September 28, 2006
Michigan automotive and white collar jobs
Loss of market share from the traditional Big Three automakers to global competitors has impacted Michigan’s economy, leading to some deep concerns about its future. To date, most attention to this issue has focussed on job loss related to automotive production activity. Auto assembly and parts production continues at a strong (though eroding) clip in the United States, but it is rapidly shifting away from Michigan. So far, the “new domestic” carmakers have avoided siting new production plants in Michigan, preferring to site them in the South, as well as in Ohio and Indiana, such as Honda’s recent announcement to build a plant in Greensburg, Indiana. However, another important employment component for Michigan also relates to the health and sales market share of the Big Three—that is, the nonproduction activities of these auto assembly companies. These activities include research and development (R&D), sales, finance, and management operations, which form an outsized economic engine for the state. In what ways does the survival (and growth) of Big Three companies go hand in hand with the nonproduction jobs located in Michigan?
Nonproduction employment of auto assembly companies typically amounts to a surprising 35%–45% of total employment and an even larger share of payroll. While Michigan is highly concentrated in automotive production—with 15 auto assembly plants—it is also the domicile of the Big Three's headquarters along with significant company R&D and other operations. For this reason, it is not surprising in Michigan to find that nonproduction automotive employment is more concentrated than elsewhere. In counting Big Three nonproduction employment at their production plants, headquarters, R&D centers, and other auxiliary facilities in Michigan, nonproduction employment likely outnumbers production employment, making up a minimum of 55%–60% of total Big Three jobs in the state.
Moreover, additional Michigan personal income and jobs are generated from local services purchased by headquarters-type operations. As Chicago Fed economist Yukako Ono has found in recent studies, headquarters operations often purchase key services for the entire company network. These purchases may include financial services, R&D, information technology (IT) products and services, strategic management consulting, and many more. From the regional economy’s standpoint, these purchases are often sourced locally to a large extent. In fact, Ono discusses the possibility that the choice of location by headquarters may be influenced by the cost and availability of such business services.
Similar behavior of automotive headquarters makes Detroit and its surrounding environs much more than just a factory economy. Specifically, much of the value of Big Three automobiles derives from product development and design, and most of that R&D activity is conducted in Michigan. As derived demand from the domestic automotive industry, key business services are largely produced in Detroit. My blog entry from August 16 shows that the Detroit metropolitan area far and away tops other midwestern metropolitan areas in its concentration of professional and technical services employment. Among Detroit’s top sectors are engineering services (employment at 51,594 jobs in 2002) and scientific research and development (18,126 jobs in 2002).
Nationally, much R&D is funded and performed by automotive companies and their affiliates. According to the most recent survey of industry funds for research and development, which is conducted by the National Science Foundation, the automotive industry accounts for $14–$15 billion in annual R&D funding in the U.S. To be sure, in recent years, as auto assemblers have increasingly relied on their first-tier suppliers for entire components and automotive modules, some significant R&D responsibilities have been shifting away from assembly companies and toward automotive parts companies. Still, today, the lion’s share of this R&D is performed in-house, that is, largely by auto assembly companies themselves.
These practices have kept Ford, General Motors (GM), and Daimler-Chrysler among the largest R&D performers in the U.S., with Michigan at the hub of such activity. For this reason, Michigan ranks second only to California in funds for industrial R&D. And for 2003 as the figure below shows, the motor vehicle assembly and parts industries in Michigan accounted for $10.7 billion of the $15.2 billion industry-performed R&D in the state. The ties between these expenditures and local employment is apparent. According to a parallel survey by the National Science Foundation, the Detroit metropolitan area employed 102,500 research scientists and engineers in 2003—a concentration of 5.2% of the work force as compared to 3.9% nationally.
Would Michigan retain this important function in the event that Big Three sales shares continued to decline? On the positive side, there are some indications that the Detroit area’s role in automotive research is in the process of growing beyond its historic roots. For example, the “new domestic” automakers have all sited research, development, and design facilities in the Detroit region, such as Toyota’s recently announced $150 million R&D center investment in Ann Arbor. Others, such as Hyundai and Nissan, have also recently expanded their facilities or announced plans for similar expansions.
So, too, Detroit’s attractiveness to automotive company headquarters operations displays some sparks of growth. Major automotive parts producer Borg Warner moved its headquarters from Chicago to the Detroit area last year. More generally, Chicago Fed economist Thomas Klier has documented an upswing in auto parts company headquarters moving to Michigan. The presence and growth of automotive parts headquarters in Michigan probably bodes well for company-sponsored R&D activity as well.
Still, competitive challenges are at play both here and abroad. Domestically, figures from the U.S. Bureau of Economic Analysis show that the annual R&D funding in the U.S. by Asia-domiciled automotive companies, at $125 million, makes up a very small share of automotive R&D in the U.S., amounting to less than 2 percent. And while the Detroit metropolitan area has so far attracted many of these transplant R&D activities, historically, it is not uncommon to find that attendant service activities eventually follow production in manufacturing. In this direction, the movement of U.S. automotive production from the Midwest toward the South is drawing the attention of those seeking R&D activities as well. For example, Clemson University in South Carolina has launched a research program and industrial park to foster technology development and transfer in cooperation with companies such as BMW and others.
And so, Michigan has several important economic activities at stake amidst the current upheaval among automotive companies.
September 20, 2006
Midwest housing market update
Following unprecedented home price appreciation nationwide in recent years, homeowners are much concerned about price reversal. In their current Economic Perspectives article, Chicago Fed economists Jonas Fisher and Saad Quayyum find that, on average, much of the recent surge in housing can be attributed to fundamentals such as rising income and favorable demographics, as well as innovations in home lending markets that have allowed renters to become homeowners. (Many of these innovations—such as interest-only loans and adjustable rate mortgages—were discussed in detail at the Chicago Fed's Bank Structure Conference this spring. The proceedings of the conference were summarized in the September issue of the Chicago Fed Letter.)
While such arguments may provide some comfort to those who worry about the possibility of a bubble in average U.S. home prices, experiences and current conditions differ widely from place to place. Should Midwestern homeowners be more or less concerned about the cooling of residential real estate markets?
Senior Business Economist Mike Munley has been tracking home price developments in the Midwest. Mike reports that, on September 5, the Office of Federal Housing Enterprise Oversight (OFHEO) released its estimates for home price appreciation in the second quarter of 2006. The report included data on the national average of home price changes as well as state averages.
Home prices for the U.S. increased at a 4.8% annual rate between the first and second quarters, the slowest quarterly appreciation since the end of 1999 and just below the average since 1980. As measured year over year, U.S. home prices were up 10% from the second quarter of 2005, which was also slower than the rate of appreciation has been—it topped out at 14% in the middle of 2005.
Recent home price appreciation here in the Midwest has also slowed noticeably, and the long term back drop has been much less robust. For the most part, home prices in the Seventh District states have been increasing more slowly than the national average of home prices (see figure 1). On a year-over-year basis, price appreciation in every District state lagged behind the national average in the second quarter of 2006, and Michigan had the lowest appreciation of any state in the nation. In comparison to the first quarter, home values in Indiana and Michigan actually declined. (Maine, Massachusetts, and Ohio were the only other states with declines.) However, home values in Iowa managed to rise slightly faster than the national average.
The city-level data told a similar story. Of the District MSAs (Metropolitan Statistical Areas) covered by OFHEO, only Michigan City-La Porte, IN, showed year-over-year appreciation (10.6%) faster than the national average. Of the bottom 20 MSAs in the U.S., 14 were in the Seventh District, and Ann Arbor, MI, was at the bottom with home prices, down 1.3% from a year ago.
The OFHEO home price data is only one of several sources of information about home prices for the U.S. and some cities. The National Association of Realtors (NAR) releases data on the median sale price of existing single-family homes. In general, the two data series tend to tell the same story—that is, the trends in both data series are similar over time. But, their results are often different in a given month (for regional and national data) or quarter (for city data). The NAR data tends to be more volatile. The NAR data set measures exactly what it sounds like: it is the price of the typical home sold during that quarter. Still, the median price depends on the mix of homes sold during that quarter. If, for example, a large number of inexpensive, starter homes were sold in the second quarter, this would lower the median sale price. By contrast, the OFHEO index is designed to track how the value of an individual home changes over time. OFHEO looks at the appraised value of homes each time a new mortgage is taken out—it is updated when a home gets sold or when the homeowner decides to refinance. OFHEO looks at the value of a large number of homes and is able to estimate the index quarterly. One drawback to the OFHEO index is that it only looks at home mortgages serviced by Fannie Mae and Freddie Mac, and those agencies only service mortgages that are less than $417,000—so the OFHEO index excludes most luxury housing.
The NAR publishes price breakdowns for regions and select metropolitan areas (but not states). In the second quarter of 2006, median home prices nationally were up 3.7% from a year earlier, while median sales prices in the Midwest (which includes the Seventh District, Ohio, and some plains states) were down 2.0%. Of the 24 District MSAs covered by the NAR, five (Chicago, Champaign, Milwaukee, Peoria, and Waterloo, IA) beat the national average, and 14 saw home sale prices down from a year ago. Although the NAR data are more volatile, this data series does confirm that home prices in the Midwest have been increasing more slowly than prices nationally.
There are a couple of reasons why home values have been rising more slowly in the Midwest than the rest of the country. Looking over the long term, the Midwest has generally seen slower home price appreciation since the early 1980s. As shown in figure 1, home values nationally have increased an average of 5.4% per year since then, whereas average appreciation in the District states has ranged from 3.5% in Iowa to 5.1% in Illinois. In part this difference reflects the slower population growth in the Midwest than in the rest of the nation. Since 1980, the U.S. population has increased an average of 1.1% per year, while population growth in Seventh District states has averaged only 0.4%. It follows that demand for housing in the District is not growing as rapidly, which in turn puts relatively less pressure on prices.
Regional differences in home prices also arise from the supply side of the market. In many metropolitan areas, available land for home building is limited by natural barriers such as mountains and waterways. In the face of rising demand for housing, such barriers to expanded supply tend to drive up land and home prices, especially for single-family homes. Further, some areas have chosen to place legal restrictions on home building by imposing growth boundaries or strict zoning requirements and building codes. In some cases where regulations are not well-crafted, evidence suggests that the effect is the same; rising demand for homes is met by rapidly rising prices rather than by expansion of the housing stock. A recent survey report of land use regulations verifies that the Midwest is not especially noted for manmade barriers to housing expansion. Over the long term, the elastic nature of home building in the Midwest region has likely contributed to less pressure on home prices.
More recently, much of the relative weakness in home prices can further be explained by the relatively sluggish economic growth in the region. As I discussed in my recent Mid-Year Jobs Report, job growth here has lagged behind national job growth. That limits income growth in the Midwest, which in turn restricts demand for housing. So while the U.S. has seen a sharp rise in home price appreciation in the past several years, the run-up in the District was less extensive or non-existent. (See figure 2.)
Among states, home price appreciation has recently been running in direct relation to the pace of economic growth. In particular, appreciation has been lowest in Michigan and Indiana, the two states with economies weighed down by structural change in automotive industries.
The figure below illustrates home price appreciation over the past year among metropolitan areas. Those metro areas experiencing depreciation tend to be found in Michigan and in central Indiana. A look back at the metropolitan area map of auto industry job concentration in figure 1 of last week’s blog shows a fairly close correlation between auto-intensity and weak home price appreciation.
The relative stability of home prices in some Midwest locations is a double-edged sword for the region. Homeowners in other parts of the country were able to cash in on the sharp increases in the value of their homes and use those funds to support their spending. Midwesterners weren't able to cash in as extensively, limiting growth in retail sales locally.
On the plus side, given all of the popular concerns about a home price bubble, steady appreciation helps abate those worries here. If a sharp run-up in prices is a warning sign of a potential bubble, that sign is largely absent in the Midwest. But this is not to say that the Midwest is immune from the risk of a slowdown in appreciation or price declines going forward. Certainly, overall economic conditions will feed into home prices, as they have in parts of Michigan.
Less appreciation in home prices can also be advantageous in that it keeps homes here more affordable. According to the NAR's affordability index, homes have historically been affordable in the Midwest in relation to other regions and recently this affordability advantage has improved. Midwestern cities can use this attribute to help attract new businesses and workers.
September 13, 2006
Where is automotive employment in the Seventh District?
Perhaps the most notable economic development taking place in the Seventh District is the market shift away from the traditional "Big 3" domestic auto makers--General Motors, Ford, and (Daimler)-Chrysler--and their parts suppliers. Lost sales are shifting toward the "new domestics" such as Toyota and Nissan and their parts suppliers. The sales gainers tend to be located outside of the Midwest to a greater degree than the Big 3. This shift is documented and analyzed in a recent Economic Perspectives article by Thomas Klier and Dan McMillen. This market upheaval is tending to idle and displace workers in many Midwest communities. Per Klier and McMillen, Michigan automotive employment is down almost one-third since 1979 while southern states such as Kentucky, Tennessee, Alabama, and the Carolinas have experienced a tripling of jobs.
But despite these shifts, Detroit and much of the Midwest continues to be the center of the production. Which particular communities remain most sensitive to future swings in automotive fortunes? The data below attribute automotive employment to particular metropolitan areas in the Seventh District. Those metropolitan areas with green shading had an employment concentration in automotive that exceeded the nation; those shaded in red had a lesser concentration. Looking across metropolitan areas in the entire Seventh District region, an east-west split in auto employment concentration becomes very apparent. The Michigan-Indiana corridor contains most of the metropolitan areas having an above-average concentration. Darkly-shaded metropolitan areas in southeast Michigan are exceptionally concentrated in automotive. So too, an east-west band of metropolitan areas across north central Indiana is steeped in automotive employment.
A numerical listing of automotive employment below shows just how concentrated some communities can be. Metropolitan areas including Detroit/Livonia/Deaborn, Flint, Holland, Saginaw, Battle Creek, and Lansing/East Lansing in Michigan all reported concentrations over 5 times the national average, as did the Kokomo and Lafayette metro areas in Indiana.
The final table below further illustrates the sharp geographic rift in employment fortunes over the 1990-2005 period. As a whole, the state of Michigan lost over 64,000 jobs in automotive, on net accounting for all job losses nationally. Largely due to the Michigan experience, the Seventh District states experienced an 18 percent decline in automotive jobs since 1990 while the remainder of the U.S. experienced a 3 percent gain in similar employment.
September 6, 2006
What industries are key to Midwest economic performance?
Urban economist Wilbur Thompson once said, “Tell me your industries, and I’ll tell you your future.” A region’s industries do tell us a lot about its economy. In the Midwest, manufacturing industries often drive fluctuations and trends in the region’s overall economic growth because manufacturing is a much larger part of its economy, on average, than the rest of the nation’s. So, too, manufactured goods are traded far and wide—that is, they are exported and imported across national boundaries as well as across regions that make up the U.S. economy. Accordingly, shifts in demand for manufactured goods can have an outsized impact on states and communities in the Midwest. For example, a national shift in buying behavior toward foreign nameplate autos, or toward smaller and more energy efficient autos, may well impact automotive production, investment, and employment in some parts of the Midwest region.
On a short-term basis, fluctuations in aggregate economic activity, such as recessions, diminish demands for durable goods such as capital equipment, thereby making the Midwest economy more sensitive to national “business cycle” fluctuations.
So, too, many Midwest manufacturing industries are impacted by global competitive shifts. Production operations of some home appliance manufacturers have shifted to Mexico, for instance.
But how can we identify which particular industries to observe and follow in the Seventh District? First, we must ascertain how concentrated is an industry in a local economy as compared with the national economy. Analysts often construct a “location quotient” to do so. In one such application, each industry’s employment share of total employment in the region is compared with its national counterpart. The comparison is constructed as a ratio with the local share on top. For example, if a locality’s labor force had 20 percent of its workers in manufacturing as compared with 10 percent nationally, the index (ratio) takes on a value of 2.0, i.e., 20/10. Parity with the nation would take on a value of 1.0.
While such an index is useful by way of comparison, it says little about the actual size of a particular industry in a state or region. For this reason, the chart below identifies manufacturing industries in the Seventh District states by relative concentration and by employment size. The horizontal scale depicts the concentration, and it is centered at the index value of one, or parity with the nation. The vertical scale is centered at the value of the median-sized manufacturing industry in the District (as measured by payroll employment).
By construction then, we may quickly characterize the most prominent industries in the District as they are located in the upper right hand quadrant of the graph. For the District, it is clear that transportation, food processing, and machinery are the most prominent industries, with transportation (representing automotive) winning hands down. The fabricated metal products sector also looms large; however, these industries represent many diverse intermediate products that are eventually used to produce more final goods such as autos or machinery. Primary metals, principally steel foundries as designated by the industry code 331 on the chart, is the most concentrated industry (as measured by employment) in the District. Yet, its employment is relatively small in comparison.
Charts for each individual state will soon be available on our Midwest Regional Website. Iowa is reproduced below. As the chart suggests, employment in food processing stands out as the largest and the most concentrated in the state. In large part, this activity represents Iowa’s further processing of corn and soybeans into meals and oils, as well as its meat packing industry, chiefly pork. Iowa’s large and highly concentrated machinery industry reflects its focus on its manufacturing of farm machinery and equipment.
Analysis of the District’s lesser industries can also be informative. In the overall U.S., the computing and electronic products industries have grown rapidly into a large component of overall U.S. manufacturing. In virtually every Seventh District state, for example, employment in this sector exceeds the median manufacturing sector. But at the same time, the states’ concentration of this sector is universally below the national average. In this instance, the sector’s lower concentration and lesser expansion here have contributed to a slower pace of overall economic growth.
Of course, these glimpses are only a superficial beginning to understanding the structure and behavior the region’s economy. For one, individually identified sectors often have important linkages to others that merit further consideration. Such industries as machinery and autos, for example, purchase great volumes of intermediate materials and parts locally, including those found in rubber and plastics, fabricated metals, and machinery (e.g., tool and die and metal cutting machinery). Also, in varying degrees, sectors may purchase local services as diverse as management consulting and transportation. Specific industry linkages can be found in the input–output tables of the U.S., which are produced by the U.S. Bureau of Economic Analysis (BEA).
However, the U.S. input–output tables may often be misleading for regional analysis. That is because specific inter-sector buying and selling relationships will differ greatly and vary widely from region to region. For one, local firms will purchase intermediate goods and services from many possible places. For the most part, we know little about the varying geography of such relationships. In response, the BEA has adapted and estimated the national relationships for individual regions of the U.S. in its RIMS II modeling system. This system and others like it, which are available commercially, are often used to estimate the broader economic impacts of small changes to a community or local industry.