ABSTRACT. As shown by the declining prevalence of Regional Economics programs in US universities, it is now more important to interest the next generation of regional scientists at an earlier educational level. The following course outline gives students a solid foundation in regional economic concepts using their own communities as the analytical subject. This hybrid project-based learning approach gives students knowledge of several methods of regional economic analysis. Economic base theory, location quotients, coefficients of specialization, pull factors, and input-output analysis are demonstrated. Students incorporate these concepts into an analysis of their home county. This course has inspired student interest to the extent that many have obtained graduate degrees or employment in regional economics. It is our hope that other economics programs who do not have a version of this class can adopt this method of teaching regional economics.

ABSTRACT. Charter schools have become one of the most popular school choice programs in recent years. While being accused of endangering the fiscal health of local school districts, charter school proliferation could also be caused by the fiscal distress of local school districts. Yet, the the reverse causation has not been fully studied in the literature. Using data of local school districts in Ohio from FY 2006 to 2013, I adopt Instrumental Variable approach to investigate the possible scenario that the rapid growth in charter school enrollment could be caused by fiscal distress of school districts. The preliminary results seem to suggest that a fiscally distressed school district would fail to sustain the desirable level of education expenditures, resulting in lower education inputs and lower student achievement. Students could be left with no better options but choosing charter schools in those local school district, especially for the case of rural school districts in Ohio.

ABSTRACT. Nowadays the role of Higher Education Institutions (HEIs) goes beyond education and research. They have become also an important economic actor also for regional economic development. In the academic area, recent studies have attempted to understand, qualify, and quantify this impact process, mostly based on case studies. However, few impact studies are available on a more generalizable level, especially for Brazil. That way, this study aims to estimate the long run consequences of the presence of HEIs in the Brazilian municipalities. For this purpose, the quasi-experimental approach is used to test if the emergence of HEIs has changed the path of regional economic development variables, as GDP per capita and labor income, over the last 30 years. In this sense, the presence of HEIs in the municipality would represent the treatment variable and the model is estimated for 5572 municipalities of the Brazilian economy, for the period 1985 to 2015.

ABSTRACT. The promotion of multi-jurisdictional regional approaches for development have been used since the 1920s. Metropolitan Councils of Government and Regional Planning Commissions were incentivized and promoted in States from 1950s on. State agencies historically used sub-state districts, through the boundaries would vary agency to agency. By Executive Order in 1972, Virginia Governor A. Linwood Holton required all State agencies using sub-state districts to use the 1968 defined Planning Districts or multiples. This administrative action enabled a rich regional data base for agencies and the public, as all data was grouped by the PD number which acted as a FIPS region code. Though most states have sub-state district systems, there is no consistent numbering system that may serve as a FIPS code. There is no such Census Bureau code. As a consequence, the only sub-state national data set of regions is at the County level. The only multi-jurisdictional data set is the MSA, but it is a market region that expands its boundaries based on commuting, rarely supporting long term analysis for this reason. In the 1970s, Eurostat began the Nomenclature of Territorial Units for Statistics (NUTS) classification system for dividing up the EU territory in order to produce regional statistics for the community. The author developed a geocode prototype that would enable building a U.S. sub-state region database that enables as well combinations within and across state lines. It is proposed that an effort begin now for a system that would be ready for use of 2020 Census data.

ABSTRACT. Residential and employment location decisions for working households frequently are commingled. Economic and social factors such as job accessibility, wage differentials, demographics, housing markets, travel times, trip-chaining opportunities, dual employment families and other quality of life considerations influence where a household ultimately chooses to locate. These choices in turn shape commuting and migration patterns within a region. The growing availability of origin-destination employment data has allowed for increasingly textured analyses of these joint residential-employment decisions across the rural-urban continuum. Our analysis adds to this research by considering the phenomenon of long-distance commuting, or those individuals who travel 50 miles or more between their place of residence and place of employment (i.e. “stretch commuters”). Specifically, we analyze the determinants of long distance commuters and whether period effects stemming from the Great Recession or broader factors attributed to regional economic restructuring have influenced patterns of stretch commuting. Understanding these patterns across this economic period offers insights to the evolving factors that influence the substitution of commuting for migration and the changing geographic extents of regional economic influence.

In considering this period, we model determinants of stretch commuters by considering the interactions of: 1) individual characteristics of the commuter in terms of age, earnings and industry of employment; 2) the economic, labor market and housing characteristics of the region receiving the commuter; and 3) the economic, labor market and housing characteristics of the region sending the commuter. To overcome methodological issues associated with using county-level data and Euclidean measurements to estimate commuting distances, we calculate stretch commuters using the Census Bureau’s LEHD Origin-Destination Employment Statistics (LODES) at the sub-county level using network analysis.

ABSTRACT. Cities are the most unequal places in the world. In a broad review of the concentration of poverty in central cities in the U.S., Glaeser, Kahn and Rappaport [2008] test out various determinants of residential sorting by income and attribute the urban concentration of poverty mainly to better access through public transportation. While a majority of Americans commute by car, congestion and maintenance costs leave cars too costly for the urban poor, who have an alternative in public transit. This paper investigates how the distribution of urban public transit services affects the location choices of the rich and the poor in U.S. cities. In particular, under what circumstances might improved public transit access induce even more spatial segregation by income?

Public transit services are atypical of most urban public goods in that they disproportionately benefit low-income households over high-income ones, such as through increased access to labor market opportunities and more desirable residential neighborhoods further away. So a well-connected transit network has the potential to reduce income segregation over space by reducing commuting costs for the poor. However, another distinguishing feature of public transit networks is that they tend to be sparsely connected relative to alternative travel networks. Neighborhoods that are relatively more accessible by private vehicles may be relatively inaccessible by public transit (and vice versa). If high-income households prefer more driving accessibility and low-income households prefer more transit accessibility, such a public transit network might exacerbate segregation by income. On the other hand, if the sparse public transit network is uniformly distributed over the city’s driving network (so that driving and transit access are positively correlated), it might induce more income diversity across neighborhoods.

A rigorous modeling of transit-induced population sorting and income inequality has yet to be undertaken. Existing spatial models of travel mode and location choice induce perfect income segregation by construction (LeRoy and Sonstelie [1983], DeSalvo and Huq [1996]), making them unsuitable for studying levels of segregation. And existing studies of residential segregation have typically focused on public amenities that only vary along one dimension across a fixed ordering of neighborhoods. I combine aspects of both literatures and propose a new theoretical framework to study income segregation as a function of accessibility along multiple modes of travel and to formalize my conjectures above.

To reinforce this theory and determine the magnitude of transit’s role in income segregation, I provide empirical evidence from 142 large metropolitan areas I construct measures of income mixing using blockgroup-level demographic data from the 5-year American Community Survey. Based on household trip characteristics observed in the National Household Travel Survey from 2009, I use Google’s Places API to identify potential trip destinations for each trip type from each neighborhood. I then simulate these trips on Google Maps to measure neighborhood accessibility along the walking, driving and transit networks in my cities. We know that mode choices vary significantly by income and income compositions vary significantly across neighborhoods in a city. This paper shows that a non-trivial share of this variation can be explained by differences in accessibility for drivers and transit riders.

References Joseph S. DeSalvo and Mobinul Huq. Income, residential location, and mode choice. Journal of Urban Economics, 40:84-99, 1996. Edward L. Glaeser, Matthew E. Kahn, and Jordan Rappaport. Why do the poor live in cities? The role of public transportation. Journal of Urban Economics, 63(1):1-24, 2008. Stephen F. LeRoy and Jon Sonstelie. Paradise lost and regained: Transportation, innovation, income and residential location. Journal of Urban Economics, 13(1):67-89, 1983.

ABSTRACT. This paper uses input-output analysis to estimate CO2 emissions of embodied freight transportation for each industry in the U.S. from 2007 to 2012. Demand for freight transportation is accelerating with the globalization of increasingly fragmented production processes. More freight activities bring more CO2 emissions. I use embodied freight transportation to measure the freight transportation needs across the supply chain for each industry. The dataset used in this study are U.S. Bureau of Economic Analysis Input-Output tables and Commodity Flow Survey Data from U.S. Bureau of Transportation Statistics. Results show that some sectors have larger embodied freight transportation than the direct freight transportation, such as wholesale and professional services. Truck and rail are the dominant modes. The share of rail becomes larger in total embodied freight. In addition, CO2 emissions of freight are estimated so that environmental mitigation strategies can be tailored to industries.

ABSTRACT. Suburban development once again has established itself as the principle form of metropolitan growth (Brookings Institute 2017). In Sunbelt states, such as Georgia, this growth is happening on a large peri-urban periphery, potentially effecting counties that until recently were far from the urban core. For transportation planners, policy makers, and managers, the effect of rural areas in a state “trending urban” presents unexpected administrative operational challenges, due, in part, to the bifurcation of urban and rural transit funding between the Urbanized Area Formula Funding program and the Formula Grants for Rural Areas program. The baseline for funding for these Rural Areas program programs is determined primarily by a state’s population residing outside of Census defined Urbanized Areas, and is primarily administered by state departments of transportation. The federal Urbanized Area funding program is based on the population in Census defined Urban Areas, which are continuous areas with Decennial Census populations of over 50,000, with urban land cover and minimum population densities. This program has various levels and means of funding, and while the planning and administration requirements for Small Urban areas with populations under 200,000 are not significantly more onerous than the rural programs, there is a steep change for Urbanized Areas with populations over 200,000 that should administer programs through their own metropolitan transit agency. This potential discontinuity is especially challenging in rapidly urbanizing regions where principally rural counties can be absorbed in to large urban areas between Decennial Censuses. Accordingly, there is a need to create accurate and accessible methods for policy makers to estimate new Urban Areas and future funding scenarios based on population projections and an understanding of land use patterns in the near future. To address this challenge, we use 1) population projection methods to estimate changes in Urbanized Area Formula Funding program and the Formula Grants for Rural Areas program for Georgia, and 2) logistic and spatial Poisson methods to estimate changes in urbanized areas at the county level in Georgia for 2020. Our results will validate a methodology for policy makers to incorporate demographic and land use change modeling methods for rapid assessments of future planning scenarios in rapidly urbanizing regions.

ABSTRACT. CUSHING SPECIAL SESSION Opioid overdose is the leading cause of unintentional death in the U.S. Narcan TM (Naloxone) is a prescription medicine that can reverse overdose effects. This research investigates the effect of Naloxone access laws on overdose death rates using state and temporal variation in the enactment of these laws. We also explore possible spillover effects of Naloxone access laws and overdose death rates across states. Our analyses reveal that when broken down by access law provisions, there exists a mixture of positive and negative effects on overdose death rates depending upon the provision. The results indicate that Naloxone access provisions have regional impacts by influencing overdose death rates within the state enacted and have a spillover effect in neighboring states. The magnitude of spillover effects is larger than direct effects in the states. Looking across multiple provisions, our findings provide no statistical evidence that these laws reduce opioid death rates.

ABSTRACT. The opioid epidemic in the United States has gained widespread public attention, but this attention has focused mainly on its effect on adults and those who have died of overdoses. There has been less consideration of the potential effects of the opioid crisis on children. However, it is likely that millions of children have been affected, either directly—through parental addiction or death—or indirectly through the experiences of friends, classmates, or neighbors.

In this analysis, we will combine county-level data from the U.S. Centers for Disease Control and Prevention and the U.S. Census Bureau to investigate state and regional patterns and trends in children’s potential exposure to drug overdose deaths—especially accidental deaths from opioid use. Data will be disaggregated by race/ethnicity, age, and poverty status to determine whether children in certain population subgroups are at higher risk of being exposed to drug overdose deaths than others. We will also classify counties by metro/nonmetro status to investigate whether children’s risk of exposure is higher in rural areas—where the opioid crisis has received the most attention.

Data from 2012-2016 will be compared with data from 1998-2002 to determine whether children’s potential exposure to drug overdose deaths has increased over time, and to compare trends across different geographic areas and population subgroups. We will also investigate changes in children’s living arrangements during this time period (e.g., living with grandparents, other relatives, or foster parents) to see if there are more children in nonparental care in hard-hit areas.

Our goal is to identify groups of children who are the most vulnerable so that policymakers, health professionals, school administrators, and others can devise programs and strategies to meet their needs.

ABSTRACT. This paper focuses on the variation in mortality rates across gender, age cohorts, and sub-state region in Kentucky during the period 1999-2016. The analysis considers mortality rates for the major components of both natural causes and external causes. The discussion is conducted at three levels, Appalachian versus non-Appalachian, sub-state regions, and county level. The county level analysis examines the extent to which mortality rates are correlated with socioeconomic and behavioral variables. The current paper expands previous work by considering gender and the age cohorts from 15-24 through 75-84 years.

ABSTRACT. The rise in drug overdose deaths in the United States since the turn of the millennium has been extraordinary. Changes in the medical profession’s view of prescription opioids have dramatically increased the availability of potent drugs, but concurrent increases in suicides and liver disease suggest a demand element to the crisis. In particular, drug overdoses are heavily concentrated among those with a high school degree or less, a population that has seen its declines in economic opportunity, physical heath, and relationships in recent decades (Case and Deaton 2017). Some evidence exists linking employment to despair deaths (Hollingsworth 2016; Pierce and Schott 2016). We further this literature by examining the role of wages in addition to employment and decompose employment and wage growth across high-, medium-, and low-paying industry tiers to identify heterogeneous employment and wage growth effects on overdose rates within nonmetro and metro counties. We find significant variation in the effect of employment and wage growth rates on drug overdose death rates at different points on the industry wage distribution. In general, we find that changes in wage growth rates—and in particular bottom-tier wage growth rates—are more important than employment growth rates for nonmetro counties. Changes in both employment and wage growth rates are important for metro counties, but the effects are strongest for bottom-tier employment growth and top-tier wage growth. Lastly, we find effects differ between male and female overdose death rates and black and white rates.

ABSTRACT. Entrepreneurial ecosystems (EE) have become a popular topic in recent years because it provides an innovative approach to regional economic development. The concept of EE has been popularized in the literature (Herrmann et al. 2015; OECD 2013; Stangler and Bell-Masterson 2015), and recent literature shows that the local economic contributions have a significant impact on the entrepreneurship process. As an example, Sussan, Sloboda, and Hall (2018) shows that Las Vegas aspire to have vibrant entrepreneurial ecosystems that is relevant to the digital economy. Their approach carefully examined the economic history of Las Vegas, a city known for its gambling and entertainment industries, and reports the recent entrepreneurial activities in the city along with the data that measures the vibrancy of the ecosystem and growth of the EE in Las Vegas. In this paper, we carefully examine the EE of Detroit. What are the factors that drove Detroit to develop a vibrant EE? The first part of the analysis will delve into the economic history of Detroit mainly in terms of how it got started with whom, in what industry and which industry (ies) dominated over the years. The second part of the analysis will analyze through various data sources that describes the EE and related systems for Detroit in the digital economy.

Reference

Sussan, F. & Acs, Z.J. (2017). The digital entrepreneurial ecosystem. Small Bus Econ, 49 (1): 55 -73. DOI: https://doi.org/10.1007/s11187-017-9867-5 Sussan F., Sloboda B., Hall R. (2018) Is There a Path from Sin City to Tech City? The Case for Las Vegas. In: O'Connor A., Stam E., Sussan F., Audretsch D. (eds) Entrepreneurial Ecosystems. International Studies in Entrepreneurship, vol 38. Springer, Cham. DOI: https://doi.org/10.1007/978-3-319-63531-6_8

ABSTRACT. The paper explores firm behavior across increasing levels of innovation activity, and compares rural and urban firms at particular levels. Data are from the 2014 Rural Establishment Innovation Survey (N=10,913) which consists of 257 firm-level variables including 42 measures of innovation. Four categories, ranging from “non-innovative” to a “high innovative activity,” are constructed. Results from a multinomial logit (MNL) regression with rural interactions are used to compare firm behaviors across innovation activity categories and those in rural and urban areas. While MNL parameters suggest a hierarchy of innovation activity intensity exists, marginal effects show this hierarchy is less pronounced. Additionally, behaviors of rural and urban firms vary little at particular intensity levels, but there are differences in the types of industries where innovation occurs. Finally, policy simulations demonstrate how encouraging (discouraging) selected behaviors may impact the make-up of firms in each category.

ABSTRACT. In Qatar, hydrocarbons revenues still form the main part of national income. However, as the country moves forward with Qatar National Vision 2030, the government is increasingly seeking to diversify the economy away from hydrocarbons. Qatar 2030 explicitly promote transformation into knowledge-based economy to sustain growth and prosperity, and raises entrepreneurship and innovation as fundamental tools to thus shift.

Entrepreneurship and innovation activities are pillars of economic growth and the local environment is key to understand both. Entrepreneurship and innovation are not isolate phenomena but rather depend on network with various actors that have distinct roles. The government contributes a great deal towards entrepreneurship and innovation. Accordingly, policy makers are now beginning to recognise the merit of a more systems-based form of support for entrepreneurship and innovation. One emerging approach is the ‘entrepreneurial ecosystem’.

Based on an systemic perspective via the concept of “entrepreneurship ecosystem”, the main objective of this research is to identify and analyze the main dynamics and challenges of the ICT Qatar’s entrepreneurial ecosystem. Our results shows much progress has been made to boost entrepreneurial activity thank to a strong top down intervention. The Qatari government has invested heavily in education, entrepreneurial financing schemes, training and development and business incubators. Our second, result shows that the ICT-sector enjoys a far more developed ecosystem than the rest of the sectors. In fact, the ICT sector forms the “backbone” of several industries (financial services, healthcare, and retail). But this development is limited some critical obstacles: access to finance, the small size of the local market, the lack of market openness, and the weak connections and cooperation between the different stakeholders of the ecosystem.

ABSTRACT. Cities in matured economies around the world face downtown revitalization problems and there is no magic wand to solve these problems. From Detroit in the U.S. to Wakayama in Japan, downtowns have deteriorated due to aging infrastructure, crime, depopulation, aging population, and simply have become unattractive comparing to the newer modern suburbs with large retail malls and planned neighbourhoods. In spite of these obvious deteriorations, politicians and communities for varying motivations have exerted effort to revitalize downtowns through public, private, or a combination of public and private funding. While there are some examples of successful cases (e.g., Las Vegas in the U.S.), it is unclear if economic revitalization is synonymous to population revitalization. In other words, bringing in economic dollars may not be the same as bringing back the population needed to maintain economic vibrancy. This paper selected 9 such downtowns dispersed in various geographic locations in Japan and examined their longitudinal revitalization processes. Based on publicly available data that we analysed through content analyses, we concluded with a 2 (Approach: public, private) x 2 (Benefits: societal, commercial) conceptual framework with four types of revitalization strategies. We further discussed the tradeoff of commercial success and population growth. Our results have managerial implications for regional development.

ABSTRACT. We study aspects of economic growth in a region that is creative a la Richard Florida. Members of the creative class possess creative capital and they fall into one of two possible groups---they are either artists or engineers. We describe the optimal income redistribution rule that maximizes the creative class’s average steady state income. Because this average income is increasing in the physical capital per creative class member ratio, the rule requires a regional authority to redistribute income away from (towards) the group that saves a lower (higher) fraction of its income. This is a negative finding in the sense that the rule’s implementation will tend to favor the group that already saves more. Even so, the finding is consistent with the observation made by some researchers that there is a connection between income inequality and regions in which the creative class performs a large part of all economic activities.

ABSTRACT. According to the Census Bureau’s Population Estimates data, the total population in the United States increased during the year 2010 to 2016. Unlike total population trend, some metropolitan areas, such as Pittsburg and Chicago, underwent population decrease. This paper examines the impacts of individuals, amenities and labor market factors on individual’s internal migration decision. We divide the migration by its types; migration among metropolitan areas, metropolitan areas to non-metropolitan areas. We use Census Bureau’s 2016 1-year ACS data and probit regression model to estimate the motivation factors. The results show that the individuals, amenities, and labor market factors have a different impact on each migration type.

ABSTRACT. The purpose of this paper is to study how labor market characteristics affects the geographical, industrial, and occupational mobility of displaced skilled workers. Literature on the probability of reemployment has mostly focused on individual characteristics of the workers, but the jobs that the individuals will choose to take after displacement will depend on what the region has to offer. Urban economics literature argues that denser markets should create more possibilities for workers to find a (relevant) job in the same region. Moreover, following the arguments of Marshall (1920) regarding labor market pooling, the decomposition of the industries in the market, beyond market size, should also matter. In these lines, a concentration of similar industries should make it easier for individuals to find a relevant job to what they had before. A worker who loses her job in a thinner market does not only have a tougher time into getting back into employment, but also experiences more difficulties finding a relevant job matching her skills. She might thus be more willing to take any job, even if it is not suited for her skills as an alternative to unemployment, or might also be more prone to geographically relocating to get a more fitting employment. Yet, the importance of regional characteristics in reemployment possibilities of displaced workers is rather scarce in the literature. Preliminary results confirm Marshall’s hypothesis on market pooling and show that if the region hosts a large share of workers employed in the same (or similar) industries like the one the displaced workers were working at, the probability that workers move, switch occupations, or industries reduces considerably. This indicates that, these markets provide better matching to the skills of the individuals, and that dealing with displacement might be easier in such cases.

ABSTRACT. The spatial spillover of compensation is a long-examined strand of regional/spatial literature (see Anselin, 1988; Burridge & Gordon, 1981; Manning, 1994; Molho, 1982). These standard literatures use the spatial contiguity matrix. Instead of this, Járosi (2017) use the interregional interindustry trade network as the network based weight matrix and tests the hypothesis that– the labor compensation in U.S. spillovers not just spatially but also based on the inter-regional trade networks of various industries. However, this paper contributes to the causal inferences of the labor compensation spillover by introducing the counterfactual interregional interindustry trade network using the scale free random interregional interindustry trade network.

The data of the interregional interindustry trade is retrieved from the Commodity Flow Survey Public Use Microdata File known as CFS PUMS from the United States Census Bureau website for year 2012. This data records information of approximately 4.5 million shipments which includes shipment origin, shipment destination, NAICS code of shipper and value of shipment for state level as well as metropolitan area  (ESMPD, 2012). The data of compensation, employment and industrial output of 51 states of U.S. for 71 different NAICS industries are retrieved from the IO-SNAP[1] software. With the trade network based weight matrix, the preliminary results show few newer insights of labor compensation spillover when compared to the results generated using the spatial contiguity matrix.

[1] IO-Snap uses national tables of input output table from the U.S. Bureau of Economic Analysis (BEA) to generate input output accounts and coefficients for state, national, and user customized regional economies using input-output impacts assessment tools and basic economic information on employment, compensation, and value-added data (RRI, 2017)

ABSTRACT. Empirical models of regional adjustment in the tradition of Blanchard and Katz (1992) often control for aggregate effects when measuring the incidence of regional shocks on local labor markets. In this paper we argue that the approximate factor structure is well suited to filtering out aggregate effects from regional economic data. This is because the geographic impact of aggregate shocks -- such as national recessions -- tends to vary across both space and time, and the factor model offers the requisite flexibility to control for this form of spatiotemporal heterogeneity. We demonstrate that the factor structure does a better job of controlling for national recessions in US state employment growth than the regression-based method commonly used in practice. Using the factor structure to control for aggregate shocks in a regional adjustment model, we find that region-specific labor demand shocks are primarily absorbed through changes in labor market participation; that the migration response to these shocks is limited; and that recoveries are highly protracted.

ABSTRACT. In the context of Sustainable Livelihood Analysis (SLA) framework, individuals’ vulnerability is closely related with their livelihood risk, livelihood strategy and livelihood capitals. Farmers’ livelihood vulnerability issue would be drawn more attention in development countries. Recent years, a huge number of rural land consolidation projects have been invested throughout China, to improve the irrigation and transportation infrastructure, with the goal to boost the livelihood vulnerability of farmers. In order to evaluate the actual effect of rural land consolidation, a case study is carried out in Jingmen and Yichang, Hubei Province, Central China. Firstly, we employ full sample and sub-sample OLS models to analyze the significant determinants of farmers’ vulnerability. The regression results verify the hypothesis of SLA other than several social capital variables. In the second place, Sharp Regression Discontinuity (SRD) specification is used to assess the local average treatment effect. The local Wald estimator indicates that, after rural land consolidation, there exists vulnerability discontinuity in the geographic boundary, that is, farmers’ vulnerability within rural land consolidation projects is lower than their counterpart outside at a significant level of 10%. Lastly, Difference-in-Differences (DID) specification with Propensity Score Matching (PSM) is adopted to calculate the treatment effect controlling both time-specific and group-specific effect. According to econometrical modelling, the rural land consolidation is statistically significant to decrease farmers’ vulnerability at a significant level of 5%. For the robustness check, we successively replace several key covariables of livelihood risk, livelihood strategy and livelihood capitals in the PSM-DID model. Moreover, we try to match observations between Jingmen and Yichang to conduct a counterfactual experiment. Both results sustain the baseline model.

ABSTRACT. This paper uses primary survey data to identify and characterize the various types of New Hampshire food service establishments currently sourcing local food products (“local” being grown or raised in New England), and to assess the potential for increasing institutional purchase of locally grown food products. Results from a recent direct-to-consumer survey revealed that consumers in northern New England have a negative propensity to consume produce that is purchased directly from growers or farmers markets, but consumers had a positive propensity to consume local and organically grown items overall. In other words, consumers in New England would like options for consuming local besides purchasing directly from farmers. Increasing local sourcing to intermediate channels, such as food service establishments, may help lower the opportunity cost of buying local food products for consumers. Currently, there is an information gap between New Hampshire restaurants and local food producers. This study examines which variables contribute to the likelihood that a New Hampshire food service establishment will make a local food purchase. The implementation of a state-wide survey is used to help examine various restaurant characteristics, perceptions, and practices that affect purchasing decisions for local food products. Based on previous research, it is hypothesized that smaller/midsized restaurants (serving less than 1,750 meals per week) will be more likely to source local food products. The data set is drawn from a survey of all New Hampshire establishments with help from hospitality associations, state agencies, and individual restaurateur contacts. Using a logistic regression model allows us to use a threshold parameter for the binary dependent variable. The dependent variable represents one of two values: y=1 when the respondent’s percentage of monthly local food purchases is ≥ 41%, and y=0 when the respondent’s percentage of monthly local food purchases is <41%. The threshold parameter is based on the average percentage of monthly purchases of locally grown food products among foodservice establishments found by the Food Processing Center study. This screening process prevents establishments purchasing small percentages of local food from being classified as local buyers; therefore, the model identifies characteristics of only the major purchasers in the market. Independent variables examined include business type, segment type, meals, number of store locations, level of autonomy, average number of meals served per week for each establishment, purchasing volume, and three composite variables: supplier attributes, production, and food attributes. This research has the potential to bridge existing information gaps between producer and intermediate buyers, as well as inform future policy initiatives such as the New England 50/60 Food Vision (an initiative to promote 50% of New England’s food supply being produced in the region by 2060), provide missing information on purchasing trends, and help develop strategies for expansion of the local food economy in the Northeast.

ABSTRACT. Limited resource farmers, a subcategory of small farmers, face a multitude of marketing issues. Alternative markets have been fostered in response to the traditional marketing structure from which a vast local food economics research has recently developed. Therefore, this contribution is focused on evaluating the farm income effects that local food marketing has brought to the limited resource farmers. This paper analyzes the difference between gross cash farm income produced by farmers selling locally and by farmers selling through non-local markets. The effects were comparable once normalizing the variance of errors from both samples. In addition to a linear regression, a multinomial logit model was performed to observe the factors that determine farmers marketing decisions. The statistically significant elements included urban proximity and off-farm income sources. In the other hand, growth in gross cash farm income showed significant dependence in variables with less farm family consumption and greater investment in management and storage. Those results have policy implications, in which positive coefficients obtained from the regressions should be incentivized for alleviating the financial performance of limited resources farmers.

ABSTRACT. In this paper, we employ a model using spatially lagged explanatory variables (SLX) to model population change in South Carolina – a predominantly rural state with large pockets of persistent poverty counties. In addition to the normal impacts of the explanatory variables in a standard OLS regression, the SLX model measures the impact of the independent variables in contiguous areas. It also allows for separation of spillovers from rural counties versus those from urban counties, something missing from the existing literature. We employ various economic, demographic, and social variables to model poverty rates pre-, during, and post-Great Recession. We also estimate separate models for rural counties and urban counties. There were significant impacts from the local county, surrounding rural counties, and surrounding metro counties in all three time-periods for rural counties. However, local factors had no impact on metro poverty rates in any time-period.

ABSTRACT. Heat waves are the deadliest type of natural hazard among all weather extremes in the United States. Every year, more than 1,000 heat events occur, causing an average of 120 deaths.A study in Science (Meehl and Tabaldi, 2004) predicts the future heat waves in North America will become “more intense, more frequent, and longer lasting”. Given the life-threatening consequences of extreme heat events and the predicted increase in the heat-related risks in the coming years, it is critical to identify the major determinants of heat wave vulnerability in order to minimize potential human losses. While weather events such as heat waves are exogenously determined by natural forces, the vulnerability to such hazards is shaped by various human and social components. This study uses all heat and excessive heat events that occurred over the 1996 and 2015 period using U.S. county level data to examine the determinants of heat wave fatalities. Our examination emphasizes the vulnerability of the elderly to heat waves, taking into consideration increasing senior poverty rates and elderly social-isolation. The elderly population is projected to increase to 83 million by 2050, which is nearly doubling today’s elderly population (U.S. Census Population Reports). Findings of this study are as follows. Urbanization tends to increase the adverse impacts of extreme temperature events, leading to more fatalities. Results also suggest there are learning effects from risk history that reduce the heat wave vulnerability. Our analysis also confirms that higher income reduces vulnerability to heat waves. Also, population composition is important; estimates indicate that vulnerability is greater in counties with higher proportions of elderly, young, and African-American populations. Findings also suggest that seniors living below the poverty line are the most vulnerable population sub-group. The magnitude of the effect is very large: A one percentage point increase in the percent of elderly living in poverty leads to a 50% increase in heat-related fatalities. We also find that several housing related factors are critical predictors of heat wave vulnerability; living in mobile homes or rental homes intensify disaster vulnerability. Overall, these findings increase our understanding of the socio-economic nature of heat wave impacts and inform targeting efforts designed protect and assist the most vulnerable populations.

ABSTRACT. Since 2009, the Census Bureau has also estimated poverty using Supplemental Poverty Measure (SPM), which extends Official Poverty Measure (OPM) by taking account of various government assistance programs, setting poverty thresholds based on consumer expenditures on food, clothing shelter, and utilities, and adjusting the thresholds geographically for housing expenditures. Applying SPM historically, it could be shown that near-cash government transfers assisted families with poverty alleviation, especially during economic downturns (Fox, 2015). Furthermore, latest research into long-term trends in rural and urban poverty indicates that rural areas have experienced drastic reduction in poverty under SPM, especially in recent years. Also, according to Council of Economic Advisors (CEA) report “Opportunity for All: Fighting Rural Child Poverty”, the estimated poverty rate and child poverty rate in 2013 were lower in rural areas than in urban areas under SPM but higher under OPM.

However, there are still concerns regarding whether the lowered poverty thresholds are appropriately adjusted. For instance, the higher reduction in rural poverty compared to urban poverty might just be the result of the adjustment factors rather than an actual trend. The research question is then whether the accounting of regional amenities rather than just housing expenditures could explain the difference in poverty measures between OPM and SPM and possibly bridge the gap between them.

Using the Census data, our methodology applies the Roback spatial equilibrium model for quality of life measurement and distinguishes effects of consumptive amenities and productive amenities on quality of life measures. The relative importance of their differences in explaining adjustment differentials across various urban and rural areas could shed a light on the differences between OPM and SPM in rural areas. Depending on how consumptive amenities are accounted for, the validity of the SPM as a measure of the quality of life could be affected.

ABSTRACT. It has been empirically shown that countries with abundant natural resources have exhibited slower growth compared to countries with smaller resource endowments. This paradoxical result has been termed the resource curse. This paper examines the resource curse in the US at the county level focusing on the impact of resource intensity on high school dropout rates. There is a tendency for natural resource based industries to demand less high-skilled labor, causing resource intensive economies to invest less in the education system. The resulting decline in educational attainment rates can harm long run growth (Gylfason 2000, 2001). Papyrakis and Gerlagh (2004) confirm this correlation between resource abundance and schooling and note it is able to explain a large portion of the poor economic performance.

ABSTRACT. On the back of robust GDP growth of over 5 years, India launched a rural employment guarantee scheme (MGNREGA) in 2007. While the scheme was meant to create rural assets and ease out people from under-employing agricultural sector, it utterly failed in achieving the stated goals. Instead, the corruption ridden scheme caused fiscal profligacy and high inflation in the ensuing years. Further, it crowded out other labor intensive sectors, particularly agriculture. As a result, the following years saw heavy mechanization of agriculture in India. This has led to several unintended consequences. First, the harvester machine cuts only top half of the wheat stem (a major crop in India), then extracts the wheat grain and blows the remaining crushed stem blowing in the wind, which leads to air pollution. Also, the remaining lower half of the stem (stubble) is left in the field that has to be burnt. This causes massive air pollution and smog in entire northern India and particularly in the national capital region. Interestingly, earlier, the whole of the stem was cut and converted to animal feed. Now, fodder has also become costlier and unavailable. Therefore, keeping animal-based farming has become costlier. Once animals are removed from the value chain, organic fertilizers are also out of question due to costs. Back in the field, remaining very small stem was also plowed back as fertilizer. Why don’t they do it any longer? Because it requires manual labor, which is slow, costly, requires monitoring, and mostly unavailable. Therefore, we see that fiscal profligacy driven schemes often cause more unintended damages than the problems they claim to resolve.

ABSTRACT. Since the passage of the U.S. Clean Air Act Amendments of 1990, the Environmental Protection Agency has updated the pollution standards for several critical air pollutants as new information has been unveiled related to the health impacts of these pollutants. As standards increase, this leads to a greater number of U.S. regions to be found in “nonattainment,” meaning their air quality is not up to the new standards. Since industry composition and climate are the main drivers of air pollution, for areas in nonattainment, this designation could impact the region’s labor market and its overall economy. An increase in environmental standards could significantly increase firms’ production costs leading to the shutdown of older (more polluting) industries and could also freeze the development of new plants and other economic activity in a region. If the requirements of these regulations are too stringent it could make it nearly impossible to establish a new plant or modify an older plant in a nonattainment region. However, a region could also diversify away from polluting industries and change its nonattainment status, which could protect the current economic status or even lead to growth. Thus, to understand the impact of the changes in air quality standards, we consider how it has affected the economy of counties in the eastern United States. Much of the region has hot summer weather that can contribute to air pollution and affect compliance with the standards. There is also variation in terms of industrial composition, with some areas within this region having a long history of high levels of employment in polluting industries. Since many of the same places with polluting industries also have a history of economic distress, if polluting industries are affected by this policy it could enhance this distress. However, if the counties experience industrial diversification as result of the decline of polluting industries, such as manufacturing, they may experience higher levels of long-term growth. Using county-level data, we determine the net effect of the environmental regulations on long-term employment changes and industrial diversification as proxies for the overall effect on the economy.

ABSTRACT. Google Trends (GT hereafter), a data product of Google Inc., provides voluminous and easily accessed time series data on people’s internet search activities. GT data also feature various geographic levels, including national, state, and metro levels, greatly enriching the data toolbox for researchers who are interested in spatial and temporal data. GT data have been used to predict the outbreaks and spread of disease, deaths from drugs and suicide, automobile sales, unemployment, job search activities, etc. This paper explores the use of GT data to predict gross job gains, gross job losses, and the ratio of the two at the state level in the United States. A set of Google search terms is selected to reflect the behaviors, sentiment, and economic thinking that are possibly related to the three economic variables. We employ a simple variable selection procedure to select a Google search term that is fed into regression models that use gross job gains, losses, and the ratio of the two as the dependent variables and use lags of the dependent variable, the GT term and its lag as independent variables. The regression models are estimated for each state and the nation separately and passed into diagnostic tests for model adequacy, unit-root nonstationarity, and Granger causality. Preliminary results show that the term “startup” is a good predictor for gross job gains and that the reverse Granger causality test that uses the dependent variable and its lags as regressors is insignificant in most cases. The term “layoff” is the most correlated term with gross job losses but two-way causality is present in most cases. For the ratio of gross job gains and losses, the current set of Google terms does not contain a single prominent predictor. Although most regressions adequately capture the autocorrelation of the dependent variable, they fail in the Dickey-Fuller tests and exhibit unit-root nonstationarity. The results motivate us to expand the search for more relevant Google search terms and explore machine learning techniques to select models and variables.

ABSTRACT. A couple of research topics received a great deal of focus during the late 20th century. One was the issue state PCPI convergence over time. Papers by Barro and Sala-i-Martin (1992), Levernier, Partridge, and Rickman (1995), and Bernat (2001) established the convergence of PCPI among the states through 1990. The second research topic looked at the rural urban divide within states and within countries. By the beginning of the 21st century, the popular conversation has shifted away from society trending toward more equal outcomes to one of winners and losers. The Great Recession seems to have exacerbated the trend and led to “Occupy Wall Street” and “the rich are getting richer and the poor are getting poorer”. The 2016 presidential election turned out to be a referendum on the changing outcomes of United States citizens in the early 21st century. While the popular media looked at the results as a surprise, a closer look at the underlying economic trends could have predicted the outcome. Since 2007 just 30 U.S. metropolitan areas have accounted for 62 percent of U.S. GDP growth, yet they only account for 40 percent of the country’s population. Looking at it another way, 60 percent of the U.S. population only received 38 percent of the country’s growth since 2007. While the “system” may or may not be rigged, where you live does matter. This paper looks at this diverging trend in the country’s economic outcomes and attempts to shed insight into both the economic causes and the potential longer term political outcomes. Both external factors (demographic trends, overall state economic performance, and regional influences) and internal factors (structure of the metro economy) affecting GDP growth are addressed as explanations of this new trend. John E. Connaughton, UNC Charlotte Caroline Swartz, UNC Charlotte

ABSTRACT. Regional wage disparities in the United States have been a topic of research for decades (Segal 1961; Roback 1982;Krugman 1991; Partridge et al. 2015) and policy makers often develop strategies with the intent of raising wages in their region. These strategies, however, are hampered by a monolithic understanding of how wage differences can arise. While there are countless reasons for why average wages vary by region, a fundamental deductive distinction can be made between regions that are concentrated in higher than average paying sectors (i.e. an above average paying industry mix) and regions that pay above the national average wages for their respective industry mix. Therefore, two disparate regions may have the same average wage for two very different reasons; region 1 may have a relatively low paying industry mix, but they pay above average wages for those relatively low paying sectors and region 2 may have an industry mix that is concentrated in relatively high paying sectors, however those sectors in the region pay below the national average for the respective sector. Even though the two hypothetical regions may have the same overall average wage, the underlying structure of the region’s economies are likely very different. Similarly, a region’s low average wage could arise from these two very disparate structural problems to varying degrees. Furthermore, the policy prescriptions for the economic development of a given region are likely very different for regions which possess a relatively low paying industry mix than for regions with a relatively high paying mix of industries, but for which the wages paid in the region are bellow national average wages for that region.

This analysis starts with developing a deductive wage decomposition methodology which splits the difference in the region’s wage to the national wage into a “wage effect’, “industry mix effect”, and “interaction effect”. This decomposition shares some similarity to a shift-share analysis (Stevens and Moore 1980) for wages, however, the decomposition developed here is fundamentally different from shift-share analysis and does not trivially reduce to a shift-share. In decomposing the difference in any given region’s wages to the national average into these three components which perfectly identify the wage differences, the specific reasons for why a region’s wages differ from the national average can be more elegantly diagnosed. Results of the regression model indicate that factors such as income inequality, percent union participation, and educational attainment affect the components of the wage differential decomposition very differently.

ABSTRACT. Although the literature has established an important link between socioeconomic status (SES) and health outcomes, little attention has been given to what effect a change in SES has on individual health outcomes and how these effects vary across regions. Using panel data from the 1979 cohort of the National Longitudinal Survey of Youth (NLSY), including the confidential restricted geocoded data, we investigate what impact a change in SES has on the likelihood that an individual will be obese during his or her lifetime. The NLSY geocoded dataset also allows us to examine regional differences in the impact of changes in SES on obesity prevalence by incorporating regionally specific socioeconomic controls. Preliminary results suggest that, while controlling for individual level demographic and socioeconomic characteristics, individuals moving from a lower to a higher SES are less likely to be obese, while individuals who move from a higher to a lower SES are either more likely to be obese, or no statistically significant effect is observed. This suggests that, to some extent, the various health disadvantages encountered by individuals in lower SES groups can be ameliorated by upward SES movement. We also find some important differences by region.

ABSTRACT. The purpose of this paper is to analyze the effect of corroboration of medical care service and nursing care service on equilibrium. We consider two kinds of region with different population density and analyze the equilibrium medical care service level and nursing care service level with a traditional Hotelling model. Moreover, we compare each optimal medical and nursing care service fee with integrated medical and nursing care fee. We show that some patients in rural region may choose the hospital located in the other region beyond boundary.

ABSTRACT. Ambient Air Pollution adversely impacts air quality and human health. According to World Health Organization (WHO) 235 million people around the world currently, suffer from asthma. In the United States, about 25 million Americans have asthma. In Pennsylvania asthma remains a serious health concern. There is substantial epidemiological evidence indicating a link between outdoor air pollution and asthma symptom more specifically between particulate matters and asthma. Using county-level data for 2003-2011 in spatial panel framework (imposing a prevailing wind pattern weight matrix) we investigate the direct and indirect impacts of Ambient Air Pollution on asthma hospitalization in Pennsylvania. Aside from PM2.5, we control for a set of explanatory variables such as population density, precipitation, smoking rate, and population breakdown. Results show PM2.5 concentration has a positive effect on asthma hospitalization. Also, asthma hospitalization is effected by PM2.5 concentrations in neighboring counties (One-unit increase in PM2.5 will add on average $175,000 yearly direct cost and $65,000 yearly indirect cost of asthma hospitalization). The study highlights the need for a more realistic and accurate impact of Ambient Air Pollution on asthma that considers the impacts on the neighboring region as well. In case of the health-related effects of Ambient Air Pollution, we suggest considering a precise algorithm to identify the neighbors.

ABSTRACT. Prior studies that reported the impact of innovation on organizational performance reveal significant findings but did not focus on oil and gas industry. While mostly viewed as a commodity that is subject to market-based pricings that dictate the industry boom or bust, oil and gas industry leaders in fact have continuously attempt to consider innovation as the solution to combat market fluctuations such as the oil price crash crisis in 2014. In other words, instead of the normal response of downsize capital investment and reduce workforce, transformational oil and gas leaders seek innovation including innovative processes in the entire supply chain that drive new products, markets, or productivity enhancement. The objective of this study is to examine the relationship between innovation and performance of public exploration and production (E&P) upstream firms. Some of these firms operate only in certain geographic regions (petroleum basins) due to reasons of size, location, or technical expertise. Other firms operate in multiple regions. We collect data from the US patent office and analyze the relationship between patents and financial performance of E&P firms to evaluate if public oil and gas firms that operate regionally are more innovative compared to firms that operate in multiple regions. The results of this study provide a new lens to understand the impact of innovation within the oil and gas industry while adding to the field of knowledge in response to the call of industry leaders in the petroleum industry innovation.

ABSTRACT. In recent decades, both hydraulic fracturing and wind energy have been increasing rapidly across the United States. This surge in both types of energy development may be an important source of employment in rural areas, particularly those located in shale play regions and / or windy environments. In some states, wind farm installations overlap with the shale drilling period; other states have both types of activity but occur in mutually exclusive timeframes; still others have only one type (and not both). The majority of the previous literature on this topic has examined the economic contributes of unconventional drilling in various regions or states, and a small number of studies have looked at impacts associated with wind development. However, these studies typically focus on a single type of energy development, and generally do not consider the possible effects of alternative energy expansion.

The purpose of this study is to estimate the regional employment impacts associated with recent energy development (both shale drilling and wind farm activity) in Arkansas, Kansas, and Oklahoma. The amount of energy development activity varies highly across these states, with the number of unconventional wells during the period 2001-2016 ranging from 918 in Kansas, to 4,600 in Oklahoma, to over 6,400 in Arkansas. Similarly, wind farm activity across these states resulted in 0 megawatts (MW) of energy capacity in Arkansas, 4,400 MW in Kansas, and over 6,600 MW in Oklahoma. Importantly, Kansas and Oklahoma rank highly in terms of wind energy employees, while Arkansas and Oklahoma rank highly in terms of oil and shale employment. The time span of our dataset (2001-2016) encompasses both pre and post- energy development growth periods. We take advantage of the fact that each county has a different boom period for the alternative types of energy development (including many that never experienced any type of boom). We estimate the treatment effect of the two types of energy activity by applying a panel regression model, with total county-level employment as the dependent variable. We also estimate the model for employment at the 2-digit industry sector levels. The results indicate that energy development activity does increase regional employment, but only for a small subset of industry sectors and regions. In particular, the employment impacts for unconventional drilling activity are larger than those for wind farms in region with both types. One notable contribution of this paper is the use of newly available county-level employment data from the Upjohn Labor Institute. Most previous studies on this topic use official governmental county business patterns (CBP) employment numbers, which suffer from missing data in many rural counties. This updated data applies Isserman and Westervelt’s (2006) technique to overcome this employment suppression. This, in turn, allows for the estimation of more precise impacts, particularly in rural counties where drilling or wind development can play a large role in the local economy.

ABSTRACT. Recent technological advances in oil and gas drilling have enabled the productive extraction of natural gas using unconventional drilling techniques. Shale development tends to occur in rural areas and benefits from employment and income opportunities can help stimulate economies in lagging regions. However, drilling activity can alter natural landscapes and give rise to increased industrialization and the associated unwanted negative externalities. One way to assess the impact of shale gas extraction on its surrounding area is to examine how drilling activity is capitalized into nearby house prices. The presence of nearby drilling may increase house prices from residents recognizing the “boom” benefits and royalty and lease payments, or decrease house prices from the unappealing aesthetics of wells and the associated negative externalities, such as groundwater contamination. Using an array of econometric techniques, we examine how residents value the proximity to producing horizontal wells in West Virginia. Much of the prior research in the Marcellus shale region has focused on Pennsylvania, especially the more urbanized areas around Pittsburgh, and little attention has been paid to the resource rich state of West Virginia. Additionally, West Virginia is predominantly rural and using the hedonic method in such thin housing markets requires additional techniques. Understanding how drilling activity is capitalized into house prices in a previously unexplored region that accounts for a large proportion of the Marcellus shale gas supply can provide insight for policy discussions related to shale gas development.

ABSTRACT. State and local policymakers must balance taxing economic activity to generate government revenue with the potential loss of activity with higher taxation. A prime example of this occurs when oil- and gas-producing states set tax rates on the production of nonrenewable natural resources. Proponents of increases in the tax rates see opportunities to increase government revenue, especially during periods of high oil prices and activity. Conversely, opponents argue that increases in taxation will lead drilling companies to go to neighboring states to drill for oil and gas. Some states have even chosen to subsidize drilling by forgoing severance tax revenue on certain wells in order to attract more drilling activity. For example, Oklahoma lawmakers reduced the state’s severance tax on new horizontally-drilled wells to one percent for the first 48 months of production, which equated to $379 million in tax breaks in 2015 (Oklahoma Policy Institute, 2015). The state now faces a significant budget deficit following the precipitous drop in the price of oil and drilling activity from peak levels in mid-2014.

Despite the important fiscal and economic implications of how drilling activity responds to changes in severance tax rates, there is little empirical evidence on the sensitivity of drilling activity to state severance tax rates. Moreover, little is known about the degree of spatial spillovers from tax rates in neighboring states. The purpose of this paper is to estimate how responsive firms are to differences in severance tax rates across states and over time. We investigate if firms respond differentially to oil prices and severance taxes, and whether or not firms respond to neighboring states’ severance tax rates.

Using data from Drillinginfo, we look at all oil wells drilled between 1980 and 2015 in reservoirs that cross state lines. Our identification strategy is based on using intra-reservoir variation in severance tax rates interacted with oil price. By focusing on reservoirs that cross state lines, we use fixed effects to control for both reservoir (i.e., unobserved geologic quality) and unobserved state characteristics such as permitting and regulatory ease. We find a stable and inelastic short-run price elasticity of drilling equal to 0.8. We also find an inelastic response to severance taxes, with a 1 percent increase in the dollar-per-barrel tax rate associated with a 0.3 percent reduction in drilling. The positive response to neighboring states’ tax rates is small and does not differ statistically from zero. It is also statistically smaller than the own-state elasticity.

Our estimates indicate that an increase in revenue from higher severance tax rates likely offsets the resulting decrease in drilling activity. The policy implication is that small increases in severance tax rates could allow states to increase revenue. Furthermore, engaging in tax competition may lead to revenue losses in competing states.

ABSTRACT. The water required to meet demand for goods produced and consumed in the southeastern US is estimated using a multiregional Environmental Input-Output/Life Cycle Assessment model (EIO-LCA). The analysis suggests that direct and indirect water requirements embedded in the production of goods and services is heterogeneous across the study region. The resulting water footprint could inform the design of water management policies for local and regional institutions and open up the possibility for modeling virtual water trading markets.

ABSTRACT. SPECIAL SESSION ON WATER - CINDY ROGERS

ABSTRACT. This research investigates approaches for estimating water demand and responses to water rate changes. Water utility pricing generally involves different rates for different levels of consumption. Such block rate structures complicate demand estimation. We challenge the long standing implicit assumption that a city’s demand can be represented by a single city-wide (or region-wide) demand curve in the presence of rate block pricing schemes. We employ the longest and most detailed panel dataset of household level water consumption, weather records, and housing characteristics available in empirical studies. This allows us to estimate demand curves for separate user groups in an increasing block rate scheme. We analyze two water utility rate changes implemented in Norman, Oklahoma: the first in 2006 and the second in 2015. We find downward sloping demand curves in all but one case, and estimate price elasticity to be between -0.13 and -0.53. Our estimates suggest that, in general, the more water a customer consumes, the more price sensitive she is.

ABSTRACT. In many rural places, the food industry is an important one, tying production agriculture to the manufacturing sector. One outcome that might be expected as a result of increasing consumer interest in healthful, niche and regional foods is more place-based value-added processing nearby the farms and ranches where food is produced. We posit the rise of a bimodal food manufacturing structure with both a stable set of large players but also a small and dynamic set of businesses more regionally dispersed across the U.S. (e.g., the rise of microbreweries). We are especially interested in these new and small food and beverage manufacturers, as they may be engines of entrepreneurship, innovation, and job creation in rural areas of the U.S. Preliminary findings suggest a re-localization of food processing facilities. New businesses, in certain food-related sectors, are smaller and more dynamic, suggesting they may be more entrepreneurial and innovative than large, stable establishments.

ABSTRACT. Internal migration, job changing, and business dynamism have declined in the United States since 1980. Using individual-level, geocode data from the National Longitudinal Survey of Youth’s 1997 cohort, I more closely investigate these trends by examining the relationship between labor market-specific measures of business dynamism and an individual’s decision to change jobs or to migrate. I find that business dynamism reduces out-migration of low-skill workers and either spurs or hinders job-changing among high-skill workers depending on the measure of dynamism used. The fact that business dynamism influences migration decisions more for low-skill workers and job-changing decisions more for high-skill workers underscores the need to allow for heterogeneous effects when discussing the impact of business dynamism on job-changing and migration decisions.

ABSTRACT. This paper discusses how an analysis of skill demand within a supply chain context can advance economic and workforce development practice. Taking the case of the U.S. automotive industry, the author investigates how manufacturing skills are distributed among related industries using publicly available data. Related industries are organized by the volume and direction of their economic transactions, demonstrating variation in skill demand across supply chain tiers. This skill-centric perspective allows policymakers to see opportunities for development. State data analysts can use this skill-centric approach to evaluate the skill dependence of automotive parts suppliers and to identify undervalued or underutilized skills in related industries. They can see where business and training services might strengthen lower-tier firms’ position in the supply chain. The results of this analysis demonstrate that both automakers and their first tier suppliers should be engaged as conveners and the ultimate beneficiaries of investments in supply chain skills. The paper highlights how additional research would help to evaluate skill transferability across and within supply chain tiers.

ABSTRACT. The continuing decline in the U.S. coal industry has implications that differ from industry to industry and from region to region. These differences in industry implications arise because of the nature of the coal industry ecosystem (CIE), in which each industry occupies its unique position in the coal industry supply chain. There are industries that provide goods and services directly to the coal industry, and those that supply the coal industry indirectly via provision of goods and services to the first-level suppliers and to their suppliers throughout subsequent supply chain levels and linkages. Likewise, each region has a unique mix of industries, so as collections of industries, regional economies also vary in terms of their position in the CIE, and in terms of the extent to which they are aligned with the coal industry. These critical differences, in turn, give rise to impacts of coal decline that can vary substantially from region to region, from state to state, and from county to county. In this report, which is one of five reports recently prepared under contract with the Appalachian Regional Commission, we describe an effort to provide for the Appalachian Region an approach to identifying counties that have experienced impacts from the decline in the Appalachian Region coal industry. We develop and compute three measures that reveal meaningful characteristics of the CIE at the county level, in terms of: industry and place-based CIE dependence, changes in CIE-dependent employment, and susceptibility to anticipated continued coal industry decline. We then combine these three measures―Dependence, Impact, and Risk―to form a typology that we apply to identify and focus on counties in three identifiable categories: Hardship counties, Vulnerable counties, and Depressed counties. The 123 Hardship counties rank below the median of all counties on the Dependence and Risk dimensions, and above the median on the Impact dimension. These counties have experienced high levels of negative impacts, but because they have lost most of their CIE-dependent industry jobs, they are not at risk for further substantial coal decline-driven job loss. The 124 Vulnerable counties rank below the median score on the Impact dimension, but above median scores on the Dependence and Risk dimensions. These counties did not exhibit negative consequences of CIE decline between 2005 and 2015, but because they are CIE-dependent and at risk of further coal industry decline, they form a group of counties with higher potential for future coal decline-driven job loss. The 83 Depressed counties score above median on all three dimensions. They suffered CIE-dependent job losses between 2005 and 2015, they continue to host CIE-dependent industries, and they are at risk of further future coal decline job-loss. Finally, we narrow focus on the Depressed counties that rank highest on all three CIE-dimensions. Specifically, there are nine Depressed counties that rank in the top 20 on all three CIE dimensions. These are Mingo and Boone Counties in West Virginia; Harlan, Leslie, Martin, Pike, and Perry Counties in Kentucky; and Buchanan and Dickenson in Virginia. Unemployment rates in these counties range from a low of 8.9 percent to a high of 13.2 percent; job losses have been substantial, as high as 34 percent and averaging 21 percent. On average in these counties, roughly one in ten employees works in the coal industry itself, and their wages account for almost 23 percent of county totals. In two of these counties, coal-industry employment shares exceed 20 percent, and corresponding wage shares of total exceed 37 percent. These Depressed counties clearly warrant special attention and concern if current coal industry trends continue.

ABSTRACT. Popular belief often praises technology for its ability to create jobs and promote economic development. The rationale used to justify this positive thought is that knowledge, which is claimed to be a nonrival and a public good, is partially or completely transferred from a frontier region to a follower. However, urban technological progress can also increase rural unemployment through product market competition. In other words, the competition among urban and rural firms can reduce output demand of rural firms, and therefore, it can negatively affect labor demand of firms located in rural regions. Technological development in the urban region could translate into the glut of urban manufacturing goods into an economy. Urban technological progress also raises demand of urban goods and services due to an improvement in their qualities. Hence, urban technology could harm firms operating in the rural area, and so product market competition could increase the rural unemployment rate.

Additionally, brain drain (which is often considered as a backwash effect of urban technology on rural economic development) can have an ambiguous influence on the rural labor market. That is, brain drain can attract human capital away from rural regions, but it can also create jobs in urban areas for unemployed workers living in rural regions. Therefore, the net effect of urban technology on rural labor market performance is theoretically ambiguous. Alas, these facts have not yet received serious attention in the academic arena. The relationship between urban technological progress and the rural labor market is particularly important for three reasons. First, without taking the negative effects of urban technology into account, any place-based policy aiming at increasing local and regional jobs by technological advancement might have unintended consequences. Second, one cannot ignore the effects of technology on regional labor markets because the interregional flow of labor (specifically, within the U.S.) is self-evident. To put it differently, brain drain could strongly manifest when there are a few constraints on the flow of labor between urban and rural areas. Third, given that the impacts of technology on future labor market is a pressing concern among policymakers, business leaders and scholars, this paper could also provide a piece of evidence for the importance of technology in determining the performance of the labor market.

I first present a basic conceptual framework that consolidates the endogenous growth theory of Romer (1990) and the spread-backwash notion of Hoselitz (1955), Myrdal (1957), and Hirschman (1960). The framework yields insight to understand the interaction between urban technology and the rural labor market. I also provide some suggestive evidence regarding the existence of the aforementioned effects of urban technology.

Next, I set up an empirical analysis to study the net effect of urban technological stock on the rural unemployment rate. Using patent counts at the county level in the United States, there is evidence that a 1 percent increase in urban technological stock, which is constructed from the perpetual inventory method and the basic inverse distance matrix, correlates with a rise of 0.1-0.2 percent in the rural unemployment rate in a short or medium run. Several competing hypotheses and econometric issues have been examined. Results of the examination indicate that the main finding is highly robust. Finally, I perform a simple assessment on rural welfare and find that urban technology has a statistically significant negative impact on the average wages but not the average incomes of the rural population. This crude assessment suggests that employed workers in rural regions might suffer from welfare loss caused by the progress of urban technology, although government transfer might alleviate some this loss. While this simple assessment cannot vividly distinguish between different kinds of government transfers to a specific rural subpopulation (for instance, unemployed versus employed), it could serve as a motivation to study further the impact of urban technological progress on a particular group of rural dwellers.

ABSTRACT. We focus on a region that is creative in the sense of Richard Florida. The creative class is broadly composed of existing and candidate entrepreneurs. The general question we analyze concerns the effects of Schumpeterian competition between existing and candidate entrepreneurs on economic growth and innovation policy in this region. We perform four specific tasks. First, when the flow rate of innovation function for the existing entrepreneurs is strictly concave, we delineate the circumstances in which competition between existing and candidate entrepreneurs leads to a unique balanced growth path (BGP) equilibrium. Second, we examine whether it is possible for the BGP equilibrium to involve different levels of R&D expenditures by the existing entrepreneurs. Third, we show how the BGP equilibrium is altered when the flow rate of innovation function for the existing entrepreneurs is constant. Finally, we study the impact that taxes and subsidies on R&D by existing and candidate entrepreneurs have on R&D expenditures and regional economic growth.

ABSTRACT. We review the mechanisms, goals and awards of the federal Small Business Innovation Research (SBIR) program, which is a set-aside program for all major federal research agencies that attempts to foster small business, spur economic growth, and assist socially disadvantaged persons. Using award data provided by SBIR, we explore the geographic distribution of awards, which, while clustered in high tech states, are found in most states. We aggregate the award data to the zip code level and match it with federal data on socio-economic characteristics of each zip code. Using a count modeling approach beginning with Poisson regressions to explore factors associated with awards in a zip code, we find that zip codes high in socially disadvantaged groups receive on balance, disproportionately fewer awards. We then explore this finding for each agency to highlight which enjoy the most success at delivering awards to socially disadvantaged zip codes.

ABSTRACT. State and local policymakers are very interested in how attending college in one’s home state affects the likelihood of living in that state after college. This paper uses cohort-level data from the American Community Survey, decennial censuses, and other sources to examine how birth-state college enrollment affects birth-state residence several years later. Ordinary least squares and instrumental variables estimates both suggest a statistically significant positive relationship. The preferred instrumental variable estimates suggest that a one percentage point increase in birth-state enrollment rates increases later life birth-state residence by roughly 0.41 percentage points. Implications for policy are discussed.

ABSTRACT. In the housing market literature, the most common approach to measure the user cost is a weighted average of quality-adjusted prices. This procedure is often used to compute national average estimates, even though several spatial differentials and variations in housing rental and costs are not being captured. At the same time, there are other differentials that can arise from differing equilibrium processes in the rental and housing markets, which are both less understood and more problematic. In fact, while rental markets reflect current supply and demand processes, housing costs reflect a long-term expectation of the future price of the house: forward-looking markets capitalize long-term expectations for the growth of a city in the housing stock. Therefore, housing prices will be higher in the most desirable (high amenity) places. Using detailed household and county level data covering from 2005 to 2016, this study first computes a quality-adjusted measure of rents and housing values across the U.S., and secondly, attempts to empirically estimate the characteristics associated with differential quality-adjusted rent/housing-value ratios for several years -focusing attention on natural amenities- and their evolution over time. Consistent with our hypothesis, our results show first that places with high levels of natural amenities have both higher rents and housing values. Second, there is a monotonic increase in rents and values as amenities increase. Third, the ratio between rents and values decreases as amenities increase, suggesting that amenities are capitalized more rapidly in values than in rents.

ABSTRACT. The purchase of a home is the largest investment made by most American families, and home equity is the largest component of family wealth. Homeownership, an essential part of the “American Dream,” has not only been a financial goal for many American families but is also central to U.S. housing policy. Scholars have long documented the social and economic merits of homeownership and explored the factors that influence access to it. However, despite the abundance of literature on homeownership and housing tenure choice, there lacks a study that focuses on whether and how debt and wealth influences a household’s decision to own or rent a home. This paper explores the effect of household debt and wealth on subsequent tenure choice. Using 2004 and 2008 panel data from the Survey of Income and Program Participation (SIPP), this study attempted to identify the causal effect of household debt and wealth on household’s decision to change tenure choice by examining what factors influenced transition from homeowner to renter or from renter to homeowner. Data analysis shows that household secured debt, household wealth, and household income play a significant role in household’s change in tenure choice – owners becoming renters and renters becoming homeowners.

ABSTRACT. Gentrification is of growing interest to policy makers as there are concerns about the effects of higher income families pushing low income individuals out of their homes. While many things are considered to be possible causes of gentrification, there are empirical problems when attempting to highlight a specific causal mechanism. In this paper, we examine one of the possible causes of gentrification – the influx of gay and lesbian couples into a community. Anecdotal evidence suggests that there is a relationship between gays and gentrification, but it could be that gays endogenously sort into neighborhoods that are more likely to gentrify. Using an instrumental variables strategy where we use support for a law similar to the Defense of Marriage Act (DOMA) as an instrument for the number of gay couples, we highlight a causal relationship between gay couples and the change in income of a tract. Our first stage results suggest that this instrument is appropriate. Preliminary results suggest that neighborhoods with more gay and lesbian couples are more likely to experience gentrification.

ABSTRACT. Many rural communities throughout Wisconsin are expressing concern over local housing markets. As rural communities are embracing the idea of placemaking, or working toward improving the quality of life within the community, as an economic development strategy. Central to this approach is the local housing market. Rural communities are voicing concern over both housing affordability and quality of the housing stock. This exploratory research will examine rural housing markets with an eye toward the drivers of, and impact of, housing stress. Here housing stress is related to both housing costs relative to income and quality of the housing stock.

ABSTRACT. This study aims to examine how transparency can undermine the effects of natural resource abundance in a country by using membership in the Extractive Industries Transparency Initiative (EITI) as a way to measure a country's willingness to have a transparent and accountable government. Based on previous research, this study uses panel data from 1996-2016 to analyze EITI membership's effect on various governance and development indicators. Consistent with other studies on the effectiveness of the EITI, this study shows that there are some significant effects on governance and development, but long-term effectiveness of the EITI remains unclear.

ABSTRACT. Research involves the World Bank data on nations’ taxation revenue (% of GDP) and the Fund for Peace’s data on human rights abuses, economic inequality, group grievances, and state legitimacy. I also included data with and without Moody’s ratings to test for the investment worth of these countries. My sample is the 110 most fragile nations and the bottom 20 most fragile nations according to the Fund for Peace. I used data from the year 2014 because it is the most recent year with sufficient data on taxation (% of GDP). My observations were 71 nations. I ran a linear regression model to test for how much does taxation (% of GDP) influence human rights abuses, and used state legitimacy, group grievances, and economic inequality to control for potential abuses of tax revenue. My research found that tax revenue as a percent of GDP has a statistically significant, negative effect on the abuses of human rights. I also ran a correlation matrix with my independent variables to make sure there was no multicolinearity, and none was found.

ABSTRACT. The purchase and use of local agricultural products has been a growing trend across the United States, particularly in New England, over the past few years. While there is extensive research on direct marketing through farmers markets and community supported agriculture programs, there is a notable gap in research on intermediate markets, such as restaurants and supermarkets. In this project a large-scale web survey on local agricultural product purchasing habits was administered to restaurant food buyers across New Hampshire. The demographic indicators for local food purchasing were analyzed. According to statistical data analysis, profiles of the average restaurants that do or do not purchase locally will be presented. The statistical trends in the data will be used to suggest possible approaches to encouragement of local food use in New Hampshire restaurants through policy and programs.

ABSTRACT. Industries such as medical, food, construction, and cosmetics etc. have created a high demand for the building of complex molecules. This leads to an increase of research focus towards finding a more productive, safe, efficient, and reliable method of building these molecules. One approach to the efficient synthesis of new bonds is through the process of biocatalysis. While some research focuses on optimizing known biocatalytic transformations, the discovery of new biocatalysts is also important. Our lab approaches this goal by introducing small organic substrates to living bacterial cultures. We utilize carbon-13 NMR as the primary screening tool by enriching our substrates with a carbon-13 isotope label. By using copper cyanide (13C labeled) generated from commercially available KCN we have shown a simple microwave reaction can be used to build a variety of carbon-13 labeled substrates. With this library of carbon-13 labeled substrates, we will explore reactivity in bacterial cultures.

ABSTRACT. Philadelphia is using a 1.5 cent per ounce sweetened beverage tax (SBT) to fund various new social programs in the city. One target program is pre-kindergarten child daycare for poor families. Preliminary research suggests that two major benefits of such a program are decreases in the consumption of sweetened beverages and improvements early childhood education outcomes. The main intention of the program, however, is to release qualified parents into the labor market to help them lift their families out of impoverished conditions. I measure the human capital embodied by all women in households who will qualify for the program--those under 2005 of the poverty level. I do so using microdata from the 2011-2015 American Community Survey for all women in Philadelphia. Since the idea if to include the results of this analysis in a computable general equilibrium (CGE) model, I use an approach that enables the estimation of wage income for each potential worker by industry that is identified in that model. We will use criteria that are employed to select families to qualify for the program to inform the types of individuals, who are not presently part of the labor force, who are likely to enter the labor market.

ABSTRACT. This study examines various economic and demographic factors and their impacts on two different levels of educational attainment. The two different levels are secondary and postsecondary education and will be broken down by state as well as the difference from national average. The data is gathered from multiple government websites and is applied using OLS regressions to determine the impacts of the control variable in each model. Previous research done in this field showed that many variables included in this research have been found to effect both levels of educational attainment. The results show evidence that multiple variables controlled for are highly significant and have an effect on secondary and postsecondary school graduation rates.

ABSTRACT. This is a case study of the potential for implementing the ACCC conductor, a high-capacity low-sag transmission wire. With this technology, the United States could better meet its energy demands, increase efficiency, and build resiliency of its grid. As energy demand grows and the nature of energy supply rapidly shifts, we need now more than ever to have a grid we can rely on. The United States, however, been a little slow on the uptake. As much as the Americans enjoy seeing ourselves as a global leader, we would be wise to turn our gaze toward developing countries. Indonesia has been robust in its efforts to take advantage of the ACCC conductor. I will explore a case study of both countries to better understand the technological, economic, and business conditions that make, or break, a brighter energy future.