How are Economic Development Districts Riding the Waves of Economic Development Strategies?

Abstract

Previous studies have examined the local adoption of various development strategies primarily by surveying economic development practitioners, although the survey method suffers from subjectivity and recollection errors. To overcome these problems, I adopt a text mining method to canvass the comprehensive economic development strategies (CEDSs) across economic development districts (EDDs) in the United States. Twenty-eight main development strategies are identified, most of which belong to the “second wave,” while the “fourth-wave” strategies—oriented toward social equity and sustainability—are popular enough to justify separation from the “third wave.” Moreover, EDDs combine various waves of strategies to form their strategy portfolios. Two main types of portfolios have been identified—one is dominated by the second-wave strategies, while the other is more balanced across four waves and places a focus on both second- and fourth-wave strategies.

Keywords

economic development strategies text mining strategy portfolio equity and sustainability-oriented strategy

Mining the Texts of Economic Development Strategies may Yield new Insights

Economic development planners across the nation have adopted various development strategies to promote prosperity. These strategies include traditional tax incentives, infrastructure improvement, and industry targeting, as well as more modern tactics such as workforce development programs, cluster development, and local business incubators (Bradshaw & Blakely, 1999; Olberding, 2002). Having a well-developed comprehensive economic development strategy (CEDS)—a strategy-driven plan for regional economic development—is a prerequisite for designation by the U.S. Economic Development Administration (EDA) as an economic development district (EDD).¹

Economic development researchers have developed various typologies to describe these strategies. The most prominent typology classifies development strategies into three waves, which have characterized the strategy dynamics since 1850 (Bradshaw & Blakely, 1999; Clarke & Gaile, 1992; Hu & Zheng, 2021; Olberding, 2002; Zhang et al., 2017). More recent studies suggested that there might be a fourth wave on the rise (Stokan et al., 2020; Zhang et al., 2017). The definitions of these waves, their time periods, and the percentage of texts for which each of them account in the current collection of CEDSs are listed in Table 1.

Table 1.

Waves of Economic Development Strategies.

Wave	Time period	Definition	Percentage of texts this wave accounts for in studied CEDSs (%)
1	1850s–1980s	“Low road” development strategies to attract footloose firms to locate in a jurisdiction, such as tax abatements, subsidies, and low-cost land (Hanley & Douglass, 2014)	18
2	1980s	Investment and entrepreneurial policies that foster innovation, nurture new businesses, breed high-technology industries, and promote local businesses (Clarke & Gaile, 1992; Hanley & Douglass, 2014; Leicht & Jenkins, 1994)	47
3	1990s	Various definitions Combines the previous two waves of strategies and aims to build general institutional or individual capacity (Elisinger, 1995) Enhancing the local context instead of meeting a particular firm's needs (Bradshaw & Blakely, 1999), such as promoting industrial clusters and regional competitiveness Regional partnerships to pursue economic development (Green, 2014; Olberding, 2002)	17
4	1990s	Strategies focusing on sustainability and/or social equity issues (Stokan et al., 2020; Zhang et al., 2017)	18

Note: Some researchers disagree with the time period distinction between waves and suggested that jurisdictions bundle strategies together (Lowe & Feldman, 2018)—consistent with what this paper finds. CEDSs = comprehensive economic development strategies.

Alternative typologies have been developed, including supply- versus demand-side strategies (Eisinger, 1989), strategy mixing (Hanley & Douglass, 2014; Kassekert & Feiock, 2009; Lowe & Feldman, 2018), quantitative versus qualitative growth strategies (Miranda & Rosdil, 1995; Morgan et al., 2019; Vanderleeuw et al., 2011; Zhang et al., 2017), and “shoot anything that flies; claim anything that falls” (Rubin, 1988). Recently, some studies have used factor analysis to cluster strategies and formed several other typologies (Hanley & Douglass, 2014; Reese, 2006).

Current typologies suffer from two major problems. First, most studies are based on surveys that ask local economic development planners or politicians to report their strategies. However, self-reporting has been proven to be error-prone (Presser & Traugott, 1992; Stone et al., 1999) and therefore may not truthfully reflect the strategies adopted. Second, in most survey designs, researchers prespecify the categories of development strategies (Green et al., 1996; Morgan et al., 2019; Reese, 2006), which is prone to confirmation bias and doesn’t account for innovative and nuanced strategies. Moreover, these surveys are administered either by the International City/County Management Association (ICMA), which ignores local governments with a population of less than 10,000, or by the researchers themselves, which is costly and hard to replicate. The heavy reliance on the ICMA surveys also means that development strategies are being forced into predetermined characterization. In contrast, a data-driven approach based on actual CEDS documents has the potential to discover new insights and new typology.

With the development of machine learning technology, planners are able to directly analyze the CEDS documents with text mining techniques and overcome these problems. Text mining techniques harness the computational power of a computer to analyze multiple texts, which makes the analysis cost-effective, easily repeated and updated, and less subject to human bias. Planners have already started to use text mining methods to analyze urban planning documents and articles. For example, Etienne et al. (2019) adopted the technique to analyze the content of urban plans. Gobster (2014) and Fang and Ewing (2020) have applied it to study the dynamics of major topics in planning literature. Hu and Zheng (2021) conducted a text mining and content analysis to compare smart city initiatives between the United States and China. Similar technology can be applied to examine the strategies in CEDSs.

This paper takes on this task. I’ve adopted the text mining technique “topic analysis” to examine major development strategies adopted in the most recent CEDSs across all EDDs in the United States. I find 28 main strategies, with business and workforce infrastructure provision, business attraction, and technology industry targeting as the most frequently mentioned. Of these 28 strategies, three have not been identified by extant studies: rural and agricultural area development, citizen participation, and creative class and event organization. I also find that a four-wave analogy best describes the current landscape of development strategies. The second wave dominates and accounts for about half of the texts in all CEDSs, while the other three waves share the other half almost equally. Finally, EDDs combine different waves of strategies to form a comprehensive strategy portfolio, which confirms the strategy mixing hypothesis (Hanley & Douglass, 2014; Lowe & Feldman, 2018). Two types of portfolios have been identified: a second wave dominated portfolio, and a more balanced portfolio that mixes all four waves while emphasizing both the second and fourth waves.

Practically, economic development planners should consider whether their strategy portfolio is carefully put together with different strategies that complement each other and look at other jurisdictions’ strategy portfolios for inspiration from emerging strategies and innovative policies. This paper also shows that text mining can be a useful tool for detecting emerging strategies and various strategy portfolios.

Existing Development Strategy Typologies

Since the 1980s, economic development researchers have developed multiple typologies to describe development strategies adopted by different jurisdictions. Rubin (1988) interviewed economic development practitioners and found that they, facing an uncertain environment, felt that their ability to bring about economic development was dependent upon factors largely beyond their control. As a result, they “look(ed) for credit” (i.e., adopted a wide range of actions that may or may not be effective), and sometimes, were even known to be ineffective. But then they claimed credit for any positive economic outcomes, fully aware that many did not come from their actions. Rubin summarized this strategy as “shoot anything that flies; claim anything that falls.” While Rubin’s interviews are based on a small number of practitioners, this opportunistic strategy was adopted by 190 out of 856 (22%) studied municipalities in the United States and Canada (Reese, 2006). Stokan (2013) partly supported Rubin in finding that U.S. municipalities with populations of 10,000 or more tended to adopt a large number of approaches to promote development, but disagreed with Rubin because these municipalities were selective in the approaches that they adopted based on local characteristics. Similar results were found by Morgan et al. (2019) with data from North Carolina cities and counties. Thus, Rubin's simple typology appears to explain part but not all of the local practitioners’ approaches for adopting development policies.

Another earlier typology classified local development strategies into supply- versus demand-side (Eisinger, 1989). Supply-side strategies aim at reducing firm costs. Examples are tax breaks and subsidies. In contrast, demand-side strategies invest in local attributes, support indigenous growth, and restructure local market contexts (Clarke, 1993). Examples are setting up business incubators and forming active partnerships between the state and private enterprises (Hanley & Douglass, 2014). Hanley and Douglass, however, rejected the conceptual distinction that a state adopts either a supply-side or a demand-side strategy in an empirical study. They found that most states combine the two sides. They identified six strategy types through a factor analysis: (a) export-driven recruitment, (b) entrepreneurial, (c) rapid-response export-driven recruitment, (d) education-driven recruitment, (e) high-tech recruitment or “chip chasing,” and (f) minority development.

The most popular and well-developed typology is the waves typology, which characterizes the dynamics of development strategies since the 19th century (Bradshaw & Blakely, 1999; Clarke & Gaile, 1992; Hu & Zheng, 2021; Olberding, 2002; Zhang et al., 2017). The first wave began in about the 1850s and lasted until about the 1980s. It focused on Eisinger's supply-side strategies, offering tax abatements, subsidies, and low-cost land to attract footloose firms to locate in a jurisdiction (Hanley & Douglass, 2014). This is often characterized as “low road” development because it invokes zero-sum competition among jurisdictions, and puts downward pressure on wages and employment stability (Hanley & Douglass, 2014; Luria & Rogers, 1999). These strategies were modestly effective (Dewar, 1998) or ineffective (Peters & Fisher, 2004) in empirical studies.

The second wave emerges in the 1980s as economic development practitioners seek to break out of those low road approaches. This wave focused on investment and entrepreneurial policies that foster innovation, nurture new businesses, breed high-technology industries, and promote local businesses (Clarke & Gaile, 1992; Hanley & Douglass, 2014; Leicht & Jenkins, 1994). Exemplary policy tools include technology transfer programs between universities and local businesses, incubator programs to breed start-up companies, and enterprise or innovation zones. These tools are still widely adopted across the United States.

Researchers have debated about what the third wave entails (Bradshaw & Blakely, 1999; Hanley & Douglass, 2014). Elisinger (1995) suggested the third wave combines the previous two waves of strategies and aims to “build general institutional or individual capacity” (p. 153). Bradshaw and Blakely (1999) expanded on this idea and argued that the specific policies adopted in the third wave since the 1990s are similar to those of the second-wave entrepreneurial policies, but are geared toward enhancing the local context instead of meeting a particular firm's needs. The focus is placed on factors outside of individual firms, but inside the local business environment (e.g., enhancing regional competitiveness; Porter, 2000), promoting industrial clusters (Bradshaw & Blakely, 1999; Feser et al., 2008), and fostering interfirm networks (Hanley & Douglass, 2014). Olberding (2002) argued that the fundamental difference in the third wave is the organization of economic development practice. Jurisdictions have worked toward regional partnerships to pursue economic development (Green, 2014). This is also partly consistent with Bradshaw and Blakely's idea, because some industrial clusters and industrial competitiveness expand beyond a single jurisdiction (Feser et al., 2001; Hill & Brennan, 2000), and naturally require a regional effort to pursue.

However, a few recent studies have found that the waves typology does not accurately represent how jurisdictions put their strategies together. Rather, jurisdictions bundle or mix different waves of strategies to form a managed strategy portfolio. For example, Lowe and Feldman (2018) found that, in practice, older traditional development tools coexist with newer approaches that orient toward entrepreneurship and innovation. Thus, a managed portfolio of strategies, or strategy mixing, more accurately describes the actual adoption of development strategies in U.S. localities (Lowe & Feldman, 2018). Hanley and Douglass (2014) found that most states combine low road with high road approaches. Zhang and Warner (2017) also confirmed that business incentives are often combined with local business expansion and retention strategies. Kassekert and Feiock (2009) explained that bundling economic development policy instruments may be motivated by economic, political, and administrative reasons. Economically, the adoption of one policy may lower the marginal cost of implementing another specific policy instrument; thus, policy bundling helps jurisdictions obtain an economy of scale. Politically, policy bundling can create a more attractive package for voters than a single policy that benefits very specific industries or clients. Administrative and professional norms, as well as path dependency, can also explain the bundling of policies: Existing policy tools are hard to get rid of and therefore become a part of the policy bundle.

More recently, some researchers suggested that a fourth wave of development strategies that deal with sustainability and social equity has started to emerge. Issues such as affordable housing, equitable education, and social justice have made it to the agenda in a number of cities since the 1990s (Clarke, 1993), and the concern for equity has increased over time with local governments directing tangible growth-related benefits to low-income groups (Goetz, 1994). Miranda and Rosdil (1995) found that local jurisdictions have become more inclined to embrace qualitative growth and progressive development policies. In addition, Zhang et al. (2017) found that economic developers couple sustainable economic development policies with community development strategies to address the broad range of barriers faced by communities. Stokan et al. (2020) found that local governments that face less competitive pressures, have more resources, and experience greater intergovernmental involvement are more likely to embrace equity-enhancing development strategies.

Thus, a few questions emerge. Is there a fourth wave of strategies about social equity and sustainability? If so, how prevalent is it? Do jurisdictions focus on one wave of strategies, or do they mix various waves? Using the most recent CEDS documents in EDDs of the United States, this paper answers these questions and reveals a typology that best describes the current landscape of economic development strategies.

Data and Method

CEDSs

My data preparation process includes three steps: data collection, conversion, and verification. The first step is the collection of the CEDSs. In 2020, I reached out to the EDA and obtained the most updated information on current EDDs, which is organized in an Excel spreadsheet that documents the EDDs’ names, years of establishment, official websites, and locational information. For designation as an EDD, a multicounty region must gather commitments for supporting economic development activities in the district from a majority of these counties and obtain the concurrence with the designation request from the state(s). Moreover, at least one county in an “economically distressed region”² must be of sufficient size or population and have the resources to foster development on a multicounty scale. In addition, it must have a CEDS approved by EDA. Theoretically, EDDs are not nationally representative, but in reality, the vast majority of the nation now has established EDDs (as shown in Figure 1 by the few white spots signifying non-EDD regions). EDDs are the basic regional units of economic development planning and thus appropriate for an analysis of development strategies.

Figure 1.

The spatial distribution of different types of strategy portfolios.

In 2020 and 2021, my research assistant and I searched the official websites of each EDD to download their most updated CEDSs according to this list of EDDs. Most EDDs have their CEDs in downloadable pdf files, while some maintain them as web pages, which we copied and pasted into text files. There are a few EDDs that do not publish CEDSs on their websites. We were able to collect a total of 376 CEDSs from a pool of 397 EDDs.

The EDA defines a CEDS as “a strategy-driven plan for regional economic development. A CEDS is the result of a regionally owned planning process designed to build capacity and guide the economic prosperity and resiliency of an area or region” (U.S. Economic Development Administration [EDA], 2019). Thus, these documents should capture the main development strategies of an EDD and are reliable materials for this analysis. Admittedly, some regions lead economic development through a local nonprofit economic development organization that has its own development strategies. They use EDD and CEDS merely for formality and EDA funding purposes. In cases like these, CEDS may not capture the actual strategies employed.

In the second step, I converted all collected CEDSs into plain text files. As mentioned above, most of the collected CEDSs are in pdf formats that need to be converted to plain text files before they can be read by the computer algorithms. There were 364 CEDS that were relatively easy and straightforward to convert. Eight were image-converted pdf files that contained no characters and only images; therefore, I adopted an image recognition procedure to convert them. Four were in secured mode (only readable) and could not be converted. A total of 372 CEDSs were converted into plain text files and included in the analysis.

In the third step, I verified the timeliness of the included CEDs by examining their envisioned timelines, as well as the date of updates. The vast majority of these CEDSs are up to date, with their envisioned timeline covering 2021; only 23 are out of date. Even among these 23 out-of-date documents, 13 have just recently expired, with the end year of the strategy being 2020. Several EDDs mentioned on their websites that they are currently working on a new version covering the next 5 years. EDDs with CEDSs included in the analysis are visualized in Figure 1 in different grey scales.

I also collected the county-level unemployment rate data from the U.S. Bureau of Labor Statistics’ Local Area Unemployment Statistics program in 2018, when the majority of the EDDs in the data set published their current CEDSs. I used these data to explore whether the unemployment rate is correlated with the EDDs’ adoption of various strategies.

Topic Modeling

To analyze the contents of these CEDSs objectively and efficiently, I adopted the Latent Dirichlet Allocation (LDA) topic modeling method. Topic modeling is an unsupervised machine learning method to analyze texts and form major topics. This method can efficiently process large amounts of texts and cluster them into main topics, quantify how many topics each document contains, and calculate the percentage of text for which each topic accounts in each document. There are a number of different topic modeling methods but LDA is the most commonly adopted. LDA assumes a straightforward distribution of topics in each document, and uses the maximum likelihood method to classify text into topics (Jelodar et al., 2019). It fits the research purpose of this paper—to find a combination of development strategies in each EDD's CEDS and to quantify the relative importance of each strategy.

The number of topics is specified by the researcher in the LDA model. To avoid subjectivity, I adopted statistical methods, including Arun et al. (2010), Cao et al. (2009), Deveaud et al. (2014), and Griffiths and Steyvers (2004), to help find the optimal number of topics (i.e., strategies). This data-driven approach of detecting strategies overcomes the subjectivity and replicability issues in prior studies.

I then classified these strategies into waves, following primarily the definition of Osgood et al. (2012). However, unlike Osgood et al. (2012), who grouped together the third and fourth waves of strategies, I allowed the fourth wave of strategies to be separated, if the percentage of text for which it accounts was as large as the other three waves. Thus, I empirically verified whether the fourth wave of strategies is prevalent, and if so, whether a four-wave typology should replace the three-wave typology.

After classification into waves, each EDD ends up with a strategy portfolio comprising a specific combination of various waves. To identify the main types of strategy portfolios, I adopted a k-means clustering algorithm to group EDDs by maximizing within-group similarities and minimizing intergroup overlaps. These groups form the topology of how local jurisdictions mix and match waves of development strategies. From them, I could tell whether EDDs focused on one particular wave of strategies or combine various waves. Finally, I visualized the spatial distribution of development strategy portfolios and examined the correlation between the unemployment rate and strategy adoption, to explore whether geography and economic condition shape each locality’s choice of portfolio.

Results

Development Strategies in the CEDSs

After combining the results of the four statistical methods, I found that the optimal number of strategies in the collected CEDSs is 28. Each of the statistical methods mentioned in the previous section identifies a range of the optimal number of strategies to be formed. These optimal numbers ensure maximum separation between different strategies. The only number that is found optimal by all methods is 28. This is not to say that 28 is the only appropriate number, though. A greater number of strategies can be formed with a more detailed separation of topics, and fewer strategies can be formed if several closely related topics are merged. For example, in the identified 28 strategies, several belong to a broader category of industry targeting strategy, but each targets a different industry. It is imaginable that in an alternative clustering algorithm, these industry-targeting strategies can be merged into one. Nonetheless, the 28 formed strategies are reasonable to characterize what EDDs do to promote economic development. Separating several industry targeting strategies will not hamper an understanding of economic development planning practice because their similarities can be compared when interpreting the results. This approach allows a more in-depth look into which specific industries are the main targets of EDDs. I have also formed more (up to 35) and fewer (down to 22) strategies to compare with the 28 strategies to check for sensitivity of the results. I did not find another number to make better sense.

Prior studies have specified various numbers of development strategies, and this optimal number, 28, sits in the middle of the spectrum. For example, Currid-Halkett and Stolarick (2011) examined 13 strategies adopted by economic development practitioners. Hanley and Douglass (2014) studied 15 economic development policies, while Reese (2006) examined 34. In a survey of 900 local governments, Green et al. (1996), found that an average local government adopted 18.3 strategies. Their survey asked about a total of 64 strategies. Stokan (2013), in comparison, found an average of 12.6 strategies adopted out of 46. In a more updated survey conducted by Morgan et al. (2019), the average number slightly increased to 19.2, out of 54 possible strategies. Clarke and Gaile (1992) and Osgood et al. (2012) characterized 47 and 51 strategies into three waves, respectively. Note that it is sensible for the number of strategies identified in this paper to be more than the average number of strategies adopted, but fewer than the total number of strategies surveyed. This is because, very infrequently, adopted strategies will not have enough texts to form text clusters (and thus fall off from this analysis), but the emergence of a text cluster does not require a strategy to be adopted by more than half of the jurisdictions.

Table 2 shows the 28 strategies, and Table 3 includes a concrete example for each strategy to help contextualize what these strategies actually look like in CEDSs. Two strategies stand out as the most commonly adopted. They are business and workforce infrastructure provision (with top keywords broadband, workforce, education, skill, and business); and business attraction (with top keywords business, new, partner, industry, and economy). The Kenai Peninsula Borough CEDS 2019, for example, adopted a strategy belonging to the business and workforce infrastructure provision category, stating “expanded broadband access, industry-focused infrastructure” (p. 5). An example of the business attraction strategy is found in the Central Alabama CEDS 2017: “attraction of new business development, especially those reliant on data processing and transmission and open-access fiber networks” (p. 26).

Table 2.

Development Strategies Identified.

Rank	Development strategy	Wave of strategy	Percentage of CEDS texts (%)	Top 10 keywords
1	Business and workforce infrastructure provision	1 & 3	6.17	Broadband, workforce, education, skill, business, access, resource, support, infrastructure, provide
2	Business attraction	2	5.95	Business, new, partner, industry, economy, opportunity, market, attract, grow, collaborate
3	Technology industry (industry targeting)	2 and 3	4.87	Industry, technology, company, workforce, college, business, support, train, skill, facility
4	Economic development strategy	1 & 2	4.50	CED, EDA, update, committee, report, EDD, strategy, action, comprehensive, SWOT
5	Population change and livelihood support	3 \| 4	4.28	Population, census, household, age, income, community, school, poverty, survey, resident, race
6	Project and funding	1 & 2	3.98	Project, program, fund, loan, goal, grant, council, agency, complete, invest
7	Natural resource-reliance industries (industry targeting)	2	3.95	Product, forest, mine, natural, resource, gas, manufacturing, land, agriculture, wood
8	Disaster recovery and mitigation	3 \| 4	3.95	District, disaster, recovery, community, assist, build, fund, infrastructure, mitigate, activity
9	Basic public infrastructure provision	1	3.87	Water, system, project, parking, sewer, construct, road, site, street, wastewater
10	Transportation development	1	3.64	Project, parking, transportation, facility, highway, availability, location, airport, infrastructure, mile
11	Housing affordability	3 \| 4	3.57	Local, housing, program, commission, afford, income, service, need, support, improve
12	Innovation cluster	2	3.55	Innovation, occupation, cluster, industry, employment, growth, job, economy, competition, concentration
13	Rural and agricultural area development	2	3.52	Agriculture, school, lack, opportunity, quality, availability, rural, resource, farm, tax
14	Citizen participation	3 \| 4	3.38	Encourage, citizen, public, participate, support, mayor, service, local, improve, provide
15	Business district	2	3.36	Business, district, strategy, resource, support, industry, plan, govern, partner, priority
16	Creative class and event organization	3	3.31	People, event, corridor, provide, pioneer, creativity, image, collaborate, organize, discuss
17	Economic analysis	2	3.18	Employment, growth, rank, data, statistics, poverty, report, measure, project, analysis
18	Economic plan	1 & 2	3.18	Plan, goal, business, community, program, data, provide, implement, committee, board
19	Land use planning and environmental conservation	3 \| 4	3.11	Use, water, land, shore, plan, future, protect, resource, zoning, environment
20	Industrial cluster	2	2.94	Cluster, project, new, business, industry, technology, sector, capital, education, asset
21	Job creation and income growth	3	2.90	Income, job, growth, wage, age, live, provide, economy, local, family
22	University and technological transfer	2	2.88	University, foundation, director, fund, research, education, technical, innovation, commercial, product
23	Energy industry (industry targeting)	2	2.87	Oil, cluster, water, gas, energy, infrastructure, manage, public, govern, industry
24	Manufacturing industry (industry targeting)	2	2.76	Manufacturing, industry, employment, transportation, product, metal, sector, investment, capital, utility
25	Workforce development	3	2.74	Job, industry, occupation, worker, workforce, growth, labor, employ, population, school
26	Port development	1 & 2	2.62	Industry, port, business, job, energy, opportunity, federal, fish, resource, ferry
27	Site development	3	2.57	City, site, build, support, history, program, district, investment, community, location
28	Tourism (industry targeting)	2	2.54	Tourism, industry, sector, service, transportation, market, recreation, food, visitor, retail

Note: & means this strategy is a combination of more than one wave. 3 | 4 means that by different classification system, this strategy can either belong to a broader definition for the third wave of strategies (Osgood et al., 2012), or the fourth wave of strategies that specifically focus on sustainability and social equity (Miranda & Rosdil, 1995; Stokan et al., 2020; Zhang et al., 2017). CEDS = comprehensive economic development strategy; EDA = Economic Development Administration; EDD = economic development district.

Table 3.

Examples of Development Strategies in Studied CEDSs.

Rank	Development strategy	Example CEDS	Example strategy in CEDS
1	Business and workforce infrastructure provision	Kenai Peninsula Borough 2019	“Expanded broadband access, industry-focused infrastructure”
2	Business attraction	Central Alabama 2017	“Attraction of new business development, especially those reliant on data processing and transmission and open-access fiber networks”
3	Technology industry (industry targeting)	Mid-America 2019	“Increase high-tech industry employment and firm innovation”
4	Economic development strategy	South Central Alabama 2018	“Using inventory and SWOT analysis, clearly define specific issues to be addressed so that forward progress can be made.”
5	Population change and livelihood support	South Central Alabama 2018	“Develop partnerships among all sectors to support public education and the general welfare of the community.”
6	Project and funding	Franklin 2020	“Active engagement to encourage or promote a policy, program or project”
7	Natural resource-reliance industries (industry targeting)	Northwest Missouri 2018	“Increase the participation rate of potential and existing small businesses in the targeted clusters of tourism, agribusiness, and forest and wood products…”
8	Disaster recovery and mitigation	South Central Ozarks 019	“Participate in the facilitation of the local COAD (Community Organizations Active in Disaster) to further pre-disaster recovery planning in an effort to bolster economic resilience.”
9	Basic public infrastructure provision	Region 4 2020	“To provide adequate public sewer service to all parts of the Region to eliminate health hazards and comply with the goals and objectives of the Federal Clean Water Act.”
10	Transportation development	Roanoke Valley-Alleghany Regional 2020	“Improve the Multimodal Transportation Network of the Roanoke Valley-Alleghany Region.”
11	Housing affordability	Tri-County Council for the Lower Eastern Shore of Maryland 2020	“Increase affordable and workforce housing options throughout the region.”
12	Innovation cluster	Northwest Pennsylvania 2016	“Develop a state-of-the-art facility, the Advanced Manufacturing and Innovation Center (AMIC), that co-locates industry, research and academia under one roof.”
13	Rural and agricultural area development	Southern Iowa Council of Governments 2020	“Develop agriculture-related technical jobs that utilize agriculture culture & knowledge.”
14	Citizen participation	Naugatuck Valley Corridor 2021	“Encourage growth and participation in the Philanthropy efforts in the NVC/EDD, through the private sector, individuals and other stakeholders’ efforts and contributions.”
15	Business district	North Central Illinois 2017	“Promote and utilize economic tools including, but not limited to enterprise zones, business districts, tax increment financing districts and port districts to encourage economic growth.”
16	Creative class and event organization	East Central Iowa 2021	“Increase investments in cultural/arts events, organizations, and institutions.”
17	Economic analysis	Central Arkansas 2019	“Use data analysis to meet the labor force needs of business and industry throughout the six-county service area”
18	Economic plan	Mo-Kan 2020	“Provide technical assistance for community and economic planning to improve quality of life …”
19	Land use planning and environmental conservation	North Olympic Peninsula Resource Conservation & Development Council 2016	“Encourage adaptable land use/zoning/ code to encourage innovative business and housing development.”
20	Industrial cluster	Central New York 2015	“Promote industrial clusters through a regional research and technology collaborative.”
21	Job creation and income growth	Cape Codd 2019	“Focus educational and workforce training programs on our existing and emerging industries, such as marine sciences and technology, to support job growth, wage equity, and economic resilience through diversity.”
22	University and technological transfer	Metropolitan Area Planning Council 2015	“Support research programs and public/private collaboration at public post-secondary universities.”
23	Energy industry (industry targeting)	Metropolitan Area Planning Council 2015	“Grow the clean energy and clean tech economy in the region.”
24	Manufacturing industry (industry targeting)	Metropolitan Area Planning Council 2015	“Support and grow the Advanced Manufacturing industry within the region.”
25	Workforce development	The Tri County Council for Western Maryland 2019	“Provide specialized industry-specific training to ensure that new businesses have a knowledgeable workforce in a minimal timeframe.”
26	Port development	Coos, Curry, and Douglas Counties 2019	“Port of Port Orford redevelopment project, including building, and infrastructure.”
27	Site development	Southern Tier West Region 2019	“Improve the region's “shovel readiness” through activities including site development / industrial park and acreage expansion / infrastructure expansion / permitting / development along rail and highway corridors and broadband backbone corridor / interstate interchange development / spec building development / building redevelopment, all consistent with smart growth principles.”
28	Tourism (industry targeting)	Georgia Mountains 2020	“Continue to support the region's tourism sectors and assist every community in promoting their tourism assets.”

Note: Keywords (consistent with those listed in Table 2) are in bold. CEDSs = comprehensive economic development strategies.

These two strategies each account for about 6% of the text in all CEDSs, and significantly more than any other strategy. This finding echoes some prior studies that also identified infrastructure provision and business attraction to be among those most commonly adopted by local jurisdictions (Olberding, 2002; Reese, 2006; Zhang et al., 2017). However, at the state level, Bradshaw and Blakely (1999) found these two strategies to be of a lower rank compared to business retention, workforce development, international trade, and entrepreneurship. According to the classification of Osgood et al. (2012), infrastructure provision is a first-wave strategy; however, since the infrastructure provision here includes a component on workforce training, which is a third-wave strategy, I concluded the top strategy as a coupling of the first and third waves. The second-ranking strategy—business attraction—belongs to the second wave. A summary of the waves to which each of the 28 strategies belong is also included in Table 2.

Industry targeting also stands out among the two most prominent strategies. Specifically, in CEDSs, several industries have been targeted. These include the technology industry (5% texts), natural resource-reliance industry (4%), energy industry (3%), manufacturing industry (3%), and tourism (3%). Taken together, industry targeting is the most prevalent strategy, consistent with findings of previous studies (Bradshaw & Blakely, 1999; Olberding, 2002). Even separated by industries, the keen pursuit of the high-tech industry and resource-reliance industries is still among the top 10 strategies. This echoes the findings of Olberding that more jurisdictions are in pursuit of newer high-tech industries, and relatively fewer (but still a significant number) target traditional industries. Industry targeting is primarily a second-wave strategy (Saiz, 2001), but can belong to the third wave if coupled with the business incubator, microenterprise program, or other small business support systems, and to the first wave if incentives and subsidies are the main tools adopted to support these industries (Osgood et al., 2012).

Other top strategies include basic public infrastructure provision and transportation development, which are first-wave strategies; economic development strategy and project and funding, which are a combination of first and second waves; and population change and livelihood support, and disaster recovery and mitigation, which are either third-wave or fourth-wave strategies, if the fourth wave is recognized as a separate wave.

Three new strategies are identified, different from most configurations in prior studies. These include rural and agricultural area development, citizen participation, and creative class and event organization. I define rural and agricultural area development as a second-wave strategy, which emerged after the EDA prioritized economically distressed areas of the nation. Citizen participation is a third- or fourth-wave strategy, and a practice of progressive development (Goetz, 1994; Miranda & Rosdil, 1995). Finally, creative class and event organization is an increasingly popular strategy among jurisdictions (Hatcher et al., 2011). This strategy originated from the second-wave industrial and innovation clusters; but with a focus on human activities, it is primarily a third-wave strategy.

Overall, the EDDs’ aggregated development strategy portfolio suggests a combination of various waves. The majority of the strategies belongs to the second wave and account for about 47% of the texts, followed by the third wave (about 35% of the texts). I let two waves equally share the texts if one strategy is a combination of both. Still, several first-wave strategies remain in the portfolio and account for about 18% of the texts. This finding lends support to both strategy progressiveness and strategy mixing (Hanley & Douglass, 2014; Kassekert & Feiock, 2009; Lowe & Feldman, 2018). On the one hand, three waves coexist, showing strategy mixing. On the other, first-wave strategies no longer dominate the strategy portfolio, and jurisdictions are now leaning toward the latter two waves, indicating potential progressiveness in development strategies.

Why do most EDDs combine different waves of strategies? A potential explanation, following Kassekert and Feiock (2009)'s framework of economic, political, and administrative incentives, is that either political or administrative motivations are driving this phenomenon. Politically, different waves of strategies appear to have different interest groups supporting them and therefore bundling may be appealing to broader constituents. Administratively, previously implemented past waves of strategies may become a path-dependent part of the strategy bundle, for at least some time. However, to confirm whether these are the underlying reasons, future studies are necessary.

Beyond the classic three-wave typology (Osgood et al., 2012), the fourth wave focusing on sustainability and social inequity (Miranda & Rosdil, 1995; Stokan et al., 2020; Zhang et al., 2017) warrants a separate wave in the strategy portfolio. Fourth-wave strategies include population change and livelihood support (i.e., supporting the poor and increasing their income), disaster recovery and mitigation, housing affordability, citizen participation, and land use planning and environmental conservation. Osgood et al. (2012) classified them into the third wave and let the third-wave strategies contain two separate components—small business development activities and community development activities. They placed sustainability- and social equity-related strategies into the latter. But sustainability- and social equity-related strategies can become their own separate wave since they account for about 18% of the texts, a significant portion of the CEDSs. Thus, their quantified importance in the overall portfolio justifies treatment as a new, stand-alone wave of strategies. In fact, if the four-wave typology is adopted, the second wave remains predominant, accounting for about 47% of the texts while the other three waves share the rest of the texts nearly equally, each accounting for about 17%18%.

Development Strategy Portfolios of EDDs

Each EDD employs a unique combination of the 28 strategies in its CEDS. Using the four-wave typology, I quantified their strategy portfolios by calculating the percentage of text for which each wave accounts. Then, using the optimized k-means clustering algorithm, I classified EDDs’ strategy portfolios into two major types.

The larger type is the second wave dominated strategy portfolio, which consists of 230 EDDs with development strategies geared primarily toward the second-wave strategies. In fact, on average, 58% of the texts in their CEDSs are second-wave strategies. The first and third waves each account for about 15%–17% of the texts, and the fourth wave is the least important, accounting for only 12% of the texts.

The second type consists of 135 EDDs with more balanced development strategies. The second-wave strategies still account for the largest share of texts (34%) in CEDSs in this group, but they are less dominant. The fourth-wave strategies follow next, with 27% of the texts. The first and third waves share the rest of the texts, accounting for 20% and 18%, respectively. This portfolio is progressive, with acute concerns for sustainability and equity, and at the same time, providing infrastructure and promoting industries. It shows a more balanced combination of all four waves of strategies.

These findings are consistent with prior studies that found a large number of local jurisdictions adopting opportunistic development strategies focused on business attraction and marketing (i.e., second-wave strategies; Reese, 2006). The identification of a significant group of local jurisdictions increasingly interested in social equity and environmental sustainability is consistent with the findings of Stokan et al. (2020) and Zhang et al. (2017). This group of EDDs adopts a more balanced strategy portfolio across four waves, which may lend some support to Rubin's (1988) idea that local officials and practitioners try nearly everything to promote development. However, only about one-third of EDDs belong to this group, which is not as common as Rubin suggested and Green et al. (1996) identified.

These different strategy portfolios are promising future directions for researchers to explore: Why do different EDDs choose different portfolios? How do they combine different strategies into a portfolio? What economic development outcomes are achieved through various portfolios? In the following section, I explore a few tentative explanations, but a fuller understanding of these questions calls for future studies.

The Geographical Distribution of Strategy Portfolios

The geographical distribution of the two portfolio types is visualized in Figure 1, which exhibits significant spatial homophily. For example, patches of second wave dominated strategy portfolios are present in the south and along the coast of the Gulf of Mexico (Florida, Mississippi, Louisiana, and Texas), north-middle (Minnesota, North and South Dakota), and southwest (Arizona, New Mexico, and Utah). Similarly, smaller patches of the balanced strategy portfolios show up in Montana, Nebraska, Arkansas, Iowa, Vermont, and New Hampshire. A formal spatial autocorrelation test reveals a positive Moran's I index of 0.06, statistically significant at the 0.1% level.

Three reasons may have accounted for the spatial homophily. First, geographical characteristics may have dictated the adoption of a strategy portfolio. For example, environmentally sensitive areas may be inclined to adopt the balanced portfolio with a focus on the fourth-wave strategies to promote sustainable development. Similarly, a dire economic condition can cause a jurisdiction to lean toward second-wave probusiness strategies. To test this hypothesis, I collected the unemployment rate for each county in 2018, when the majority of the EDDs published their current CEDSs and conducted a correlation analysis between a county's unemployment rate and its adoption of different strategy portfolios. The results show a statistically significant negative correlation between the adoption of the balanced strategy portfolio with the unemployment rate. A logistic regression further quantifies that when the unemployment rate increases by 1%, the possibility of a county adopting the balanced strategy portfolio decreases by 21%. This relationship is statistically significant at the 5% level.

Second, state strategy adoption may affect local adoption. Most states have developed state-level CEDSs, and many EDDs have consulted or even cited the state strategies in their own strategy development. Thus, the type of strategy portfolio a state adopts may significantly influence EDDs to adopt the same type of strategy portfolio. Many spatial homophiles in Figure 1 are exhibited along state borders, lending some support to this explanation.

Third, EDA field office leadership may have shaped local adoption of strategies. For example, Figure 1 and the EDA regional offices’ territory, show that states led by the Atlanta Regional Office (AL, FL, GA, KY, NC, SC, and TC) and Austin Regional Office (LA, AR, NM, OK, TX) are almost all dominated by second-wave strategies, with only KY and AR as exceptions. In contrast, states corresponding to the Chicago and Denver regional offices mostly adopted the balanced strategy portfolio, with a few exceptions such as MN and ND. These preliminary observations lend support to this explanation, but follow-up interviews with EDA regional directors are needed to verify whether this explanation holds, and to understand how EDA regional offices may affect local adoption of strategies.

Most studies exploring topologies of development strategies did not examine the geographical distributions of these typologies (Clarke & Gaile, 1992; Hanley & Douglass, 2014; Miranda & Rosdil, 1995; Reese, 1993, 2006). For example, Olberding (2002) listed characteristics of different MSAs’ development strategy portfolios in an appendix table, but has not explicitly pointed out the role geographical factors may have played in forming these portfolios. Similarly, Bradshaw and Blakely (1999) listed different expenditure patterns for economic development programs in various states in a table, but did not explore the geographical factors and spatial homophily in those patterns. In contrast, a few recent studies such as Kassekert and Feiock (2009), Osgood et al. (2012), and Stokan (2013) have considered local characteristics to explain the bundling of development strategies. For example, Kassekert and Feiock (2009) found that, with the increase in city council size, long-term debt, and economic development capacity, more incentives will be adopted. Similarly, Osgood et al. (2012) examined how community characteristics affected the probability of adopting different waves of strategies. They found that communities with a larger population and higher rates of poverty and racial minorities are more likely to adopt third-wave strategies. (Their third wave is broadly defined to include both the third and fourth waves in this paper.) Stokan et al. (2020) explicitly examined the factors that propel local governments to prioritize equity-oriented development strategies. They found that fourth-wave strategies are more prevalent among local governments facing less competitive pressure, having greater resource capacities, and experiencing more extensive intergovernmental involvement in the economic development planning process. This study has taken a step toward mapping and statistically verifying the spatial homophily in the adoption of different strategy portfolios. The causes of spatial homophily are still in need of future empirical studies.

Conclusion and Practical Implications

Typology is a succinct way of summarizing the wide range of development strategies adopted by local jurisdictions. This paper adopts a text mining method to analyze the CEDSs of EDDs across the nation and quantifies the adoption of four waves of economic development strategies. Three main findings emerge. First, 28 major strategies are identified, with business and workforce infrastructure provision, business attraction, and technology industry targeting as the top three. In addition, new strategies are revealed: rural and agricultural area development, citizen participation, and creative class and event organization. These are not common strategies identified by extant studies but have started to emerge among the most recent CEDSs. Second, the fourth wave of equity- and sustainability-oriented economic development strategies is on the rise, accounting for about 18% of the texts in CEDSs. In the current CEDS landscape, the fourth wave shares a similar percentage of texts with the first and third waves of strategies, while the second wave dominates with about half the texts. Third, EDDs do not focus on only one wave of strategies. Instead, they combine various waves to form a comprehensive strategy portfolio. Two types of portfolios have been identified: Two-thirds of the EDDs adopt a second-wave portfolio focusing mostly on the second-wave strategies while adopting a small percentage of each of the other three waves. The other one-third adopts a portfolio that is more balanced across waves and places an emphasis on both the second- and fourth-wave strategies.

Practical lessons are offered. Local economic development planners should carefully evaluate whether their strategy portfolio is put together with strategies that complement each other, and they should understand the underlying reasons that cause them to choose this combination of strategies. Moreover, frequent regional collaboration in planning activities and development-related regional conventions can help form interpersonal and intergovernmental networks that promote the diffusion of novel and future-oriented development strategies.

This paper also shows that text mining can be a useful method to keep track of strategies and strategy portfolios at the local level. This tool saves from having to read a large volume of local CEDSs, and/or administering surveys to local economic development planners. Moreover, it is responsive to new strategies because bottom-up changes in local CEDSs can be immediately and automatically detected. In contrast, in a survey instrument, the sensitivity to new strategies is often affected by survey designers’ perceptions of prevailing strategies. In most surveys, survey designers list different strategies for local planners to check whether they have adopted them. If the survey designs do not correctly perceive emerging trends, it would be difficult for the survey to reveal them. Even when the designers intentionally add an open-ended option, the well-documented inertia in survey responses is likely to lead to a significantly higher rate of nonresponse (Emde, 2014). Moreover, text mining allows easy replication of the analysis over time with each round of new CEDS release, and therefore affords a consistent way of detecting the shift of strategy landscape over time. In comparison, repeating surveys is costly, and the inconsistency in questions (which is inevitable if they are to reflect new trends in development strategies) also undercuts the comparability of results. This study shows that text mining is a useful tool to detect both well-established and emerging development strategies. I hope more researchers will adopt this tool to conduct dynamic strategy studies.

A few questions emerging from this study beg for future examination. For example, what specific characteristics of EDDs have led them to adopt or postpone their adoption of the fourth-wave strategies? What are the specific considerations when EDDs put together their strategy portfolios? Which strategies and strategy portfolios have led to more desirable economic outcomes? These follow-up questions emerge naturally as this study unfolds, but they go beyond the scope of a single paper. I hope to see other researchers take on the task of answering them and further completing an understanding of the development strategy landscape.

Footnotes

Acknowledgment

The author would like to thank her research assistant Paula Perez for help collecting the comprehensive economic development strategies from the websites of economic development districts.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article

ORCID iD

Li Fang

Notes

Author Biography

Li Fang is an assistant professor in the Department of Urban and Regional Planning at the Florida State University. Her research interests include economic development and land use policies, with a particular emphasis on urban innovation.

References

Arun

Suresh

Veni Madhavan

C. E.

Murthy

M. N. N.

(2010). On finding the natural number of topics with Latent Dirichlet Allocation: Some observations. In Zaki

M. J.

J. X.

Ravindran

Pudi

(Eds.), Advances in knowledge discovery and data mining (pp. 391‐402). Springer Berlin Heidelberg.

Bradshaw

T. K.

Blakely

E. J.

(1999). What are “third-wave” state economic development efforts? From incentives to industrial policy. Economic Development Quarterly, 13(3), 229‐244. doi:10.1177/089124249901300303

Cao

Xia

Zhang

Tang

(2009). A density-based method for adaptive lda model selection. Neurocomputing — 16th European Symposium on Artificial Neural Networks, 72(7–9), 1775‐1781. doi:10.1016/j.neucom.2008.06.011

Clarke

S. E.

(1993). The profound and the mundane: Analyzing local economic development activities. Urban Geography, 14(1), 78‐94. doi:10.2747/0272-3638.14.1.78

Clarke

S. E.

Gaile

G. L.

(1992). The next wave: Postfederal local economic development strategies. Economic Development Quarterly, 6(2), 187‐198. doi:10.1177/089124249200600207

Currid-Halkett

Stolarick

(2011). The great divide: Economic development theory versus practice-a survey of the current landscape. Economic Development Quarterly, 25(2), 143‐157. doi:10.1177/0891242410394358

Deveaud

SanJuan

Bellot

(2014). Accurate and effective latent concept modeling for ad hoc information retrieval. Document Numérique, 17(1), 61‐84. doi:10.3166/dn.17.1.61-84

Dewar

M. E.

(1998). Why state and local economic development programs cause so little economic development. Economic Development Quarterly, 12(1), 68‐87. doi:10.1177/089124249801200106

Eisinger

P. K.

(1989). The rise of the entrepreneurial state: State and local economic development policy in the United States. University of Wisconsin Press.

10.

Elisinger

(1995). State economic development in the 1990s: Politics and policy learning. Economic Development Quarterly, 9(2), 146‐158. doi:10.1177/089124249500900204

11.

Emde

(2014). Open-ended questions in web surveys–Using visual and adaptive questionnaire design to improve narrative responses. Working paper.

12.

Etienne

Praene

J. P.

Leducq

Gatina

J. C.

(2019). A text mining approach to investigate urban planning documents: Study case of Reunion Island. In 59th ERSA Congress Cities, regions and digital transformations: opportunities, risks and challenges.

13.

Fang

Ewing

(2020). Tracking our footsteps: Thirty years of publication in JAPA, JPER, and JPL. Journal of the American Planning Association, 86(4), 470‐480. doi:10.1080/01944363.2020.1766994

14.

Feser

Renski

Goldstein

(2008). Clusters and economic development outcomes: An analysis of the link between clustering and industry growth. Economic Development Quarterly, 22(4), 324‐344. doi:10.1177/0891242408325419

15.

Feser

E. J.

Koo

Renski

Sweeney

S. H.

(2001). Incorporating spatial analysis in applied industry cluster studies. Working paper. https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.122.7444&rep=rep1&type=pdf

16.

Gobster

P. H.

(2014). (Text) mining the LANDscape: Themes and trends over 40 years of landscape and urban planning. Landscape and Urban Planning, 126(June), 21‐30. doi:10.1016/j.landurbplan.2014.02.025

17.

Goetz

E. G.

(1994). Expanding possibilities in local development policy: An examination of US cities. Political Research Quarterly, 47(1), 85‐109. doi:10.1177/106591299404700105

18.

Green

G. P.

Fleischmann

Kwong

T. M.

(1996). The effectiveness of local economic development policies in the 1980s. Social Science Quarterly, 77(3), 609‐625. https://www-jstor-org.web.bisu.edu.cn/stable/42863505

19.

Green

T. F.

(2014). Cluster planning and cluster strategy in regional economic development organizations. University of Illinois at Urbana-Champaign. Dissertation.

20.

Griffiths

T. L.

Steyvers

(2004). Finding scientific topics. Proceedings of the National Academy of Sciences, 101(suppl 1), 5228‐5235. doi:10.1073/pnas.0307752101

21.

Hanley

Douglass

M. T.

(2014). High road, low road, or off road? Economic development strategies in the American states. Economic Development Quarterly, 28(3), 220‐229. doi:10.1177/0891242414534726

22.

Hatcher

Oyer

Gallardo

(2011). The creative class and economic development as practiced in the rural U.S. South: An exploratory survey of economic development professionals. Review of Regional Studies, 41(2–3), 139‐159. doi:10.52324/001c.8153

23.

Hill

E. W.

Brennan

J. F.

(2000). A methodology for identifying the drivers of industrial clusters: The foundation of regional competitive advantage. Economic Development Quarterly, 14(1), 65‐96. doi:10.1177/089124240001400109

24.

Zheng

(2021). Smart city initiatives: A comparative study of American and Chinese cities. Journal of Urban Affairs, 43(4), 504-525. doi: 10.1080/07352166.2019.1694413

25.

Jelodar

Wang

Yuan

Feng

Jiang

Zhao

(2019). Latent Dirichlet Allocation (LDA) and topic modeling: Models, applications, a survey. Multimedia Tools and Applications, 78(11), 15169‐15211. doi:10.1007/s11042-018-6894-4

26.

Kassekert

Feiock

(2009). Policy tool selection: Predicting the bundling of economic development policy instruments using a multivariate probit analysis. APSA 2009 Toronto Meeting.

27.

Leicht

K. T.

Jenkins

J. C.

(1994). Three strategies of state economic development: Entrepreneurial, industrial recruitment, and deregulation policies in the American states. Economic Development Quarterly, 8(3), 256‐269. doi:10.1177/089124249400800303

28.

Lowe

Feldman

M. P.

(2018). Breaking the waves: Innovating at the intersections of economic development. Economic Development Quarterly, 32(3), 183‐194. doi:10.1177/0891242418783848

29.

Luria

Rogers

(1999). Metro futures: Economic solutions for cities and their suburbs. Beacon Press.

30.

Miranda

Rosdil

(1995). From boosterism to qualitative growth: Classifying economic development strategies. Urban Affairs Quarterly, 30(6), 868‐879. doi:10.1177/107808749503000605

31.

Morgan

J. Q.

Hoyman

M. M.

McCall

J. R.

(2019). Everything but the kitchen sink? Factors associated with local economic development strategy use. Economic Development Quarterly, 33(4), 267‐278. doi:10.1177/0891242419857152

32.

Olberding

J. C.

(2002). Diving into the “third waves” of regional governance and economic development strategies: A study of regional partnerships for economic development in U.S. Metropolitan areas. Economic Development Quarterly, 16(3), 251‐272. doi:10.1177/089124240201600305

33.

Osgood

J. L.

Opp

S. M.

Bernotsky

R. L.

(2012). Yesterday’s gains versus today’s realties: Lessons from 10 years of economic development practice. Economic Development Quarterly, 26(4), 334‐350. doi:10.1177/0891242412465002

34.

Peters

Fisher

(2004). The failures of economic development incentives. Journal of the American Planning Association, 70(1), 27‐37. doi:10.1080/01944360408976336

35.

Porter

M. E.

(2000). Location, competition, and economic development: Local clusters in a global economy. Economic Development Quarterly, 14(1), 15‐34. doi:10.1177/089124240001400105

36.

Presser

Traugott

(1992). Little white lies and social science models correlated response errors in a panel study of voting. Public Opinion Quarterly, 56(1), 77‐86. doi:10.1086/269296

37.

Reese

L. A.

(1993). Categories of local economic development: An empirical analysis. Policy Studies Journal, 21(3), 492‐506. doi:10.1111/j.1541-0072.1993.tb01804.x

38.

Reese

L. A.

(2006). Do we really need another typology? Clusters of local economic development strategies. Economic Development Quarterly, 20(4), 368‐376. doi:10.1177/0891242406291630

39.

Rubin

H. J.

(1988). Shoot anything that flies; claim anything that falls: Conversations with economic development practitioners. Economic Development Quarterly, 2(3), 236‐251. doi:10.1177/089124248800200304

40.

Saiz

(2001). Using program attributes to measure and evaluate state economic development strategies. Economic Development Quarterly, 15(1), 45‐57. doi:10.1177/089124240101500104

41.

Stokan

(2013). Testing rubin’s model 25 years later: A multilevel approach to local economic development incentive adoption. Economic Development Quarterly, 27(4), 301‐315. doi:10.1177/0891242413491315

42.

Stokan

Deslatte

Hatch

M. E.

(2020). Exploring the trade-offs local governments make in the pursuit of economic growth and equity. Urban Affairs Review, online first.

43.

Stone

A. A.

Bachrach

C. A.

Jobe

J. B.

Kurtzman

H. S.

Cain

V. S.

(eds.). (1999). The science of self-report: Implications for research and practice. Psychology Press.

44.

U.S. Economic Development Administration (EDA) (2019). https://www.eda.gov/edd/ and https://www.eda.gov/ceds/.

45.

Vanderleeuw

J. M.

Sandovici

M. E.

Jarmon

C. A.

(2011). Women city leaders and postmaterialist values: Gender differences in economic development priorities. Journal of Women, Politics and Policy, 32(3), 211‐236. doi:10.1080/1554477X.2011.589289

46.

Zhang

Warner

M. E.

(2017). Business retention and expansion and business clusters–A comprehensive approach to community development. Community Development, 48(2), 170‐186. https://doi.org/10.1080/15575330.2017.1285332

47.

Zhang

Warner

M. E.

Homsy

G. C.

(2017). Environment, equity, and economic development goals: Understanding differences in local economic development strategies. Economic Development Quarterly, 31(3), 196‐209. doi:10.1177/0891242417712003