Office of Planning, Environment, & Realty (HEP)
The contractors who developed these research products were Jack Faucett Associates (lead) and Economic Development Research Group (assisting). Both have done work for FHWA on other occasions. Their web home pages are: www.jfaucett.com and www.edrgroup.com/index.shtml respectively.
Consistent with the objective, this research focused on the economic development history of completed interstates or long portions of such interstates (e.g. I-43, I-81) with similar natures to those designated by Congress (e.g., I-69, I-73/74). These histories would provide some perspective with which to anticipate what would happen subsequent to completion of future interstates of similar natures.
Nine Interstate or near Interstate corridors were examined. The research correlated data, generally county level, on population, employment, income, etc. with the time period before, during and after completion of the Interstate. In some counties, the changes in population, employment, income, etc. were similar to changes in counties without interstates. In some counties, changes were dissimilar and imply that the influence of the Interstate was positive from an economic development standpoint. Some of the dissimilarities are discussed below.
This research was close to a before/after study for a number of mostly rural interstates or near interstates (SR-99 was the near interstate). The histories are not a clean before/after study because of the incremental nature of the construction of highways (it took decades to construct some of the interstates in the study with sections being opened every few years).The research confirms previous FHWA statements that highway improvement, by itself, does not bring economic development but, in some cases, combined with other effects, the highway improvement can lead to such development. However, without additional research, it is impossible to fully understand the mechanism(s) by which positive development occurs.
In a number of counties economic growth was anemic despite proximity to an interstate highway. For example, this was the case in Emanuel, Treutlen, Wilkinson and Twiggs counties in the I-16 corridor (I-16 does not go through Wilkinson county but does go near it). Each of these counties has populations below 11,000 (three counties are below 5,000). For these counties, the advantages of the Interstate (employer access to labor, labor mobility, supplier access to market, etc.) simply may be less significant than other economic conditions (for example the relative decline of the US industrial economy, especially certain segments such as tobacco and textiles, compared with the post industrial economy). In fact, there is some speculation that in portions of these counties the I-16 suppressed economic activity: either by facilitating retail activity outside the area that would otherwise remain in the county; by diverting through traffic from non-interstate highways; or, by acting as a physical barrier to commerce[i].
I-16 was not the only corridor with some counties lagging behind in economic growth. Similar examples occur in the other corridor studies. An interesting example of such a county in the South Carolina I-26 corridor is Orangeburg County in which is the junction between I-26 and I-95. In this county, one of the problems preventing economic development is access to utilities[ii].
On the other hand, there were a number of interesting and positive developments in a number of corridors. For example, in the I-16 corridor, Bryan County, which had a population of about 25,000 in 2002 and which is within the Savannah Metropolitan Statistical Area, had about a 420% increase in jobs[iii] (and a 290% growth in population) between 1969 and 2002 (the increase for the State of Georgia during this time was about 130% in jobs and about 90% in population). Bullock County (a non metropolitan county just outside the Savannah Metropolitan Statistical Area which had about 57,000 in population in 2002) had increases in jobs and population of about 140% and 80% respectively (which exceeds the increase for non metropolitan Georgia of about 70% and 40% respectively). Similar examples occur in the other corridor studies.
Readers of the histories will note a number of other interesting things in the histories. For example:
Between 1990 and 2000 manufacturing employment increased considerably in Brown and Ozaukee Counties (by about 25% and 34% respectively). This was greater than the increase during that time period in Wisconsin (about 10%) and counter to the national trend (about a 7% decline[iv] ). These two counties (which are part of the Green Bay and Milwaukee metropolitan area respectively) also had robust general employment increases between 1970 and 2000 (about 160% and 140% respectively) supported by population increases (about 80% and 115% respectively). We are not certain what contribution I-43 made to the health of the manufacturing sector in these two counties; however, the advocates of building I-43 undoubtedly hoped for something like what actually happened[v] .
Although not disaggregated by county, the Virginia I-81 corridor also has had an about an 18% increase in manufacturing employment between 1971 and 2000[vi] . This was especially welcome since in the southern part of this corridor a large chemical manufacture closed in 1970 contributing to substantial decreases in employment in the period between 1969 and 1971. Distribution related employers have also opened in the corridor but wholesale trade employment is not available by county so the quantitative analysis of this factor is difficult[vii].
In the SR-99 corridor in California, farm employment seems to have expanded dramatically between 1990 and 2000 (by about 25%). It is not clear whether this is due to a statistical quirk (because employment in Agriculture Services is not available these two years). It is also curious that employment in wholesale trade, which would seem to be facilitated by the continuous improvements to SR-99, did not substantially increase in this decade. This may be partially due to the fact that SR-99 is not a new highway and in fact, the improvement has been incremental over a few decades as noted in the report on this corridor.
While many of the less populated counties in the Interstate corridors had minimal economic growth, a county that seems to have had some success is Garrett County, MD. Between 1969 and 2000, employment and population increased by about 170% and 40% respectively compared to same year increases in non-metropolitan Maryland of about 100% and 55% respectively[viii] . Garrett County has a considerable amount of recreational related employment (the recreational resources are more accessible as a result of the Interstate). One indicator of the extent of the recreational component of the Garrett County economy is that the peak unemployment rate in the winter has typically been over twice the minimum rate in the summer over the past decade or so[ix] . In addition, and past the time of analysis in the study, the unemployment rate in Garrett County has, per the Maryland State government, fallen from about 8.5% in 2000 to about 6.6% in 2003 (whereas both the US and MD unemployment rate has risen during that time period[x] ).
Another interesting rural county is Hale County in the Texas I-27 corridor. This county has lost population steadily between 1969 and 2002 (about 11% of its 1969 population which was about 36,000 in 2002). Generally, this is thought to be an effect of the consolidation of agriculture (fewer and bigger farms over time). However, this county had a 30% increase in employment between 1969 and 2002. This was due substantially to the success of three industrial parks near I-27 that have attracted agriculture related manufacturing employers[xi].
A county that is making an economic transition is Woodbury County in the Iowa I-29 corridor. This county was once a major meatpacking center and the destination of livestock shipped from much of the northern Great Plains. A number of employers in this industry disappeared in the 1970s and 1980s. A combination of agricultural consolidation and the closure of livestock yards and other meatpacking related business have resulted in a stagnant population. Between 1969 and 2002 the population increased less than 1%. Job growth was also stagnant between 1969 and 1983 increasing about 3%. However, beginning in about 1983, the region encompassing Sioux City began having some success attracting business. Part of the reason was the presence of I-29; part was due to other highway improvements. The most important non-highway factor was aggressive implementation of financial incentives and this is thought by many to have been more important than the highway improvements. From 1983 to 2000, employment has increased substantially (about 25%) and the increase in housing prices between 1990 and 2000 (about 85%) exceeded the average for both Iowa and the United States (about 80% and 50% respectively)[xii] . However, the story didn't end in 2000 and between 2000 and 2003, the unemployment rate has risen considerably[xiii].
Dublin is the anchor of a micropolitan statistical area (somewhat like a metropolitan area but with a smaller population) that includes Laurens County. During the 1980s and 1990s, the county was hit by the decline of the domestic textile industry. However, beginning in the 1990s, Dublin - about half way between Savannah and Macon - has been the site of a number of logistics related employers, for example in warehousing and distribution. The I-16 report shows that between 1969 and 2002, Laurens County gained about 40% and 100% in population and employment respectively[xiv] . In recent years, Laurens County has done well compared to the country as a whole. The unemployment rate in Laurens County, which, during the summer of 2000 was between 5.5% and 7.5% (unadjusted for seasonal effects), was, during the summer of 2004, between 4.0 and 5.5%[xv].
Some counties have been hurt by the decline of one dominant industry. Schuylkill County, Pennsylvania was hurt by the decline of two dominant industries. This county reached its peak population in 1930 at about 235,000 (current population is about 150,000) In the 1950s, unemployment went over 20% during the decline of the coal industry employment[xvi] and later the area was hit by a decline in manufacturing. In the past few decades, population has declined steadily and total employment has not increased. However, in recent years, some employers have placed sites in the county (e.g., distribution, manufacturing) and two interchanges have been added to support development of industrial parks. The success of these industrial parks will not be known for several years[xvii], but a review of the situation in 2005 did find some encouraging activity.
The objective of this research was to compile data to be used as a baseline and as a point of comparison to future interstates designated by Congress. To this extent, the research was successful. It is difficult to look at the results of this research and agree with those proponents of developing new interstate highways who envision increased employment all along any corridor with an Interstate. Similarly, it is difficult to agree with those opponents of developing new interstate highways who envision increased sprawl and lower income jobs all along any corridor with an Interstate. The results of this research imply that counties with partially successful employment expansion programs will have more successful programs if there is an interstate nearby. For counties where economic development is inhibited by a lack of developable sites or other barriers, the results of this research imply that a new interstate may result in little improvement in the economic development picture. The data from this research was collected and is displayed to assist future efforts to supplement the data. [viii] At this time, however, the FHWA has no plans to pursue data supplementation.
Since the initial posting (March 1, 2005), several people indicated that they felt that others might infer that success in highway economic development was mostly random. While acknowledging that luck is a component of success, the data more easily leads to a conclusion that economic development success is related to the degree of access and connectivity improvement that the highway improvement provides as well as to the nature and strength of the non-highway economic development initiatives. It is also possible that the effects of improvements to highways without access control would not result in quite the effects of improvements to highways with access control. Although as of this time (June 2005), FHWA does not have any funds available to study this possibility, such funds may be available in the future either to FHWA or to some other organization, e.g., NCHRP. One possible general starting approach for the study of economic development effects of highways without access control would be to collect information, similar to that done in this study, for 30 to 150 mile long 2-4 lane non-freeway highway improvements that have been completed in recent years. Any reader who knows of such an improvement that might be studied in this context is invited to e-mail me (see CONTACT information below).
[i] All data on Georgia I-16 is available in the study on that corridor.
[ii] This is discussed in section 4 of the South Carolina I-26 report.
[iii] The FHWA is aware that there are differing ways of measuring employment and unemployment. We have taken care that when comparisons are made they are to numbers generated by the same methodology. Thus a 30% increase in employment, when using one method of estimating employment for all data points should be fairly consistent with the increase in employment if another method of estimating employment was used for all data points.
[v] All data on Wisconsin I-43 is available in the study on that corridor.
[vi] Shown on table 5 of the study on Virginia I-81
[vii] This is discussed in section 4 of the Virginia I-81 report.
[viii] This is discussed in section 3 of the Maryland I-68 report.
[xi] All data on Texas I-27 is available in the study on that corridor.
[xii] All data on I-29 counties up to this point is from the Iowa I-29 study.
[xiv] All data on Laurens County, Georgia up to this point is from the Georgia I-16 study.
[xvii] All Schuylkill County data other than the pre 1969 historic data is from the Pennsylvania I-81 study.
[xviii} above URLs are valid as of February, 2005