In order to accurately measure the overall impacts of ARC's infrastructure and public works projects, it is important to divide these projects into two categories. 

1. Economic development projects: These are public investment projects intended to promote business development by attracting new jobs or retaining (saving) threatened existing jobs.
2. Residential projects: These are public investment projects intended to improve basic public health and quality of life by providing basic water and/or sewer services to residential areas that have lacked them.

Most of the projects (87 of 99 studied) were economic development projects—aimed at promoting economic development.  The impacts of those projects were measured in terms of jobs (new or retained), personal income (wages) associated with those jobs, private investment leveraged by the public funding, and tax revenue associated with new private investment.  This chapter reviews the overall findings for each of those measures, as well as ratios of impacts per public dollar spent on these projects. 

For the 12 residential projects, the number of households served is also measured.  Those projects were not intended to directly attract business and jobs, and hence showed no immediate impacts on those measures.  However, they did in some cases also improve the local capacity for future economic development, and those types of qualitative impacts are noted in the individual project discussions in Appendix I.

3.1 Direct Effects: Anticipated vs. Actual Results

Goals.  In the initial project applications for funding, local applicants are required to estimate the number of jobs to be created or retained, the number of businesses to be served or retained, and the number of new or existing households to be served directly by the project.  The job and business goals were applicable for industrial and commercial projects, while the household goals were applicable for residential water/sewer projects. 

Measures of Anticipated vs. Actual Impact.  The aggregate results far exceeded projections in each of those impact measures, as shown in the following table:

Table 3.1 Direct Impact: Aggregate Projections and Results

The "new jobs" fulfillment rate is noteworthy for two reasons. First, unlike certain other federal programs, ARC investments do not always require a guarantee of job creation before granting the funds, so projections of job impact can be somewhat speculative. Second, there is a natural inclination for applicants to "stretch the envelope" on job projections in order to enhance the perceived likelihood of project funding.

Of course, it would be possible for a handful of wildly successful projects to distort aggregate totals, even if the majority of projects failed to meet or even approach projections. But this didn't seem to be the case for the sample reviewed for this report. In fact, in every category for which projections were made, between 72 percent and 100 percent of all projects with projections in those categories (jobs created and households served, respectively) met or exceeded expectations.

Portion of Projects Meeting Goals. Twenty-four projects—most of them residential water and sewer investments—did not submit initial projections regarding jobs created, jobs retained, or businesses served. Ten of these eventually showed "unanticipated results" or impacts in one or more categories. All projects with initial projections were measured for achievement of projections. The portion of projects meeting or exceeding each of the major goals is summarized as follows:

• New businesses served: 59 projects of 72 (82 percent) met or exceeded expectations.
• Retained businesses served: 36 projects of 46 (78 percent) met or exceeded expectations.
• Jobs created: 58 projects of 81 (72 percent) met or exceeded expectations.
• Jobs retained: 31 projects of 38 (82 percent) met or exceeded expectations.
• Households served: 23 projects of 23 with projections (100 percent) met or exceeded expectations.
• Existing households served: 27 projects of 27 (100 percent) met or exceeded expectations.

These results should be seen as based on fairly rigorous success standards because several types of project outcomes were excluded from meeting the "minimum success" thresholds: 

• projects that approached but did not reach projections;
• projects that had large impacts but were nonetheless below projections; and
• projects such as recent industrial parks that are still in "immature" stages.

Eighty-three of the eighty-seven economic development projects studied achieved significant, measurable outcomes. Included are all industrial park, business incubator, and access road projects, as well as 35 of 39 (non-residential) water and sewer projects.  The other four projects either had an anticipated business cancel its plans or had an existing business subsequently close up (or move out).

While variations and shortfalls certainly exist for some measures in some categories, the general conclusion is that each project classification met or exceeded statistical projections. As the site visit narratives show (see Appendix A), viewing projects of all types within their economic and social contexts offers a better understanding of the project impacts and the value of the initial investment.

Results by State.  As the table below indicates the results on businesses, jobs and households served largely reflect differences in the mix of projects. The following analysis is useful mainly as information about the project mix within a state, not as a scorecard or yardstick for comparison between states.⁴ In addition, the table is useful in the review of individual projects within the context of a state's total ARC program investment.

Within each state, the number and dollar value of total investments varied, as did the  impacts generated from the project. For example, South Carolina's high job impact numbers were almost all generated by a single very large project. By contrast, Maryland's projects include a large proportion derived from workforce development and technical assistance efforts. The types of projects reflect state priorities that determined both the scale of required investment and the nature of outcomes. For instance, incubators were a high priority in New York, while access roads were a high priority in Mississippi.

Table 3.2 Direct Impact: Results by State

* includes 3,000 jobs at BMW plant in Greenville-Spartanburg, South Carolina (project 11163)

3.2 Indirect and Induced Effects

Whereas "direct effects" refer to the growth of businesses located at the project site that benefit directly from the project completion, "indirect and induced effects" refer to additional economic growth typically located elsewhere in the community that follows as a consequence of the direct effects.  These additional effects are commonly analyzed in studies of localized economic impacts associated with business relocation and expansion.  Methodology:

Definitions.  The economic development projects were intended to either, (a) support the growth or attraction of new business activity that otherwise would not occur in the area, or, (b) support the retention of existing business activity that was economically threatened and which would otherwise decline or move out of the region.  The former lead to "new" jobs and income, and the latter lead to "retained" jobs and income. 

Treatment of New Activity.  For the new jobs and income, we can distinguish three classes of impacts:

• Direct effects. The business activity of the output, jobs and income directly related to the project are the "direct economic effects" of the project. 

• Indirect effects. In addition, projects have broader impacts elsewhere in the community such as expanding business for local suppliers of products or services that service the new businesses. The additional output, jobs, and incomes for such suppliers are typically referred to as "indirect economic effects." 

• Induced effects. Another impact is the so-called induced effect which includes the expansion of local commercial business as a result of income re-spent by persons working at the new businesses (the direct new hires) and suppliers (the indirect employment effect). 

The additional indirect and induced effects are often referred to as "multiplier effects."  The total effect on jobs and associated income is thus the sum of the direct project effects and the indirect and induced effects.   Since most of these local areas are characterized by significant unemployment, and relatively low labor force participation rates, it is reasonable to expect that the additional jobs and income go to local residents, and are not merely replacing jobs and income from other existing business activities.

Treatment of Retained Activity. This study does not estimate indirect or induced effects associated with business retention since it is unclear whether or not all of the business losses would actually occur without the public investment. If the retained jobs and income would indeed be lost without further public investment, then there could be  potential negative indirect and induced effects—leading to additional job loss for existing businesses elsewhere in the local area. Nonetheless, the uncertainty about how to treat retained businesses and jobs meant that the prudent approach was not to attribute such indirect and induced effects to retained businesses.

Methodology for Analysis.   The measures of direct, on-site impacts on business attraction and retention came directly from interviews with local officials, who were asked to report the actual number of affected businesses and jobs, and to estimate the associated personal income, including existing or saved jobs.  For cases in which there were no reliable estimates of income effects, data from the state labor agency and the US Bureau of Labor Statistics were used to indicate the average wage per worker (based on separate data by county and by industry).⁵

The measures of indirect and induced effects were developed using the Impact Analysis for Planning (IMPLAN) economic model.⁶ It is important to note that multiplier effects differ by industry and by area.  Industries(types of business) can have larger or smaller indirect and induced effects, depending on the portions of dollars going to pay workers and to buy different types of equipment and supplies.  Locationscan have larger or smaller indirect and induced effects, depending on the portion of suppliers and consumer-serving businesses located within the area.  For these reasons, employment, income and business sales data for the year 1997 were obtained by industry, for each of the counties associated with the 99 projects studied. In cases in which projects involved multiple counties, impacts were estimated for the multi-county area.  The IMPLAN model was then run to calculate employment, income, and business sales multipliers associated with growth or shrinkage of each industry in each county. For each project, the types of business (industry) associated with the business expansion or attraction were identified, and the applicable multipliers were then applied.  For projects in which the specific types of business were not all known, multipliers representing an average of the area's dominant industries were applied.

3.3 Job Impacts: Direct, Indirect and Induced Effects

New Jobs.  A total of 23,377 new jobs were directly created as a result of the ARC-funded projects.  These direct effects include only jobs at the sites served directly by the ARC-funded infrastructure and public works investments.  In addition, it is estimated that another 20,954 new jobs were created away from the project sites by indirect effects on off-site suppliers and induced effects on consumer re-spending of additional worker incomes.  These indirect and induced effects follow as a consequence of the directly created new jobs.  (See Appendix C for further discussion of the calculation of indirect and induced effects.)  All of these new jobs (both direct and indirect/induced effects) were created because of the projects. 

Retained Jobs.   Another 16,387 existing jobs were directly retained or saved as a result of the ARC-funded projects.  It is reasonable to assume, based on project application data, that those directly-affected jobs would most likely have been lost without the projects.  The extent of their indirect effects on supplier businesses and induced effects on consumer-serving businesses is less clear.  Those businesses had already existed before the projects were implemented.  If the projects had not been implemented without ARC funding, the directly affected businesses may have responded by closing or by relocating, or they may have survived in their current locations by adjusting products and services for other markets.   If we assume that all of the business activity associated with indirect (supplier) and induced (consumer) sales would indeed have disappeared, then it would be reasonable to add indirect and induced effects associated with the retained jobs.  While that is a distinct possibility, this study adopted a more conservative approach that  counted additional indirect/induced effects based on new jobs, but not any additional indirect/induced effects based on retained jobs.

Total Jobs. The estimated total number of job impacts of the ARC-funded sample projects was 44,331. This includes direct new jobs and indirect/induced new jobs. It does not include the retained jobs, nor estimates of indirect/induced effects for retained jobs.  This total impact can be broken down by project type, as follows:

• 7,998 total jobs created from 22 industrial park projects (average of 364 each); 
• 3,869 total jobs created from 11 business incubator projects (average of 352 each); 
• 3,723 jobs created from 15 access road projects (average of 248 each); and,
• 28,741 jobs created from 39 water/sewer projects (average of 737 each, but this drops to 598 when the large BMW project is excluded).

Table 3.3 Total Overall Job Impacts

These numbers reflect differences in the average size and scale of the projects, and not necessarily project success.   A full breakdown of the job impacts is shown in Table 3.3. Table 3.3 also shows that average job creation was relatively greater for the projects in transitional areas than for the projects in fully distressed areas.  That reflects a combination of two factors:

• attracting business is harder in the more distressed areas, and hence the average number of jobs created per project is smaller in those areas; and 
• most of the residential projects were in the distressed areas and were aimed at public health rather than immediate economic development.

3.4 Personal Income: Direct, Indirect and Induced Effects

Additional Income.  Personal income derived from wages from newly created jobs has a variety of local economic impacts.  While the impacts of economic development projects are often tracked in terms of job creation, the most tangible benefit to people in the target areas comes from the enhancement of their incomes.  Another advantage of measuring program impact in terms of personal income is that the income measure reflects differences between the creation of high-paying jobs and the creation of low-paying jobs. Because counties in which these projects occurred were characterized by high unemployment and low income levels, it is reasonable to assume that essentially all of the additional income created (directly or indirectly) by these projects flows to existing residents of the county.

Measurement.  The estimates of direct effects on retained wages (from saved jobs at existing businesses) and on new income (from new jobs attracted) came from interviews with local officials, and were supplemented when necessary with average wage data from the U.S. Bureau of Labor Statistics.  The estimates of indirect and induced effects on personal income came from the IMPLAN model.⁷

Overall Results.  Table 3.4 shows a breakdown of the retained wages as well as the new (direct) wage income and indirect/induced income impacts, by project type, area classification and state.  Overall, it shows that the 87 (case study) economic development projects helped to directly retain $440.7 million of existing wages at threatened jobs, attract $576.9 million of new wages at the project sites, and led to a net expansion of $950.3 million of personal income.   As with the job impacts, the personal income impacts were largest for the water/sewer (rather than industrial park and business incubator) projects and for the transitional rather than distressed counties.

Wage Levels.  The new jobs directly generated by these ARC-funded projects were primarily industrial rather than commercial or service jobs, and thus would be expected to have wage levels higher than the overall average in those counties.  An attempt was made to document these differences.  However, it was concluded that it was not possible to accurately measure such differences in wage levels for this study.  The reason is that the project information on directly generated jobs and payroll from ARC-supported projects did not sufficiently distinguish levels of part-time and full-time jobs.  Published county-wide data on wage rates (from the U.S. Department of Commerce's "County Business Patterns" database), on the other hand, do adjust wage statistics to reflect hourly or full-time rates.  Thus, there was not sufficient consistency in the definition of pay per job to allow for a comparison of project and overall wage rates.  Nevertheless, there was a clear consensus, indicated in the local interviews, that the ARC-funded projects had indeed broadened available job opportunities and provided desirable types of jobs.

Note: See text for discussion of indirect and induced effects of retained jobs

3.5 Effects on Public and Private Investment

Overview.  ARC does not fully fund any infrastructure or public works projects.  Rather, ARC co-funds projects which also have some other federal funding assistance.  The other federal funding is predominantly from the Economic Development Administration, the Farmer's Home Administration of the U.S. Department of Agriculture, the U.S. Department of Housing and Urban Development (through Community Development Action Grants) or the Federal Highway Administration of the U.S. Department of Transportation.  In addition, most of those other federal programs also require some state or local matching funds.  This section reviews these funding patterns in two parts.  First, the mix of public funding is described.  Then the leveraging of private sector funding is analyzed. A full breakdown of the public funding by project type, area distress level, and state is shown in Table 3.5.

*includes BMW plant infrastructure:  $2,195,000 ARC, $595,000 state, $3,555,731 local = $6,345,731 total.

Public Funding Mix.  Because of the typical mix of public funding in ARC projects, ARC cannot take full credit for the economic impacts of any of its projects.  It can, however, take credit for helping to leverage other federal, state, and local funds, as well as private funds. Overall, ARC funding for these projects totaled $32,433,047, which is 28 percent of the total public cost ($117,147,976) for these projects.  Other federal funding averaged 24 percent of project cost, while states invested an average of 15 percent, and local funding averaged 33 percent of the total. Viewed another way, each dollar of ARC investment helped to make possible a package of $2.61 in other public funding, adding up to $3.61 of total public funding.

Overall, the ARC portion of total public funding was:

• 33 percent of all public funding for access road projects;
• 25 percent of all funding for business incubator projects;
• 20 percent of all public funding for industrial park projects;
• 31 percent of all public funding for water/sewer projects serving business sites; and
• 38 percent of all public funding for water/sewer projects serving residential areas.

The ARC portion of public funding was 41 percent for projects in distressed areas and 26 percent for projects in transitional areas.  All together, these figures show that ARC funding has played a relatively larger role in those areas that are most in need—especially economically distressed areas—and in those projects that are most critical for basic household and business operation such as water and sewer services.

Private Investment Leveraged.  Of the 87 economic development, non-residential infrastructure projects, 34 were initiated with records of commitments for private sector investment.  Other projects were initiated with expectation of private sector investment, but no record of a specific amount for it.  The records of initial commitments indicate a total $862 million, of which $400 million was attributable to one single project (the new BMW plant in South Carolina).  Excluding that one project, there was an original commitment for $462 million of related private sector investment. 

Local interviews and data collection conducted for this project showed that these projects actually had an even larger impact on private investment at their sites.  This investment included new or renovated buildings and other business facilities.  The actual private sector investment associated with (or resulting from) these projects totaled $3.075 billion, of which nearly half was attributable to the single BMW plant.  Excluding that one project, the total actual private investment was $1.675 billion.  Details of the private investment are shown in Table 3.6.

The corresponding level of public funding for these economic development projects, excluding the BMW case, was $26.4 million of ARC funds and $101.6 million of total public funds.  Thus, there was $16.65 of private investment for each dollar of total public funding.

It is notable that these private sector leveraging rates vary dramatically among types of projects for at least two reasons. First, because of the nature of various project types there is relatively large variation in the amount of permanent private investment. In some cases, such as business incubator facilities, there is little substantial permanent private investment.  On the other hand, there is substantial private investment associated with access roads and most non-residential water/sewer projects. Second, the variation in the maturity and timeline of projects affects the amount of private investment. For example, industrial parks may not yet have moved to full-scale marketing of the facilities.

Table 3.6  Private Investment Leveraged

*Each figure for South Carolina includes the following  funding for the BMW plant:  $2.195 million ARC, $6.345 million total public funding, $400 million of original private commitment for project and $1.4 billion of total private investment leveraged by the project.

3.6 Effects on Tax Revenues

For a distressed area, the attraction or expansion of business activity can also bring about more tax revenue which can help pay for such things as improvements to local schools and public services.  For that reason, there is interest in examining the likely tax impact  of these projects.

Tax revenues can be affected by economic development in several distinct ways:

• The additional private investment can lead to increased local property tax revenues;
• The additional wages can lead to additional state income tax revenue;
• The re-spending of wages on consumer purchases can also lead to additional state and local sales tax revenues; and
• The additional business income can lead to additional business income tax revenues.

Additional jobs and population growth can also lead to offset increases in public expenditures for schools, police, fire, and other public services.  However, in the case of Appalachian communities with relatively high unemployment, it can be expected that these projects will primarily serve the existing area population and hence have relatively little effect on attraction of new population, though there can be some additional costs of police/fire protection services associated with new or expanded business activity.  Such cost impacts are, however, very specific to local situations, and hence are not addressed here.

Results.  The estimated project impacts on annual tax collections are as follows:

• state income tax revenue of $14.3 million ($12.1 million excluding BMW project);
• state/local sales tax revenue of $13.9 million ($11.2 million excluding BMW project); and
• local property tax revenue of $29.2 million ($14.6 million excluding BMW project).

A breakdown of the tax revenue impacts by project type, area classification, and state is shown in Table 3.7.  The differences among states in sales and income taxes primarily reflect the levels of personal income impact, as well as state differences in average sales and income tax rates.  In addition, the differences in property tax impacts reflect the degree of local tax exemption offered as part of the public incentive package to attract some businesses.  In some places, some or all of the projects were exempted from local property taxes. 

Table 3.7  Additional Tax Revenues Generated

3.7 Benefit/Cost Analysis

Measurement Approach.  The purpose of ARC project funding for infrastructure and public works projects is to transfer federal funds to targeted local projects, in order to promote improvements to the economic development and quality of life for areas that are considered to be economically depressed (classified as either distressed or transitional).  In the parlance of benefit/cost analysis, the focus of this funding is to bring about desired distributional impacts.  In this sense, if a business is attracted to invest in and locate activities in a depressed area, then it is a desired benefit even if that business activity was attracted from elsewhere in the United States (presumably in a less depressed area). 

Given the desire to attract business activity, "success" can be measured in terms of jobs, income, or private investment.  There is no single benefit/cost ratio that is directly applicable.  Rather, it is useful to assess the returns on investment for the economic development projects in terms of several measures:

• public cost per job created;
• private sector investment leverage (ratio of private investment per public dollar); and
• personal income created per public dollar spent.

For the residential projects, the primary impacts are the provision of a basic quality of life through access to community water and sewer service, and associated public health improvements. Case studies with local interviews were conducted to assess how the residential public works projects affected the communities, but the results are qualitative rather than quantitative benefit/costs measures.

To assess the impacts associated with economic development (non-residential) projects, two perspectives were used for analysis:

• ARC investments were compared with actual results for the entire project in which the investment was made. This type of ratio is commonly used in program evaluations.  But ARC is only one of several public investment sources used in a project financing package. As a result, this type of ratio is accurate only if all of the project results depended exclusively on the ARC funding, and none would have occurred without it.
• To correct for this problem, investment ratios were also developed that compared the total public funding with actual results, and credit is assigned to ARC based on its share of total public investment. This method delivers a much better understanding of actual return on public investment, and eliminates the common problem of "double dipping" among the claims of partnering programs in development projects. 

A further discussion of this approach and its differences from other forms of benefit/cost analysis is provided in Appendix C.

Results for Economic Development Projects.  The findings on non-residential, economic development project results are summarized in Table 3.8.  Three columns of numbers are shown: 

1. The first column shows the project results in terms of private investment, jobs, and income.  Only impacts generated by the 87 economic development projects are shown, so that they can be compared with the public costs for those same projects.
2.   The second column shows results comparing total impacts with ARC dollars spent.  As previously noted, this comparison is most useful if it is assumed that the project results would not occur without the ARC funding. 
3.  The third column shows results comparing total impacts with total public dollars spent.  Since the ARC funding is always accompanied by additional public funding for other aspects of the project, the total public dollars are always greater than the ARC dollars alone. The measure of total public dollars combines ARC funds, other federal funds, state funds, and local public funds and treats them all as one package of funding.  The resulting ratio thus represents the "average impact" of public funding for these projects.  This measure is most useful when it is recognized that the marginal impact of the ARC dollars cannot be accurately distinguished from the marginal impact of other public dollars invested in these projects.

*The $1.675 billion reflects the $3.075 billion total minus the BMW project, which had a disproportionately high level of private investment ($1.4 billion)
Note: All ratios are based on non-residential project funding: ARC $28.6 million, total public $106.9 million; see text for important limitations on interpretations of these ratios.

The results are impressive.  Findings are as follows:

• Total private investment stimulated.  Overall, there was nearly $29 of private investment per dollar of public investment in economic development projects.  Even deleting the single large project, it was $16.  The rate is so high largely because of the nature of the public works projects, in which an access road, sewer line, water line, or industrial park development improvement is made that may attract other businesses to the location.  
• Job creation rate.  Overall, the economic development projects studied here cost $2,412 per new job created, including indirect and induced job creation.  If jobs saved are also counted, then the average cost drops to $1,761 per job (new and retained).  
• Personal income. The new jobs led to increased personal income for residents of the affected counties.  The ratio was approximately $9 of annual personal income to $1  of a one-time public funding investment for economic development projects.

Table 3.9 shows how the leveraging of public dollars differs among the four types of projects.  This is shown first in terms of ratios per ARC investment and second in terms of ratios per total public investment.⁸ Because of the unusually large private investment impact of the Greenville-Spartanburg BMW project on the water and sewer category, the project category is analyzed both with and without investments and impacts related to that single project (#11163). The analysis is useful as a presentation of alternative perspectives on viewing ARC investment impact, not as a comparison of the two sets of figures.  In general, it shows that rates of private sector leverage tend to be highest for the water/sewer and industrial park projects.  Public costs per job tend to be lowest for water/sewer projects. 

Table 3.9 Breakdown of Results per Public Dollar by Project Type for Economic Development (Non-Residential) Projects

*retained job totals exclude Hagerstown Technical Innovation Center

Breakdown of Overall Results for All Projects.  Table 3.10 differs from the preceding tables in that it shows the ratios of total results for all 99 projects, including the 12 residential projects.  Some of the residential projects did leverage private investment, but none of them had immediate measurable impacts on jobs and associated income.  Thus, the ratios of overall results shown here indicate slightly lower ratios for private sector leverage and income creation (26:1 instead of 29:1), and slightly higher ratios for cost per job ($2,643 instead of $2,412).

The breakdown also reflects differences by project type. Rates of income creation as well as private sector leverage tended to be higher for the water/sewer and industrial park projects, and lower for the business incubator and access road projects.  They were also higher for projects in transitional areas and lower for projects in distressed areas.

Table 3.10  Results per Total Public Dollars  for all Economic Development & Residential Projects