We may be able to garner some insight into the possible contributing factors of jobless recovery if we know which is driving the structural breaks. We then investigate changes in: 1) participation rates; 2) the percentage of women in the workforce; 3) industry composition; and 4) social benefits as possible contributing factors. While we have no formal statistical test for comparing our regressions, we calculate the percentage of the structural breaks that can be explained by the different factors for our specifications. We then look to see if the changes are outside of the standard errors for the structural breaks and by how much. (12)

Unemployment and Employment

Like Merz (1999), Shimer (2012) and Hall (2005) challenged the conventional view of the dynamics of the unemployment rate with relatively new empirical evidence. Traditional descriptions of unemployment dynamics begin with a negative shock causing mass layoffs (or separations) that increase the number of unemployed. In turn, job-finding rates decline and the duration of unemployment rises. However, the Job Openings and Labor Turnover Survey (JOLTS) greatly advanced our knowledge of separations. Hall and Shimer, backed by this empirical evidence, argue that separations play much less of a role in the unemployment dynamics than previously believed. They find that unemployment is high, at least since 2000, during a recession due to firms reducing their hiring rates rather than increasing separation rates. While there are changes in separation rates that accompany recessions, they are insignificant compared to regular, aggregate worker flows out of jobs.

JOLTS data show that the increase in layoffs and separations in 2007 are much smaller than the drop in both job openings and hirings. Furthermore, job openings and hiring rates respond much sooner to the economic downturn than layoffs and separations. The results from Fujita and Ramey (2009) dampen the findings of Hall and Shimer with their statistical analysis showing that between 40% and 50% of fluctuations in the unemployment rate are due to changes in separation rates. However that still leaves between 50% and 60% of the fluctuations in unemployment rates to the job hiring rate. Without JOLTS data going back further than 2000, however, it is not possible to directly test whether jobless recoveries are due to changes in separation rates or hiring rates. We can test whether our results are driven by changes in unemployment or changes in employment. We replace changes in unemployment rates from our main specification for the outflows subsample with log differences in unemployment and log differences in employment.

Table 5 suggests the main results in this paper are due to both changes in unemployment, DUR, and changes in employment, DE, controlling for GDP. Both structural breaks are significant for log differences in unemployment while only the 1984 structural break is significant for log difference in employment. The results for both, however, show contributions to jobless recovery. The positive coefficients on the structural breaks for log differences in unemployment mean that the structural breaks are correlated with less negative decreases in unemployment during outflows, controlling for the unemployment rate and for log differences in real GDP. The negative coefficient for the 1984 structural break with log differences in employment means that the structural break is correlated with less positive increases in employment during outflows, controlling for the unemployment rate and for log differences in real GDP.

Participation Rates and Women in the Workforce

Figure 3 shows appreciable changes in the labor participation rate over the time period we are investigating. The labor participation rate rose from 59% in 1948 to a peak of 67% in the late 1990s. Most of this increase comes from women entering the labor market. In this time frame the participation rate for women rose from 32% to 60%, while the participation rate for men fell from 86% to 75%. In a search and matching setting (Mortensen and Pissarides 1994), events that affect participation margins will affect the unemployment function. As people become discouraged and stop looking for work or find encouragement and enter the labor force to look for work, the participation rate will vary. As longer periods of unemployment persist, workers become discouraged and give up searching, which takes them off the unemployment roles and accelerates the recovery of unemployment rates. Working in the other direction, recovery from high rates of unemployment will be slowed as the number of participants increase. Changing societal norms and making it more common for women to enter the workforce could have resulted in non-working spouses joining the labor force to replace lost wages during periods of high unemployment, resulting in slower recoveries in unemployment rates. Peretto (2006) specifically models the effect of participation margins on the unemployment rate. The paper investigates the effects of product and labor market frictions in a dynamic general equilibrium model with a three-states representation of the labor market: 1) Working or employed, 2) Looking for work and cannot find it or unemployed, and 3) Not participating in the labor market (maybe discouraged or content). One of the conclusions from Peretto's model is that participation margins amplify the effects of labor market frictions generating unemployment. With increased participation margins, there will be higher numbers of people who cannot find work.

To test the explanatory power of changes in participation rates on jobless recovery, we calculate the percentage of the structural breaks explained by changes in the participation rate for our VAR specifications including changes in unemployment and changes in employment. We rerun both VARs, but we also include differences in the participation rate, DPR, as a fourth endogenous variable.

We find that changes in the participation rate have little explanatory power for our structural breaks indicating increasingly jobless recovery either through changes in unemployment or changes in employment. This finding does not mean that changes in participation rates do not affect the speed of recovery in unemployment rates, but rather that while controlling for the level of the unemployment rates and GDP growth, changes in participation rates cannot explain much of the increasingly jobless recoveries during the time periods indicated by our structural breaks.

Next, we look to see if the increase of women in the workforce may have explanatory power for our structural breaks. In our sample, the percentage of the labor force made up from women increased by 18.63 percentage points, from 28.28% to 46.91%. We rerun both VAR's in the "Unemployment and Employment" section, but we also include differences in the percentage of women making up the labor force, DPLFW, as a fourth endogenous variable. The results in Table 6, show that the increase in the percentage of women may have some explanatory power for our structural breaks indicating increasingly jobless recovery through changes in unemployment. With the inclusion of DPLFW, the 1959 structural break for changes in unemployment goes down by 0.005, or about 9.43%, and the 1984 structural break goes down by 0.011, or about 18.03%. For changes in employment, including the changes in the percentage of women in the labor force makes the structural break in 1984 even more negative by about 33.33%. This result indicates that without the increases in the percentage of women in the labor force, changes in employment would have become even smaller than before 1984. However, only the change to the structural break in 1984 for log differences in unemployment is actually outside the standard error for the original structural break in Table 5. The evidence is mixed as to whether the percentage of women in the labor force is correlated with jobless recovery, depending on the time period and whether we include changes in employment or changes in unemployment.

Changing Industry Composition of the Labor Force

While Stock and Watson (2003) find at most a small contribution from sectoral shifts in the labor force to the overall decline in macroeconomic volatility seen in the Great Moderation, a transition to a service-oriented economy may play a role in jobless recovery. Looking at Table 7, since 1948 the industry makeup of the U.S. changed dramatically. The U.S. moved away from being a goods-producing, manufacturing-heavy economy toward a service-providing economy. The efficiency wage model (Shapiro and Stiglitz 1984) explains how this shift may be contributing to jobless recovery. In this model, involuntary unemployment stems from employers not being able to observe worker efforts. MacLeod and Malcomson (1998) look at the labor market conditions in which efficiency wages versus merit pay will be endogenously determined as optimal. They find that in highly capital intensive labor markets such as goods-producing and especially manufacturing sectors, the cost of vacancy is high and therefore efficiency wages will be optimal compared to merit pay. Efficiency wages will keep workers in their posts. An example of merit pay is a per item payment for goods produced or quantities harvested. Using merit pay in the service sector, however, is difficult because of firms' inability to verify and quantify productivity. Often times in the service sector, firms depend solely on measures of quality. Quality is much more difficult to measure than quantity. The difficulty in verifying and quantifying productivity in the service sector makes shirking easier and more likely than in goods-producing industries. Firms in the service sector will choose to pay higher wages to prevent shirking.

Following the logic of MacLeod and Malcomson (1998) this shift toward a service-dominated labor force may be contributing to jobless recovery in two ways. First, as demand increases after a recession, capital intensive industries hire quickly to keep costly capital from sitting idle. Bui with the large majority of the U.S. labor economy moving away from a manufacturing and goods-producing workforce, we may be seeing firms stretch employee productivity. Looking at Fig. 4 we see the goods-producing sector is much more responsive to both the initial recession and the beginning stages of recovery. Second, if efficiency wages are being paid in the service sector, the higher marginal cost of labor will drive down employment.

We look to see how much of our structural breaks can be explained by this transition by rerunning both VAR's in the "Unemployment and Employment" section, but now also including differences in the percentage of those employed in the service industry, D PC ESS, as a fourth endogenous variable. (13) The results in Table 8 show that the increase in the percentage of employment in the service industry may have some explanatory power for our structural breaks indicating increasingly jobless recovery through changes in unemployment. With the inclusion of D PC ESS, the 1959 structural break for changes in unemployment goes down by 0.003, or about 5.66%, and the 1984 structural break goes down by 0.008, or about 13.11%. However, for changes in employment, including the changes in the percentage of employment in the service industry makes the structural break in 1984 more negative by about 33.33%. This result indicates that without the increases in the percentage of employment in the service industry, changes in employment would have become even smaller than before 1984. Again, the evidence is mixed as to whether the percentage of employment in the service industry contributes to jobless recovery, depending on the time period and if we include changes in employment or changes in unemployment. (14) Our results are in line with the Burger and Schwartz (2015) findings of the role of reallocative shocks in the labor market for jobless recovery. They find while there is no single underlying factor for jobless recovery, it is at least due in part to real-locative structural change and tends to be more pronounced in recessions that result in extensive rather than intensive decreases in labor inputs. Panovska (2017) also finds evidence for the structural change hypothesis as firms treat hours and employment as substitutes rather than complements post-1984.

Social Benefits, Reservation Wages and Search Intensity

Hiring in a search and matching setting will also be affected by changes on the supply side. The increases in social benefits in the U.S. over time may have led to a negative income effect on labor. That effect could result in a decrease in job search intensity through an increase in reservation wages. Using data from the BEA, total government spending on social benefits to persons as a percentage of GDP rose from 4.6% in 1960 to 9.7% in 1980 to 15.4% in 2010. Government assistance increased in a number of areas including healthcare, housing and food. From the U.S. Census Bureau, for example, the percentage of people on Medicaid increased from 8.4% in 1987 to 15.9% in 2010, and the percentage increase in people on food stamps outpaced the population growth, 34.4, to 21.1%, from 1990 to 2008. From the Social Security Administration (SSA), the number of people receiving Supplemental Security Income from the federal government increased by 26.6% from just 2003 to 2010, while the population only increased by 6.2%. Also, real earnings at the end of 2012 were below the real earnings level in 1979. If wages have not sufficiently kept up with the increases in social benefits, then a decrease in search intensity and an increase in reservation wages could be contributors to jobless recovery.

In order to test the explanatory power on our structural breaks of social benefits in the context of reservation wages, we construct a variable representing the ratio of total U.S. wages and salaries divided by total U.S. social benefits paid. (15) Looking at Fig. 5, this ratio declined dramatically over our sample period. (16) In the mid 1950s there was about $ 18 in wages and salaries earned for every dollar paid out in social benefits. This ratio declined rapidly to about $5 in wages and salaries earned for every dollar paid out in social benefits in the mid 1970s, and then steadily declined to around $3 in wages and salaries earned for every dollar paid out in social benefits in 2012.

We look to see what percentage of our structural breaks can be explained by changes in this ratio by again rerunning both VARs in the "Unemployment and Employment" section, but also including log differences of the ratio, DWSSBP, as a fourth endogenous variable. Our results in Table 9 indicate that changes in this ratio may have some explanatory power for our structural breaks indicating increasingly jobless recovery through changes in unemployment. With the inclusion of DWSSBP, the 1959 structural break for changes in unemployment goes down by 0.006, or about 11.32%, and the 1984 structural break goes down by 0.004, or about 6.56%. These changes are not outside the standard error for the structural breaks found in Table 5, but the results indicate the decrease in wages and salaries earned per social benefit paid out may play a small role in the observed increase in the joblessness of recoveries over time.

Combined Effect

We take this exercise one step further and include all of the variables with possible explanatory power that we have examined in our VAR specifications for log differences in unemployment and log differences in employment. The results in Table 10 show the structural break for log differences in unemployment of 0.042 in 1959 and 0.045 in 1984, which is a decline of 0.011 (20.75%) and 0.016 (26.23%) from the structural breaks found in Table 5. While adding these variables of interest to our VAR specifications individually does not always result in a change to the structural breaks outside of the standard errors found in Table 5, the additive effects on the structural breaks in 1959 and 1984 are outside of the standard errors by 37.50% and 100%. (17)

From these results, we can say that while much of the explanation for the increase in jobless recovery in the U.S. is yet to be found, we can explain some of the story with the factors we have presented here. It seems there are many possible contributors to the increase in the joblessness of recoveries in the U.S., each playing some small role. Complicating the matter further for some factors, the effects on jobless recovery depend on the time frame investigated.

Conclusions

Using VAR models and controlling for log differences in GDP, the unemployment rate and changes in the unemployment rate we show that the U.S. entered into an era of jobless recoveries. We find structural breaks for changes in the unemployment rate in the outflows subsample in the fourth quarters of 1959 and 1984 using the Quandt-Andrews test for unknown breaks. Our structural breaks support the view that the Great Moderation is part of a trend that goes as far back as the 1950s (Blanchard and Simon 2001). While some, including Lazear and Spletzer (2012), find no evidence of structural change in the labor market, our evidence supports Groshen and Potter (2003) and Chinn et al. (2014) that jobless recovery is at least due in part to structural change. When we incorporate structural breaks into our VARs, the coefficients with changes in the unemployment rate as the dependent variable are statistically significant, positive and successively larger. This result indicates that controlling for the unemployment rate and GDP growth, the U.S. is experiencing weaker recovery in unemployment rates. Using the 7.85% unemployment rate at the end of 2012, recovery back to the historical long-run average of 5.5% unemployment rate after both breaks will take eight quarters at 2% GDP growth.

We use the timing of the structural breaks to review the possible causes and related theory, including industry composition, participation rates, and social benefits. Our results indicate the combined effect of all of these factors explain up to about one-fifth of our 1959 structural break and up to about a quarter of our 1984 structural break for changes in unemployment. The increased percentage of women in the workforce and the increased percentage of employment in the service sector play the most prominent roles in explaining the structural break for changes in unemployment. Thus, the evidence in this paper shows that no single explanation accounts for the observed increase in the joblessness of recoveries in the U.S., but it provides empirical clues as to which of the factors are most worthy of deeper future research.

References

Aaronson, D., Rissman, E.R., Sullivan. D.G. (2004). Assessing the jobless recovery. Journal of Economic Perspectives, Q2, 2-21.

Andrews, D. (2003). Tests for parameter instability and structural change with unknown change point: a corrigendum. Econometrica, 77(1), 395-397.

Bai, J., & Perron, P. (1998). Estimating and testing linear models with multiple structural changes. Econometrica, 66(1), 47-78.

Blanchard, O., & Simon, J. (2001). The long and large decline in U.S. output volatility. Brookings Papers on Economic Activity, 32(1), 135-164.

Burger, J.D., & Schwartz, J.S. (2015). Productive recessions and jobless recoveries. Contemporary Econ Policy, 33, 636-648. https://doi.Org/10.1111/coep.12096.

Chinn, M., Ferrara, L., Mignon, V. (2014). Explaining US employment growth after the great recession: the role of output-employment non-linearities. Journal of Macroeconomics, 42, 118-129. ISSN 0164-0704.

Davis, S., Haltiwanger, J., Schuh, S. (2006). Job creation and destruction. Cambridge: MIT Press.

Evans, G. (1989). Output and unemployment dynamics in the united states: 1950-1985. Journal of Applied Econometrics, 4(3), 213-237.

Faberman, J. (2017). Job flows, jobless recovery, and the great moderation. Journal of Economic Dynamics & Control, 76, 152-170.

Federal Reserve Bank of San Francisco (2014). http://www.frbsf.org/economic-research/indicators-data/ total-factor-productivity-tfp/. Total factor productivity. Accessed: February.

Fernald, J. (2014). A quarterly, utilization-adjusted series on total factor profuctivity. Federal Reserve Bank of San Francisco Working Paper, http://www.frbsf.org/economic-research/files/wpl2-19bk.pdf.

Fujita, S., & Ramey, G. (2009). The cyclicality of separation and job finding rates. International Economic Review, 50(2), 415-430.

Groshen, E.L., & Potter, S. (2003). Has structural change contributed to a jobless recovery? Federal Reserve Bank of New York Current Issues in Economics and Finance, 9(8), 1-7.

Hall, R.E. (2005). Job loss, job finding, and unemployment in the U.S. Economy over the past fifty years. NBER working paper, #11678. http://www.nber.org/papers/wll678.pdf.

Jaimovich, N., & Siu, H. (2012). The trend is the cycle: job polarization and jobless recoveries. NBER working paper, #18334. http://www.nber.org/papers/wl8334.pdf.

Kim, C., & Nelson, C. (1999). Has the U.S. Economy become more stable? A Bayesian approach based on a Markov-Switching model of the business cycle. The Review of Economics and Statistics, 81(4), 608-616.

Lazear, E.P., & Spletzer, J.R. (2012). Hiring, churn, and the business cycle. American Economic Review, 702(3), 575-579.

MacLeod, W., & Malcomson, J.M. (1998). Motivation and markets. American Economic Association, 88(3), 388-411.

McConnell, M., & Perez-Quiros, G. (2000). Output fluctuations in the united states: what has changed since the early 1980's? American Economic Review, 90(5), 1464-1476.

Merz, M. (1999). Time series evidence of unemployment flows: the sample period matters. Journal of Business & Economic Statistics, 170), 324-334.

Moosa, I., Silvapulle, M., Silvapulle, P. (2004). Asymmetry in Okun's law. The Canadian Journal of Economics, 37(2), 353-374.

Mortensen, D., & Pissarides, C. (1994). Job creation and job destruction in the theory of unemployment. Review of Economic Studies, 61(3), 397-415.

Panovska, I. (2017). What explains the recent jobless recoveries? Macroeconomic Dynamics, 27(3), 708732.

Peretto, P. (2006). The employment (and output) of nations: theory and policy implications. Duke University Working Paper, #16. https://papers.ssrn.com/sol3/papers.cfm?abstractJd=1270645.

Shapiro, C., & Stiglitz, J. (1984). Equilibrium unemployment as a worker discipline device. American Economic Review, 74(3), 433-444.

Shinier, R. (2012). Reassessing the ins and outs of unemployment. Review of Economic Dynamics, 75, 127-148.

Stock, J., & Watson, M. (2003). Has the business cycle changed? evidence and explanations. Monetary policy and uncertainty: adapting to a changing economy. In Proceedings of symposium sponsored by federal reserve bank of Kansas City, Jackson Hole, Wyo. (pp. 9-56). https://hec.unil.ch/docs/files/80/599/stock- watson03.pdf.

U.S. Bureau of Economic Analysis (2014). https://www.bea.gov 1948-2012, U.S. Economic accounts and income and personal income accounts. Accessed: February, 2014.

U.S. Bureau of Labor Statistics (2014). https://www.bls.gov. 1948-2012 data. Unemployment and employment tables and current employment statistics. Accessed: February, 2014.

(1) http://www.npr.org/templates/transcript/transcript.php?storyId=113847257

(2) http://www.investopedia.com/terms/j/jobless-recovery.asp

(3) Merz (1999) defines inflows as those who become unemployed, while outflows are defined as those who leave the state of unemployment. The idea gives insight into the different parts of the unemployment cycle, but still does not address the parts in separate phases of a cycle. Merz finds that outflows have become more strongly countercyclical, lending support to Hall's (2005) finding that the large majority of labor dynamics are due to fluctuations in the hiring rate as opposed to layoffs, but it does not address the phenomenon of a jobless recovery.

(4) For significance of the QA test, we use Andrews (2003) corrected critical values.

(5) For Bai and Perron: The scaled F-statistic for 1 vs. 2 breaks is 22.6. For significance at the 1% threshold the critical value is 22.4 for trimming of 10% and 22.0 for trimming of 15%. The breaks are at the same points found using our method and both are significant to the 1% threshold.

(6) We reject the unit root in all specifications presented to ensure stationary using the Augmented Dickey-Fuller (ADF) test and the Dickey-Fuller Generalized Least Squares (DF-GLS) test.

(7) These results hold with two significant structural breaks for DUR or UR as the dependent variable with up to at least eight lags of DU R and DGDP. These results hold with two significant structural breaks for DGDP as the dependent variable up to three lags of DU R and DGDP and for the 1984 structural break up until at least eight lags.

(8) While this paper is primarily concerned with the recovery portion of the unemployment cycle, we also investigate inflows to unemployment. Performing the same procedures for inflows to unemployment as we did for outflows, we do not find a significant structural break through the QA test. Thus, we do not present VAR results.

(9) The results are robust to using the 1959Q4 and 1984Q4 structural breaks.

(10) 2.0% is the average annualized quarterly growth rate during the most recent recovery up until 2012.

(11) Unemployment rates actually reached 5.5% in the first quarter of 2015, but GDP grew at an annual rate of about 2.8% over this time with four quarters over 3.5% and three quarters over 4.5%. The participation rate declined over that time period from 63.7 to 62.7%, which is the lowest it has been since 1977.

(12) For brevity, we do not show tables for all of our results, but they are available upon request.

(13) These data comes from the Current Employment Statistics (CES) survey from the BLS.

(14) We also ran a test including both DPCESS and DPLFW in our VAR specifications. Our results indicate that little or none of the increase in the magnitude of the structural break, with the sole inclusion of changes in the percentage of women in the labor force, to our VAR can be explained due to a correlation with changes in the percent of employment in the service industry.

(15) These data come from the BEA National Income and Personal Income accounts.

(16) The large decline and rapid increase in the early 1950's dealt with payments to soldiers following World War II.

(17) The result for the 1984 structural break for log differences in employment when we include all the variables is -0.004, which is not outside the standard error.

https://doi.org/10.1007/s11293-018-9569-7

Published online: 5 March 2018

James DeNicco [1] * Christopher A. Laincz [2]

[mail] James DeNicco

jimmydenicco@gmail.com

[1] Department of Economics, Rice University, 6100 Main St, Houston, TX 77005, USA

[2] Department of Economics, Drexel University, 3141 Chestnut St, Philadelphia, PA 19104, USA

Caption: Fig. 1 Out-of-sample forecast results for recovery from the 2001 recession. Source: Own calculations using data from the Bureau of Labor Statistics (2014) and the Bureau of Economic Analysis (2014)

Caption: Fig. 2 Unemployment rates forecast results from 2012: Outflows. Source: Own calculations using data from the Bureau of Labor Statistics (2014) and the Bureau of Economic Analysis (2014)

Caption: Fig. 3 Monthly U.S. participation rate (1948-2012). Source: Bureau of Labor Statistics (2014)

Caption: Fig. 4 Industry unemployment rates (2000-2012): 6 month rolling averages. Source: Own calculations using data from the Bureau of Labor Statistics (2014)

Caption: Fig. 5 Wages and salaries earned per dollar in social benefits paid from (1948-2012). Source: Own calculations using data from the Bureau of Economic Analysis (2014)
Table 1 VARs for outflows subsample

VAR              1

Specification    Structural breaks

Variable         DUR          UR          DGDP

C                -0.009       -0.009      0.012 ***
                 (0.062)      (0.062)     (0.003)
UR(-1)           -0.068 ***   0.932 ***   0.001 *
                 (0.009)      (0.009)     (0.0004)
DUR(-2)          0.105 ***    0.105 ***   -0.001
                 (0.037)      (0.037)     (0.002)
DGDP(-1)         -4.287 **    -4.287 **   0.093
                 (1.784)      (1.784)     (0.089)
SB594            0.231 ***    0.231 ***   -0.004 **
                 (0.043)      (0.043)     (0.002)
SB844            0.297 ***    0.297 ***   -0.009 ***
                 (0.044)      (0.044)     (0.002)
QUARTER

[R.sup.2]        0.435        0.990       0.160
Adj. [R.sup.2]   0.415        0.990       0.130
F-Stat           21.40 ***    2734 ***    5.300 ***
LL               59.601       59.601      494.89
DW               2.434        2.434       1.862

VAR              2

Specification    Time trend

Variable         DUR          UR           DGDP

C                0.080        0.080        0.011 ***
                 (0.063)      (0.063)      (0.003)
UR(-1)           -0.067 ***   0.933 ***    0.001 **
                 (0.010)      (0.010)      (0.0004)
DUR(-2)          0.088 **     0.088 **     -0.001
                 (0.039)      (0.039)      (0.002)
DGDP(-1)         -5.228 ***   -5.228 ***   0.088
                 (1.914)      (1.914)      (0.088)
SB594

SB844

QUARTER          0.001 ***    0.001 ***    -4.5E-05 ***
                 (0.0002)     (0.0002)     (1.0E-05)
[R.sup.2]        0.349        0.988        0.162
Adj. [R.sup.2]   0.330        0.988        0.138
F-Stat           18.74 ***    2982 ***     6.776 ***
LL               49.303       49.30        495.072
DW               2.145        2.145        1.855

n = 145, ***,**,* denote significance at the 1%, 5%,
and 10% threshold respectively. () contains standard errors.
Data: 1948 to 2012 quarterly data on unemployment rate (UR)
and change in unemployment rate (DUR), Source: Bureau of
Labor Statistics (2014). Change in real chained quarterly
GDP (DGDP), Bureau of Economic Analysis (2014)

Table 2 VARs for expansion subsample

VAR               1

Specification     Structural breaks

Variable          DUR           UR            DGDP

C                 0.239 ***     0.239 ***     0.009 ***
                  (0.066)       (0.066)       (0.002)
UR(-1)            -0.057 ***    0.943 ***     0.001 ***
                  (0.011)       (0.011)       (0.0003)
DUR(-2)           0.120 ***     0.120 ***     0.001
                  (0.043)       (0.043)       (0.001)
DGDP(-1)          -14.257 ***   -14.257 ***   0.215 ***
                  (1.886)       (1.886)       (0.055)
SB601             0.159 ***     0.159 ***     -0.004**
                  (0.049)       (0.049)       (0.001)
SB813             0.164 ***     0.164 ***     -0.007 ***
                  (0.049)       (0.049)       (0.001)
QUARTER
[R.sup.2]         0.360         0.981         0.203
Adj. [R.sup.2]    0.345         0.980         0.184
F-Stat            23.71 ***     2149 ***      10.738 ***
LL                12.666        12.666        780.47
DW                2.128         2.128         2.005

VAR               2

Specification     Time trend

Variable          DUR           UR            DGDP

C                 0.280 ***     0.280 ***     0.010 ***
                  (0.065)       (0.065)       (0.002)
UR(-1)            -0.054 ***    0.946 ***     0.0008 ***
                  (0.011)       (0.011)       (0.0003)
DUR(-2)           0.118 ***     0.118 ***     0.001
                  (0.043)       (0.043)       (0.001)
DGDP(-1)          -13.954 ***   -13.954 ***   0.198 ***
                  (1.924)       (1.924)       (0.054)
SB601

SB813

QUARTER           0.0006 **     0.0006 **     -4.0E-05 ***
                  (0.0002)      (0.0002)      (6.7E-06)
[R.sup.2]         0.338         0.980         0.223
Adj. [R.sup.2]    0.326         0.980         0.210
F-Stat            27.068 ***    2610 ***      15.174 ***
LL                9.059         9.059         783.19
DW                2.052         2.052         2.016

n = 217, ***,**,* denote significance at the 1%, 5%,
and 10% threshold respectively. () contains standard errors.
Data: 1948 to 2012 quarterly data on unemployment rate (UR)
and change in unemployment rate (DUR), Source: Bureau of Labor
Statistics (2014). Change in real chained quarterly GDP (DGDP),
Bureau of Economic Analysis (2014)

Table 3 VARs for outflows subsample
without unemployment rate levels

VAR               1

Specification     Structural breaks

Variable          DUR         DGDP

C                 -0.411***   0.015 ***
                  (0.061)     (0.003)
DUR(-1)           0.081       -0.009 ***
                  (0.060)     (0.003)
DUR(-2)           0.141**     -0.001
                  (0.054)     (0.002)
DUR(-3)           0.097       0.001
                  (0.060)     (0.003)
DUR(-4)           -0.189***   0.006 ***
                  (0.050)     (0.002)
DGDP(-1)          -1.122      -0.106
                  (2.238)     (0.100)
DGDP(-2)          2.726       -0.056
                  (2.239)     (0.100)
DGDP(-3)          4.964 **    -0.066
                  (2.375)     (0.106)
DGDP(-4)          2.344       0.082
                  (2.327)     (0.104)
SB602             0.104 **    -0.002
                  (0.042)     (0.002)
SB951             0.214 ***   -0.007 ***
                  (0.048)     (0.002)
QUARTER
[R.sup.2]         0.391       0.270
Adj. [R.sup.2]    0.346       0.216
F-Stat            8.618 ***   4.963 ***
LL                54.216      505.085
DW                2.313       1.783

VAR               2

Specification     Time trend

Variable          DUR          DGDP

C                 -0.413 ***   0.021 ***
                  (0.073)      (0.003)
DUR(-1)           0.090        -0.011***
                  (0.063)      (0.003)
DUR(-2)           0.151 **     -0.003
                  (0.058)      (0.002)
DUR(-3)           0.104        -0.001
                  (0.063)      (0.003)
DUR(-4)           -0.182 ***   0.006 ***
                  (0.052)      (0.002)
DGDP(-1)          -1.028      -0.199 **
                  (2.396)      (0.100)
DGDP(-2)          3.055        -0.140
                  (2.407)      (0.101)
DGDP(-3)          5.611 **     -0.132
                  (2.486)      (0.104)
DGDP(-4)          3.008        0.028
                  (2.422)      (0.101)
SB602
SB951
QUARTER           0.001 ***    -4.8E-05 ***
                  (0.000)      (1.0E-05)
[R.sup.2]         0.347        0.312
Adj. [R.sup.2]    0.303        0.266
F-Stat            7.971 ***    6.800 ***
LL                49.111       509.348
DW                2.200        1.775

n = 145, ***,**,* denote significance at the 1%, 5%,
and 10% threshold respectively. () contains standard
errors. Data: 1948 to 2012 quarterly data on change in
unemployment rate (DUR), Source: Bureau of Labor Statistics
(2014). Change in real chained quarterly GDP (DGDP),
Bureau of Economic Analysis (2014)

Table 4 VARs for outflows subsample
with total factor productivity

VAR               1

Specification     Structural breaks

Variable          DUR          UR          DTFP

C                 -0.060       -0.060      2.478 **
                  (0.058)      (0.058)     (1.096)
UR(-1)            -0.067 ***   0.933 ***   0.310 **
                  (0.009)      (0.009)     (0.167)
DUR(-2)           0.130 ***    0.130***    0.963
                  (0.035)      (0.035)     (0.661)
DTFP(-1)          -0.005       -0.005      -0.037
                  (0.005)      (0.005)     (0.085)
SB594             0.240 ***    0.240 ***   -1.806 **
                  (0.044)      (0.044)     (0.832)
SB844             0.316 ***    0.316 ***   -3.071 ***
                  (0.045)      (0.045)     (0.846)
QUARTER
[R.sup.2]         0.417        0.990       0.132
Adj. [R.sup.2]    0.396        0.989       0.101
F-Stat            19.92 ***    2650 ***    4.224 ***
LL                57.377       57.377      -367.83
DW                2.343        2.343       1.852

VAR               2

Specification     Time trend

Variable          DUR          UR          DTFP

C                 0.027        0.027       1.894 **
                  (0.059)      (0.059)     (1.041)
UR(-1)            -0.066 ***   0.934 ***   0.348 **
                  (0.010)      (0.010)     (0.171)
DUR(-2)           0.117 ***    0.117 ***   1.024
                  (0.038)      (0.038)     (0.664)
DTFP(-1)          -0.009 *     -0.009 *    -0.018
                  (0.005)      (0.005)     (0.085)
SB594
SB844
QUARTER           0.001 ***    0.001 ***   -0.013 ***
                  (0.0002)     (0.0002)    (0.004)
[R.sup.2]         0.330        0.988       0.116
Adj. [R.sup.2]    0.311        0.988       0.091
F-Stat            17.24 ***    2898 ***    4.582 ***
LL                47.254       47.254      -369.16
DW                2.076        2.076       1.849

n = 145, ***,**,* denote significance at the 1%, 5%,
and 10% threshold respectively. () contains standard errors.
Data: 1948 to 2012 quarterly data on unemployment rate (UR)
and change in unemployment rate (DUR), Source: Bureau of
Labor Statistics (2014). Change in Total Factor
Productivity (DTFP), Federal Reserve Bank of
San Francisco (2014)

Table 5 Outflow VARs: Log differences
in employment and unemployment

VAR        1

Variable   DU           UR          DGDP

C          -0.062 ***   -0.096      0.018 ***
           (0.015)      (0.079)     (0.004)
SB594      0.053 ***    0.197 ***   -0.002
           (0.008)      (0.043)     (0.002)
SB844      0.061 ***    0.260 ***   -0.007 ***
           (0.008)      (0.043)     (0.002)

VAR        2

Variable   DE           UR          DGDP

C          0.005 **     -0.129      0.018 ***
           (0.003)      (0.086)     (0.004)
SB594      0.0002       0.209 ***   -0.0032
           (0.001)      (0.047)     (0.002)
SB844      -0.003 **    0.297 ***   -0.008 ***
           (0.001)      (0.045)     (0.002)

n = 145, ***,**,* denote significance at the 1%, 5%,
and 10% threshold respectively. () contains standard errors.
Data: 1948 to 2012 quarterly data on unemployment rate (UR),
change in unemployment rate (DUR), and change in employment
(DE) Source: Bureau of Labor Statistics (2014). Change in
real chained quarterly GDP (DGDP), Bureau of Economic
Analysis (2014)

Table 6 Outflow VARs: Percentage of women in the labor force

VAR        1

Variable   LDU          UR          DGDP         DPLFW

C          -0.052 ***   -0.073      0.016 ***    -0.023
           (0.016)      (0.083)     (0.004)      (0.097)
SB594      0.048 ***    0.190 ***   -0.002       -0.028
           (0.009)      (0.045)     (0.002)      (0.053)
SB844      0.050 ***    0.240 ***   -0.006 **    -0.131 **
           (0.010)      (0.050)     (0.002)      (0.059)

VAR        2

Variable   LDE          UR          DGDP         DPLFW

C          0.006 **     -0.121      0.017 ***    0.000
           (0.003)      (0.087)     (0.004)      (0.098)
SB594      0.000        0.186 ***   -0.003       -0.047
           (0.001)      (0.048)     (0.002)      (0.054)
SB844      -0.004 **    0.267 ***   -0.007 ***   -0.146 **
           (0.002)      (0.051)     (0.002)      (0.057)

n = 145, ***,**,* denote significance at the 1%, 5%,
and 10% threshold respectively. () contains standard errors.
Data: 1948 to 2012 quarterly data on unemployment rate (UR),
change in unemployment rate (DUR), and change in percentage
of women in the labor force (DPLFW) Source: Bureau of Labor
Statistics (2014). Change in real chained quarterly GDP
(DGDP), Bureau of Economic Analysis (2014)

Table 7 Percent industry composition

Year                      1950     1960     1970     1980

Goods-Producing          37.34    35.91    31.93    27.48
  Mining & Logging        2.02     1.44     0.97     1.15
  Construction            5.09     5.57     5.08     5.09
  Manufacturing          30.23    28.90    25.89    21.24
Service-providing        49.03    48.78    50.51    54.68
  Financial Activities    4.12     4.60     4.90     5.46
  Profes. & Bus.          6.54     6.76     7.38     8.22
  Education & Health      4.84     5.33     6.37     7.64
  Leisure & Hosp.         6.24     6.32     6.68     7.39
Other                     1.89     2.09     2.49     2.98
Government               13.63    15.31    17.56    17.84

Year                      1990     2000     2010

Goods-Producing          21.97    18.84    13.79
  Mining & Logging        0.70     0.45     0.53
  Construction            4.97     5.16     4.35
  Manufacturing          16.31    13.22     8.92
Service-providing        61.40    65.43    68.86
  Financial Activities    6.04     5.87     5.89
  Profes. & Bus.          9.87    12.55    12.74
  Education & Health      9.84    11.44    14.95
  Leisure & Hosp.         8.51     8.96    10.03
Other                     3.87     3.93     4.10
Government               16.63    15.73    17.34

Source: Bureau of Labor Statistics (2014)

Table 8 Outflow VARs: Percentage of
employment in the service industry

VAR        1

Variable   LDU          UR          DGDP         DPCESS

C          -0.058 ***   -0.081      0.016 ***    -0.045
           (0.015)      (0.08)      (0.004)      (0.054)
SB594      0.050 ***    0.194 ***   -0.001       0.076 **
           (0.009)      (0.047)     (0.002)      (0.032)
SB844      0.053 ***    0.246 ***   -0.006 **    0.111 ***
           (0.010)      (0.050)     (0.002)      (0.034)

VAR        2

Variable   LDE          UR          DGDP         DPCESS

C          0.005 **     -0.117      0.016 ***    -0.059
           (0.002)      (0.086)     (0.004)      (0.058)
SB594      -0.001       0.215 ***   -0.003       0.066 *
           (0.001)      (0.051)     (0.002)      (0.034)
SB844      -0.004 ***   0.299 ***   -0.007 ***   0.112 ***
           (0.002)      (0.052)     (0.003)      (0.035)

n = 145, ***,**,* denote significance at the 1%, 5%,
and 10% threshold respectively. () contains standard errors.
Data: 1948 to 2012 quarterly data on unemployment rate (UR),
change in unemployment rate (DUR), change in percentage employed
in service industry (DPCESS), Source: Bureau of Labor Statistics
(2014). Change in real chained quarterly GDP (DGDP), Bureau of
Economic Analysis (2014)

Table 9 Outflow VARs: Wage and salaries earned
per dollar in social benefits paid

VAR        1

Variable   DU           UR          DGDP         DWSSBP

C          -0.070 ***   -0.114      0.017 ***    -0.081 ***
           (0.014)      (0.075)     (0.004)      (0.028)
SB594      0.047 ***    0.177 ***   -0.002       -0.027 *
           (0.008)      (0.041)     (0.002)      (0.015)
SB844      0.057 ***    0.246 ***   -0.007 ***   -0.010
           (0.008)      (0.041)     (0.002)      (0.015)

VAR        2

Variable   DE           UR          DGDP         DWSSBP

C          0.004        -0.107      0.016 ***    -0.084 ***
           (0.002)      (0.080)     (0.004)      (0.029)
SB594      0.000        0.203 ***   -0.004 *     -0.021
           (0.001)      (0.044)     (0.002)      (0.016)
SB844      -0.003 **    0.279 ***   -0.008 ***   -0.002
           (0.001)      (0.042)     (0.002)      (0.015)

n = 145, ***,**,* denote significance at the 1%,5%,
and 10% threshold respectively. () contains standard
errors. Data: 1948 to 2012 quarterly data on unemployment
rate (UR) and change in unemployment rate (DUR), Source:
Bureau of Labor Statistics (2014). Change in real chained
quarterly GDP (DGDP) and change in ratio of total US wages
and salaries divided by total social benefits paid (DWSSBP).
Bureau of Economic Analysis (2014)

Table 10 Outflow VARs: Combined effect

VAR        1

Variable   LDU          UR           DGDP

C          -0.058 ***   -0.093       0.014 ***
           (0.015)      (0.080)      (0.004)
SB594      0.042 ***    0.177 ***    -0.001
           (0.009)      (0.047)      (0.002)
SB 844     0.045 ***    0.234 ***    -0.005 *
           (0.010)      (0.055)      (0.003)

VAR        2

Variable   LDE          UR           DGDP

C          0.003        -0.242 ***   0.013 ***
           (0.003)      (0.088)      (0.005)
SB594      -0.002       0.139 ***    -0.003
           (0.002)      (0.050)      (0.003)
SB 844     -0.004 ***   0.265 ***    -0.006 **
           (0.002)      (0.050)      (0.003)

n = 145, ***,**,* denote significance at the 1%, 5%,
and 10% threshold respectively. () contains standard errors.
Data: 1948 to 2012 quarterly data on unemployment rate (UR)
and change in unemployment rate (DUR), Source: Bureau of Labor
Statistics (2014). Change in real chained quarterly GDP (DGDP),
Source: Bureau of Economic Analysis (2014). Not shown in
table--regressions also include change in employment (DE),
percentage of women in the labor force (DPLFW), change in
percentage employed in service industry (DPCESS),
Source: Bureau of Labor Statistics (2014), and change
in ratio of total U.S. wages and salaries divided by
total social benefits paid (DWSSBP), Bureau of
Economic Analysis (2014)
COPYRIGHT 2018 Atlantic Economic Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2018 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:DeNicco, James; Laincz, Christopher A.
Publication:Atlantic Economic Journal
Article Type:Report
Geographic Code:1USA
Date:Mar 1, 2018
Words:10498
Previous Article:Best Article Award: Fleissig.
Next Article:Product Reliability, R&D, and Manufacturing Cost Shocks.
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters