Office of Operations Freight Management and Operations

2. Answers Emerging from the Literature Review (Page 2 of 4)

2.3 Has the Relationship between Highway Investment and Industrial Productivity been Discerned at Macro and Micro-level?

2.3.1 Classification of Literature Findings

A summary classification of literature findings related to highway investment and industry productivity is presented in Table 2.

Table 2: Classification of Literature Findings Related to Question B: "Has the Relationship between Highway Investment and Industrial Productivity Been Discerned at Macro and Micro-level?"
empty cell Direct Investigations Indirect Investigations
Number of Citations * 23 [9], [23], [29], [38], [45], [49], [51], [53], [56], [61], [62], [63], [66], [67], [68], [69], [70], [80], [84], [95], [96], [99], [102] 7 [7], [28], [52], [54], [71], [82], [100]
Percent of Total Citations 23 7
Range of Quantitative Findings
  • The marginal industry benefits of highway capital are in the range 0.2 to 0.6 cents per year per $1 increase in highway capital. Somewhat lower results are obtained for Germany and Sweden.
  • A study replicating the analysis of Aschauer regarding the impacts of public investment on private productivity obtains insignificant results for the State of Texas. The same methodology applied to data from several OECD countries and USA finds similar negative results.
  • Cross-section analysis of data from 98 countries showed significant associations between per capita national product and per capita length of paved road network.
  • The positive shadow prices for public capital indicate that infrastructure has a significant role in determining productivity growth.
  • Net return on investment in public infrastructure accruing specifically to the manufacturing sector could be close to zero.
  • Highway projects that appear to generate large amounts of growth may not actually cause that growth or may cause it in part by shifting economic activity from other areas.
Nature of Qualitative Findings
  • An increase in highway capital reduces cost for a given level of output for all industries.
  • The decline in public infrastructure investment has contributed to the slowdown in productivity.
  • Changes in road growth lead to larger changes in productivity growth in vehicle intensive. Industries.
  • The econometric evidence suggests that congestion reduction is productive. The effects of expanding the street and highway stock is more dubious.
  • Several case studies illustrate how an effective highway network plays an important role in private economic activity.
  • Economic rate of return, not number of "jobs created" should be the criterion for infrastructure project selection.
  • The link between transportation investment and private economic performance varies by transportation mode, industry and state.

* Numbers in square brackets give citation number in Appendix 1.

2.3.2 Review of Literature Findings Relationship between Highway Investment and Industrial Productivity

The relationship between highway investment and industrial productivity attracted the attention of many researchers. Research by Aschauer (1991) uncovered a positive relationship between general transportation spending and long-run economic performance. The study finds that an increase in transportation spending of $10 billion for ten years would raise labor productivity by nearly four percent over the level which would occur in the absence of higher transportation funding. The econometric study performed by Pinnoi (1993) suggests that highway and street capital provides positive marginal benefits to firms in the manufacturing industry. In addition, the study concludes that the productivity slowdown of 1980's may be partially explained by utilizing suboptimal levels of public and private capital.

Some recent studies also confirm the positive nature of the relationship between the highway investment and private sector productivity at macro-level. The econometric study by Nadiri and Mamuneas (1996), an analysis of 35-industry, documents the highway network's contribution to industry productivity growth, national economic performance, and international competitiveness. The study finds that the net social rate of return on investment in the non-local highway system during the 1980's was 16 percent, and the rate of return for the entire road network was 10 percent. This high return to highway capital is due to its network feature (i.e., its benefits are shared by all industries).

Fernald (1999) obtains similar results. He investigates econometrically how changes in roads affect the relative productivity performance of U.S. industries between 1953 and 1989. The study finds that changes in road growth are associated with larger changes in productivity growth in industries that are more vehicle intensive. The results indicate that the highway construction investment between the late 1950's and 1973 substantially boosted productivity. In particular, the estimates imply that public investment had above average rates of return, and contributed to an additional one percentage point to total productivity growth. The study concludes that roads were exceptionally productive before 1973 and normally productive thereafter.

Queiroz and Gautam (1992) investigate empirically the relationship between per capita income and the magnitude and quality of road infrastructure. Cross-section analysis of data from 98 countries, and time series analysis of U.S. data since 1950 shows consistent and significant relationship between economic development, in terms of per capita national product, and road infrastructure, in terms of per capita length of paved road network. The data indicate that per capita stock of road infrastructure in high-income economies is dramatically greater in middle and high-income economies. For example, the average density of paved roads (km/million inhabitants) varies from 170 in low-income economies to 1,660 in middle and 10,110 in high-income economies. Road condition also seems to be associated with economic development: the average density of paved roads in good condition (km/million inhabitant) varies from 40 in low-income economies to 470 in middle and 8,550 in high-income countries.

Another line of research finds no significant relationship between highway investment and productivity. Roeseler and Smithson (1980) address the impact of technological change in rail and highway network on the regional productivity and development. The study investigates the Houston-Beaumont region in Texas. The main finding of this study is that technological changes in transportation systems have very little measurable impact on the overall economy of the region. Husak and Glenn (1998) reached a similar conclusion. Their study attempts to replicate the analyses of Aschauer and others regarding the impacts of public infrastructure investment on private factor productivity for the State of Texas. The generated relationships are, however, statistically insignificant. Relationship Between Infrastructure Investment and Industrial Productivity

Some studies investigate the broader relationship between public capital (highway, water, sewer, etc.) and industrial productivity. Studies by Aschauer (1990b, 1991) and Munnel (1990) show that public capital has a significant, positive impact on the output and productivity at the state and macroeconomic level. The regression coefficients imply that the marginal productivity of public capital is comparable and even higher in some cases than the marginal productivity of private capital. Other research, such as, Morrison and Schwartz (1996) and Nadiri and Mamuneas (1994) have obtained similar findings. They conclude that publicly-financed infrastructure has a significant role in determining productivity growth. The magnitude of these effects is, however, smaller than has been previously reported in the literature.

At micro-level, the positive relationship between highway investment, freight transportation and industry productivity is illustrated by several case studies. Apogee Research (1990) and Hickling (1995) demonstrated the positive impact of highway investment on firm productivity. These studies find that companies such as Koley's Medical Supply, James River Corporation, and Dole Fresh Fruit achieved productivity improvements through stockless purchasing and good transportation access. In addition, the studies find that effective highway network helped other companies, like Alladin Mills and Hewlett Packard to become more competitive by facilitating labor access from adjacent communities. Firms such as Digital Equipment Corporation and Bank of Boston benefited from improved highway network, which allowed for cost-effective expansion into new geographical areas.

In the international arena, Berndt and Hansson (1992) and Conrad and Seitz (1994) investigate statistically the relationship between public investment and industry productivity for Sweden and Germany, respectively. These studies show that public infrastructure contributes to industry productivity and that public investments are a complement to private investments. This line of research is not, however, without, antagonists. Chrihfield and Panggabean (1995) study the productivity of public investment within the context of a neoclassical growth model. The model's estimations indicate that public infrastructure has at most a modest effect on factor markets, and an even smaller impact on growth of per-capita income. Moreover, most infrastructure coefficients in the growth models were found to be statistically insignificant. Neither state nor local investments were found to lead to growth of per-capita income. The study concludes that there are no substantial unexploited gains from public investments.

Ford and Poret (1991) accept the essentials of Aschauer's methodology. They applied the methodology to a broader range of data (several OECD countries and a larger data set for the United States). The results, however, provide little support for Aschauer's hypothesis. The study shows that while infrastructure growth slowed in the 1970s in all twelve of the countries examined, it was accompanied by a deceleration of private-sector total factor productivity in only half of them. Examination of a century of data for the United States implies that there is no relationship between productivity and infrastructure capital in the United States except for the post-war period examined by Aschauer. The authors conclude that overall the regression results cannot support a policy recommendation of a sharp acceleration of infrastructure investment.

In summary, most of the literature findings suggest that a positive relationship exists between public infrastructure investments and freight and industry productivity, though the strength of this relationship has weakened since 1973. There are, however, a number of valid theoretical, econometric and other reasons for exercising caution in the use of these results for policy making.

previous | next
Office of Operations