Over the past several months the labor market has been sending out some mixed signals in terms of its relative strength. For instance, initial unemployment claims data—typically a reasonable signal of the overall labor market—has been trending downward and is not far off from a new 35-year low. Once adjusted for the size of the current population, the September figure of the seasonally adjusted 4-week moving average of 305,000 initial claims would “traditionally” be associated with a very strong labor market. In October the number of claims have edged upwards, but still remain fairly low.
Recently several economists & bloggers—Marginal Revolution, The Money Allusion, Calculated Risk, and others—have noted the disconnect between the recent unemployment claim numbers and the rest of the labor market—job creation, the unemployment rate, and wage growth. Some have taken it as a sign that the labor market is not really as weak as the headline unemployment rate suggests, while others see it as a symptom of the decline of manufacturing, which frequently relies on mass layoffs to adjust output. At its lowest in early 2000 there were 0.94 claims per 1,000 people in the country. The only other periods to approach this—1989 and in 2006—the same metric was 1.16 and .96 respectively. As of the end of September, the current rate was .96 claims per 1,000 citizens—given a U.S. population estimate of 316.4 million in 2013.
However, masked in these national numbers is the marked difference in claims data between the states. Similar to our analysis that showed great disparity in job gains by state, the claims data shows a very different picture when examined at the state level. While some states are at or near record lows in terms of the level of unemployment claims per 1,000 residents, other states are still well above the number of monthly unemployment claims that would signal a return to a healthy labor market.
In contrast to the past two recoveries where a vast majority of states had also seen their claim data nadir in close correlation with the national job market, this recovery has seen several states lag well behind the leaders in terms of layoff declines. One simple measure is to see how the average state was doing in terms of its claims per 1,000 residents when the national economy experienced a new low in initial claims. The elevated figure for this simple average among the 50 states and the District of Columbia in 2013 also features a higher weighted variance for these claims per 1,000 people currently among the states as compared to the corresponding periods in 2000 and 2006.
Another way to examine this is look at the 52-week moving average of claims by state to see in each of these periods—1989, 2000, 2006, and 2013—how close each state was to its lowest ever ratio of claims per 1,000 residents. For instance in 2000 and 2006, on average when the nation claims rate bottomed out in terms of initial claims the average state was within 14 and 17 percentage points respectively above its historic low of number of claims per 1,000 residents. In contrast in 2013, as the nation approaches its historic low for initial claims per 1,000 residents, on average the states are 26 percentage points above their previous historic low.
Essentially many states approached their historic lows in 2000 and in 2006, but much fewer states are close to their historic lows currently despite the healthy national numbers in terms of initial claims. Another way to think about it would be to posit that the economic expansions in the late 1990s and mid 2000s were rising tides that spread business activity relatively broadly across most states and MSAs. Whereas today’s expansion is a more of story of states that have and have not, where several states are beginning to forge new records for employment and many other states lag significantly behind in terms of their labor market.
While it is possible that the average of initial claims per 1,000 will narrow across many more states as the nation actually reaches new 35-year low of claims per 1,000 citizens, it is unlikely that the states which have the highest current claims rate will make enough progress in the next several months to fundamentally change the shape of this distribution. This would require large drop of in claims in Alaska, Wisconsin, Pennsylvania, Oregon, and New Jersey all of which have more been averaging more than 1.5 claims per 1,000 residents for the past year. Furthermore, we find little preliminary evidence lately that the level of initial unemployment claims are associated with the size of the manufacturing sector—as gauged by the location quotient, but we will explore this relationship more fully in subsequent blogs.
Written by Patrick Clapp and Chris Chmura.
The impact of shutting down the U.S. government is much larger than the 800,000 federal workers who are staying at home and not spending money on commuting and lunch. Even though Congress has agreed to pay these workers, the reduction and shutdown of certain services is rippling through the economy.
The Wall Street Journal cites an estimate from J.P Morgan Chase economists that “each week of a shutdown would reduce the annualized pace of fourth-quarter economic growth by 0.12 percentage point.” However, the forecast only looks at effects on government workers, not the private sector or consumer confidence.
One question missing from the various lists of FAQs about the shutdown is whether any private sector industries could be affected. Fewer people are making retail purchases, going out to eat, enjoying arts or entertainment, and/or traveling and staying at hotels when tourist attractions like the Grand Canyon or Smithsonian museums are closed. They might also delay big purchases, like a new house, even if they could get a loan from an understaffed Federal Housing Administration. That gives us four North American Industrial Classification System (NAICS) sectors for closer inspection: retail trade (NAICS code 44), real estate and rental and leasing (53), arts, entertainment and recreation (71), and accommodation and food services (72).
Next, we need to look at a comparable period. The last time the government shut down was between December 16, 1995 and January 6, 1996, a total of 21 days. We consider the effects on our selected sectors around that period for metropolitan statistical areas (MSAs). We also consider the percentage of federal employees in the area during that time period—areas with a higher percentage of federal workers implies those areas would have more furloughed workers, and likely larger impacts on local industries. Data are obtained from Chmura’s database powering the EQSuite®.
For each of the four sector categories in every MSA, we first calculated the year-over-year percent change in employment by month and wages by quarter to filter out some seasonal fluctuations. The graphs later in this post shows the percentage point changes in growth rates month to month or quarter to quarter relative to the number of people employed by the federal government in each region as a percent of total regional employment. With less demand during the shutdown, we would expect firms in these industries to cut back employees’ hours, which would show as a decrease in the growth rate of wages. To a lesser extent, they might cut their staff size, which would appear as a negative growth rate in employment.
Now for some results: The clearest story is in the retail sector, where there was a strong negative relationship between the percentage of federal employees in a region and changes in the growth rate of wages quarter to quarter in the time periods surrounding the last government shutdown. Employment in the retail sector also declined, but not as dramatically as wages, implying that businesses cut hours, rather than laying off workers to cope with the government shutdown.
The Washington, D.C. metro area ranked fifth of 353 MSAs in the percentage of federal employees in the workforce. The year-over-year growth rate in wages in the retail sector slowed 0.6 percentage points between the last quarter of 1995 and the first quarter of 1996. In the metro area with the largest percentage of federal workers, Warner Robins, Georgia, the year-over-year growth rate of retail wages fell 4.5 percentage points over that time.
Tourist areas such as national parks and museums also suffered. The year-over-year growth rate of retail wages in the Flagstaff, Arizona MSA, near the Grand Canyon, fell 1.7 percentage points between the last quarter of 1995 and the first quarter of 1996.
The effects of furloughed workers rippled through the arts, entertainment, and recreation sector as well. As federal employees were unsure if they would get paid, metro areas with a large percentage of federal employees experienced sharper drops in wages for this sector over the shutdown period and a similar drop in employment. In the Jacksonville, North Carolina MSA, home of Marine Corps Base Camp Lejeune, the year-over-year employment growth rate in the arts, entertainment, and recreation sector fell over 16 percentage points between December and January.
The year-over-year growth rate of employment and wages in the real estate sector and accommodation and food services sector were not negatively affected, suggesting they are more resistant to uncertainty caused by a government shutdown. This makes intuitive sense for the real estate sector in particular—if the only thing keeping someone from purchasing a house is delayed paperwork at the FHA, they are still going to buy a house, just somewhat later than expected.
Warner Robins, GA MSA
Bremerton-Silverdale, WA MSA
Washington-Arlington-Alexandria, DC-VA-MD-WV MSA
Jacksonville, NC MSA
Lawton, OK MSA
Fairbanks, AK MSA
Fort Walton Beach-Crestview-Destin, FL MSA
Huntsville, AL MSA
Anniston-Oxford, AL MSA
Clarksville, TN-KY MSA
Anchorage, AK MSA
Cheyenne, WY MSA
Honolulu, HI MSA
Las Cruces, NM MSA
Lebanon, PA MSA
Bakersfield, CA MSA
Texarkana, TX-Texarkana, AR MSA
* Data for the real estate sector of the Washington, DC metro area is unavailable prior to 2001
Similar to the last shutdown, Congress said it will pay the furloughed employees even though they did not work, and employment and wages in these industries quickly bounced back—and we would expect the same when the current shutdown ends. Continuing to look at the retail sector, year-over-year growth in wages in the quarter following the shutdown was positively correlated with the percentage of federal workers in a region, implying renewed demand from now-working government employees rippled out to more hours in the retail sector.
However, that quick return to growth took place in a strong mid-1990s economy. In the midst of our current slow recovery, a similarly long shutdown could be a much larger and longer setback for workers in both the public and private sectors.
The above chart shows the MSAs most likely to be impacted by the federal shutdown based on the most recent data available. The size of each bubble represents total MSA employment, and the color of a bubble indicates the political party affiliations of the senators representing that state. MSAs closer to the top right of the chart have higher unemployment rates and a larger percentage of federal workers in their local workforce, and are expected to be hit harder by the shutdown.
Chmura Economics & Analytics recently presented its second annual Workforce Boardroom Conference. The conference theme was Illuminate to Innovate, and it brought thought leaders from a variety of disciplines together to discuss smart solutions to some big challenges facing our regions, states, and nation.
The event kicked off with a presentation from Marge Connelly, who brought her unique insight and private sector perspective on customer-centered innovation from her career in financial services and higher education. Following are some other highlights from the event.
Federal Issues, Sequestration and Your Economy
Chmura has been researching defense spending, the sequestration, and their impacts on state and regional economies. The team has developed one of the most complete datasets on federal defense spending ever constructed, and Chris Chmura’s remarks were designed to identify, inform, and inspire defense-dependent communities to action. Data-Driven Decision Making
Jay Dougherty, partner at Mercer and co-founder of the Workforce Sciences Institute, shared his candid experience with companies making location decisions today. He emphasized the use of quantitative data analysis to optimize the location, labor, cost of doing business, and business climate of a corporate decision.
Chmura team member Dan Meges joined Bruce Stephen of Monster Government Solutions to talk about workforce planning and the use of Real-Time Labor Intelligence. Monster, which started the online job posting revolution, is now manipulating its massive database of candidates and openings to help create a clearer picture of labor supply and demand. This perspective is invaluable for communities and higher education working to better align their program offerings with market trends.
Chmura’s chief statistician, Greg Chmura, took the stage to share some interesting analysis on the connection between education, innovation, and job growth. His research built on findings of the Fund for Our Economic Future in northeastern Ohio. The research is surprising - innovation and job growth don’t necessarily go hand in hand.
Measuring What Matters – New Paradigms for Economic Development
One of the goals of the conference was to challenge traditional thinking about economic development and Ed Burghard of the Burghard Group’s talk on the American Dream Composite Index (ADCI) did not disappoint. Ed suggested the audience push away from traditional outcome measures such as job creation and capital investment and instead focus on the well-being and satisfaction of the populace.
Ed has partnered with Xavier University to advance the ADCI, which uses household survey data to measure the pulse of the American sentiment. It’s a holistic measure that includes five components, including economic, a well-being, societal, diversity, and environmental indexes.
Four Business Perspectives
The conference brought four unique and valuable business perspectives to the forefront. The first was from a return presentation by David B. Trebing, General Manager, State and Local Relations for Daimler. David shared the new reality for manufacturers operating in a competitive, dynamic, global marketplace. This reality includes the need to locate production facilities in growing and emerging markets, a push towards cost reduction and efficiency, and the need for a qualified workforce.
The second business topic was presented by Carlos Solari, VP at Wilkitech. Carlos’s remarks enlightened the group on emerging trends in the cyber security industry and what communities need to know to attract and retain businesses in that sector.
Jeff Harris with Xerox Services spoke on the relationship between innovation, people, and the environment, and what that means for business and communities today. His futuristic remarks about the shifting trends in the definitions of work and the workplace prompted healthy discussions over lunch about how economic developers will adapt.
The conference closed with some practical public relations strategies from Hope Katz Bibbs, founder and president of The Inkandescent Group, LLC. Hope has helped many entrepreneurs and early-stage companies transform their innovative ideas into growth, and she shared her ideas about what it takes to help a new enterprise launch successfully and get the attention it deserves.
It was an exciting two days in Richmond and Chmura thanks the engaged attendees and speakers. Now, we’re busy planning for 2014. If you would like to learn more about the 2014 conference, contact Maggie Bishop at email@example.com.
Read Older Posts
Applied Economist is the official Chmura Economics & Analytics Weblog, focusing on select topics related to the Virginia and national economy.
Use the links below to read one of our staff blogs or submit a question.
Have a question for one of our bloggers that you would like to see answered?
The opinions expressed by the bloggers on this site and those providing comments are
theirs alone, and do not reflect the opinions of Chmura Economics & Analytics
or any employee thereof. Chmura Economics & Analytics is not responsible for
the accuracy of any of the information supplied by the bloggers on this site.