6
Uses of Current Rural Classification Systems
This chapter summarizes the sixth session of the workshop, in which a panel discussed uses of current rural classification systems in research and program design and administration. The panelists were Douglas O’Brien (White House Domestic Policy Council), Timothy Parker (Economic Research Service of the U.S. Department of Agriculture [ERS/USDA]), Thomas Johnson (University of Missouri), Kenneth Johnson (University of New Hampshire), and Rose Olfert (University of Saskatchewan). The panel discussion was followed by open floor discussion. Brigitte Waldorf (Purdue University) moderated the session.
STATEMENT BY DOUGLAS O’BRIEN
O’Brien said that his focus is policy rather than research, but he follows both rural policy and research closely.
The Purpose of Defining Rural
O’Brien summarized the USDA Rural Development’s (RD) report to Congress (2013), required by the 2008 Farm Bill, which described how RD defines rural and why. He explained that national programs and private investors often bypass rural investment because of low capacity and a desire for high return on investment. In rural places, resources are scarce and policy makers have decided to set aside special funds for them, so a way to delineate eligibility for funding is needed. The U.S. Congress
found over time that total population is the simplest criterion. The 2008 Farm Bill required that RD assess the various definitions used within the agency, describe the effect of the variability of definitions on program effectiveness, and make recommendations for better targeted spending.
Assessing Various Definitions of Rural
O’Brien said RD starts all application reviews with eligibility determination. RD has about 40 different programs, about 10 of which are for housing. For most business development programs, the default definition of rural area used to determine eligibility is less than 50,000 in total population and nonadjacent/contiguous to such a municipality, but lower population caps are used for water and community facility programs. Generally, housing programs use a threshold of 20,000; however, the rules can be complicated. RD’s infrastructure programs use a threshold of 10,000. He said RD uses about 15 different definitions.
As exceptions to the default, the undersecretary has exemption authority, such as “rural in character.” If the undersecretary determines that an area is rural in nature, then it is eligible, even if not eligible according to the default. For example, in an urbanized area, a project would be eligible for a guaranteed loan if the undersecretary determined that the specific address was rural in character. However, O’Brien said exemptions are rare.
O’Brien described a few, relatively new programs for which the “rural area” eligibility criteria are not used. The Rural Energy for America Program (REAP) helps to finance small renewable energy and energy efficiency projects. The Rural Electrification Act is another, as is the Food Deserts Program, now called Health Food Financing, which finances projects that help provide fresh food to areas that lack such access, whether rural or urban. Some of these programs provide a subsidy, often a loan subsidy to help grow the economy or provide economic opportunity for rural people. However, he said, in some situations, the best way to use federal dollars would be to invest them in an urban area because the urban area would create expanded markets for people in rural places. He said he would like to help policy makers make this kind of determination.
An RD application must qualify on three fronts to be eligible for funding: the applying individual or entity must be eligible for the particular program, the proposed activity must be an eligible activity for that particular program, and the location of the proposed activity must be eligible for that particular program.
In O’Brien’s view, variability in population thresholds is bad for project effectiveness. It creates arbitrary barriers to regional strategies and perpetuates community isolation and less cost-effective economic
and community development practices. Many times the different definitions make it difficult for communities to figure out which programs apply. The more nuanced and different definitions make it hard for potential recipients to understand the reason why the policy exists.
Recommendations for Ways to Better Target RD Funds
O’Brien said that the report recommended that RD make use of a common population threshold—any place outside of a city of 50,000—which would vastly simplify eligibility requirements. But on top of that, priority points, such as those in RD’s water program, could be used. In that program, a region gets points if it is far below the threshold in population, and there is a median income or poverty indicator that can also provide priority points.
He said with this approach, communities could be prioritized because they are in more remote areas or in areas of greater poverty. He also pointed out reasons why the government provides subsidies to people who live in rural areas. Rural places have relatively low capacity and they may have the greatest need.
When considering how to improve a rural definition, an analyst should consider the purpose of that classification. Understanding regions with the greatest need is one thing, but the other question is where federal dollars can make the biggest impact. In 2009 and 2010 when there was pressure to create jobs using the American Recovery and Reinvestment Act, he pointed to debate about whether to put the dollars in places that were going to create jobs, or to put them where there was huge need.
O’Brien noted that work on rural area classification has a lot more application than a change in legislation. However, he noted, a change in legislation, regulation, or program implementation is an important consideration. He also noted the status quo of a definition of eligibility for RD programs has incredible momentum. Those who are currently eligible may have a lobby to continue the current definition, while those not currently eligible do not generally organize to provide input. Even if there were good reasons to change an eligibility criterion, there might not be much public support.
O’Brien observed that if the goal of research around rural area classification is to change behavior, then it is important for the definition to be simple enough so stakeholders will understand it, act on it, and organize around it.
STATEMENT BY TIMOTHY PARKER
Parker explained that one of his responsibilities at ERS is to receive inquiries and complaints about the Urban Influence Codes (UIC) and the Rural-Urban Continuum Codes (RUCC). He focused his presentation on the use of rural area classifications in program design and administration by many federal programs.
Parker said that rural classifications serve two main purposes: to identify underserved rural areas where distance from urban centers and low-population density leads to shortages of critical services such as health care and banking; and to identify and target federal assistance to distressed rural areas where distance from urban centers and low population density lead to a lack of economic opportunity. Many government agencies look at both, he said.
Underserved Rural Areas
Several health services programs primarily concerned with underserved areas use ERS rural area classification codes in combination with other indicators. They are as follows:
- Federal Office of Federal Health Policy in the Department of Health and Human Services (DHHS) uses the Rural-Urban Commuting Area (RUCA) Codes to administer grant programs to build rural health care capacity, and coordinate funds.
- Centers for Medicare & Medicaid Service also use the RUCA Codes for payment purposes, such as cost reimbursement for Critical Access Hospitals. This is important, he said, because a considerable amount of money goes to rural hospitals.
- Office of Rural Health in the Department of Veterans Affairs has recently begun using the RUCA Codes to determine “highly rural areas” in order to target telemedicine and tele-video technologies to address distance, and for telephone care management.
Distressed Rural Areas
In addition, other agencies use these classifications in distressed rural areas. For example, the Corporation for National and Community Service uses the county-based RUCC, the UIC, and subcounty RUCA Codes to identify rural areas and award funds. AmeriCorps uses the RUCA Codes and the RUCC to identify needy areas, and Senior Corps connects people aged 55+ with volunteer organizations that need their job skills and expertise. The Social Innovation Fund awards grants to innovative community-
based nonprofit organization focused on youth development, economic opportunity, and healthy futures, and the Volunteer Generation Fund supports voluntary organizations in addressing critical community needs.
The Consumer Finance Protection Bureau uses the UIC under its new regulations to meet escrow requirements for financial institutions. Several rules have provisions that relate to mortgage loans made by creditors operating predominantly in rural and underserved counties or made in rural counties. For example, requirements under the Truth in Lending Act rule require certain creditors to create escrow accounts for higher priced mortgage loans, but rural and underserved counties are exempt from this requirement. In addition, ability to repay and Qualified Mortgage Standards under the Truth in Lending Act rule allows exceptions for mortgage loans with balloon payments that do not meet the qualified mortgage standard.
Parker also summarized why the ERS rural area classifications may not be used in other programs. First, some rural definitions used in federal programs preceded the ERS classifications, and it is hard to change codes, which may require congressional legislation. Second, county-level classifications can be too big for targeting rural assistance, particularly in the West. Third, the ERS county and subcounty classifications are complex and difficult for nonresearchers to understand.
STATEMENT BY TOM JOHNSON
Johnson focused his presentation on policy research and assessment, and on its relationship with policy development and delivery. He said that definitions have consequences. They make money and cost money for people, which provides incentives for people to change, influence, and have an opinion about them. Many programs use and depend on these definitions. Rural policy research and assessment must explicitly consider the effect of rural definitions on program eligibility and impacts. For policy assessment and analysis, real data must be synced with those definitions. As in any type of spatial research, rural policy research is constrained by data availability, he commented. The plethora of definitions and classifications frequently compounds the data availability problems.
Johnson noted many federal agencies define rural, including the Census Bureau, Office of Management and Budget (OMB), and ERS/USDA. Many programs use and depend on these definitions. From a policy analyst’s point of view, it is important to understand the goals of each program. For instance, to a transportation official, rural means low-volume roads with long distances between intersections, with high rates of single-car accidents and fatalities. It is not hard to understand why every
program has to distinguish rural from urban, he noted, and each of them is going to need a slightly different definition. To illustrate:
- The Federal Highway Administration uses a modified and flexible version of the Census definition of rural.
- The U.S. Department of Education’s National Center for Education Statistics uses its own Local Codes system, similar to the RUCC, to classify school districts.
- The National Center for Health Statistics in the Centers for Disease Control and Prevention developed and uses a variation of the OMB metro/nonmetro classification.
- The USDA/RD uses unique criteria to determine eligibility for their programs, as discussed above.
- The Department of Housing and Urban Development (HUD) has three ways of defining rural areas based on population, including small places within metro counties.
Johnson pointed out RD blurs the line between urban and rural definitions. It turns out that many of the rural development programs qualify applicants in metro counties. In the Business and Industry Loan Guarantees and Rural Business Development Grants Programs, for example, he found most investments were close to the metro center but just across the line.
Using HUD and USDA programs as examples, Johnson said every program is different for good reason but asked if it is possible or desirable to reduce, simplify, or harmonize definitions. Each policy program is designed to address an issue or problem which is often, at least partially, related to population density, distance to urban services, land use, and access to infrastructure. The goal is frequently to change economic outcomes in these areas. Each of these geographic features is affected by policy. If rural definitions were designed around these geographic features, fewer systems may be necessary, he suggested.
Johnson reiterated the political economy of rural definitions mentioned earlier. These classifications determine program eligibility, and eligibility generates economic rents. Potential economic rents generate incentives to influence and change eligibility criteria. Discrete classes, as opposed to graduated scores, lead to anomalous spatial outcomes. If the definitions are not right, outcomes are very different than what the program was designed for.
The challenge, he said, is to determine a set of simple definitions that line up with a number of different programs, suggesting ERS could work with each of the agencies that needs and uses these definitions. Many
definitions are needed, but perhaps the number should be reduced to the simplest set that meets as many needs as possible, he said.
STATEMENT BY KENNETH JOHNSON
Kenneth Johnson focused on rural definitions used in research. He said from a researcher’s point of view, a rural classification system must reflect contemporary rural America, but it also has to recognize the importance of longitudinal compatibility. His work with Al Nucci and Larry Long at the Census Bureau (Johnson et al., 2005) and recent updates compared county metropolitan classifications in 1963 to those in 2013. He found that 752 counties identified as nonmetropolitan in 1963 had been redefined as metropolitan by 2013. These reclassified counties contained 64 million people in 2013, and their transfer from nonmetropolitan to metropolitan areas accounted for the gain in the proportion of the U.S. population that was metropolitan between 1963 and 2013. In contrast, there was no change in the proportion of the U.S. population who resided in counties that were metropolitan in 1963 and remained metropolitan in 2013. Thus, all the growth in the proportion of the population that is metropolitan has come from reclassification of nonmetropolitan counties to metropolitan status. Reclassification, he said, is an important issue.
In recent work, Johnson used a combination of USDA-ERS codes and his own judgment to group rural and urban counties. He noted distinct demographic differences between the core counties of large metropolitan areas of 1 million or more and their suburbs. There were also differences between population redistribution patterns in these large urban areas and those in smaller metropolitan areas. In rural areas, nonmetropolitan adjacent and nonadjacent counties also exhibited critical demographic differences from one another.
Johnson said that the population and land area distribution of the United States in 2014 show that nonmetropolitan areas included 72 percent of the land area and 14.5 percent of the population of the United States. These nonmetropolitan areas also produce most of the country’s food, timber, water, and clean air, so they represent important sources of ecosystem services to the nation.
Rural America continues to experience significant demographic change, he stressed. A rural classification system must facilitate tracking this change. In order to understand the redistribution of the American population, a timely set of demographic information is also needed. For example, data from the most recent population estimates show dramatic variability in population change in rural America. To understand the details of both short-term and long-term demographic change, data are needed on the components of population change (natural increase and
net migration). Just looking at population change does not explain how a county is changing. Counties are the lowest level of geography for which birth and death data needed to calculate natural increase and net migration are available. Only some states have birth and death records for cities and towns. Thus, to understand population redistribution in the United States, county-level data are imperative. To him, this underscores the need for county-based rural classification systems.
To understand recent demographic trends, it is useful to view them in historical context, Johnson said. In the 1990s, rural America reflected the very familiar pattern of nonadjacent counties growing less than adjacent counties, primarily because they received less net migration. In contrast, a distinct slowdown in rural demographic growth is clearly evident in the 2000s. He provided information on rural demographic change using the Census Bureau’s 2014 demographic estimates for adjacent and nonadjacent counties for four time periods: the early 2000s; the economic boom; the recession; and the post-recessionary trends. These data reflect a dramatic slowdown in migration to rural America, leading to an actual rural net migration loss in recent time periods. They also reflect the rare phenomenon of adjacent counties growing less than nonadjacent counties in the most recent time period. To see this, contemporary data on natural increase and net migration are essential, he said. Clearly, whether considering long-term, intermediate, or short-term trends, good data and a consistent classification system are important to making comparisons between the urban and rural areas of the United States.
Rural America is a simple term describing a very complicated place, Johnson observed. A rural classification system must reflect this complexity. He said he is not convinced that any one rural-urban classification system can reflect the variability, and rural classification systems must be multidimensional and used in combination. For example, recreational and retirement counties are the fastest growing parts of rural America, but other rural areas contain slower growing agricultural and manufacturing counties. Manufacturing counties were the focus of rural economic development programs for decades and were expected to be where most new rural growth would occur. They did experience significant population growth for several decades, but recent data show population growth in manufacturing counties has sharply diminished. Farm counties continue to grow slowly, as they have for decades. He noted an urban-to-rural classification system based solely on population density cannot reflect this variability and need more than a single dimension to reflect the complexity and spatial diversity of rural America. Fast-growing recreation and retirement counties often exist in close spatial proximity to manufacturing and farming counties, which cannot be captured on a simple rural-urban continuum.
Classification systems are important because they inform policy making, Johnson stated. The policy implications of rural demographic change vary across county types. For example, recreational counties have experienced a large influx of older adults that is accelerating through time. In contrast, rural farm counties have been losing young adults for decades. These trends and their policy implications cannot be addressed by a simple straightforward rural-urban continuum, no matter how sophisticated. He said a multidimensional view of what is happening in rural America is important to facilitate understanding by researchers and policy makers.
In addition, long-term trend data are essential to capture the movement of people over time. For example, the inflow of older adults—the baby boomers—to recreational counties slowed because of the recession, but Johnson said he expects the growth to resume. Such areas will need to deal with the implications of population growth, the environmental impacts of such growth, and the need for services for an older population. And, because many of these older migrants from urban America are experienced in dealing with bureaucracies, he said they will exert considerable influence on rural policy.
Summary
Johnson summarized the importance for a rural classification system to reflect contemporary rural America but recognize the importance of longitudinal compatibility; facilitate the timely analysis of demographic trends over both the short and long term; reflect the complexity and growing diversity of rural America; and be useful for policy making and for research, such as by providing continuum measures for researchers and categorical classification systems for policy makers.
STATEMENT BY ROSE OLFERT
Olfert focused on data requirements for rural classification.
Rural Economics Research Questions
Olfert described key research questions. An old, still useful question is the size and role of rural communities, she said. Size and role remain important for people in rural areas, researchers, and policy makers. However, she said, the ways in which size and role have been approached has changed over time.
Closely related to identifying rural communities is investigating the reasons for rural population change—growth, decline, and migration patterns. There is a very rich literature on migration patterns, she noted. One
reason for population redistribution over space is the location decisions of individual firms. Areas with more economic activity are more likely to draw new industries, which results in more employment or income-earning options in a spatial context. The relationship between urban/ metro areas and the surrounding rural areas also remains a very important area of research, she said. Researchers want to better understand to what extent urban or metro growth benefits the surrounding rural areas, or whether there are negative effects.
Most of the research that comes out of the empirical investigation of these questions has very strong policy implications, Olfert said. One may consider whether policies are required. If they are required, what kinds of policies? Do they seek to influence or do they seek to accommodate the changes that are going on?
Geographic Units for Economic Analysis
Olfert stated meaningful geographic areas are essential for this economic analysis. She noted functional economic areas, once a fairly popular concept, appear to have come back into vogue. Functional economic areas are areas that are relatively “closed” in that people both live and work in the area. They earn and spend their incomes in the same area, and they access public and private services and amenities within it, although it does not mean those boundaries are absolutely closed. These functionally cohesive regions compete globally for economic activity and population, within a province or state, within the country, and in global terms.
The regional or the functional economic area population size and characteristics, its economic structure, the industry structure of that area, and the amenities will be the determining influences for growth/decline in population and employment through migration patterns. A natural increase in population is also important, she said, but that increase will be very closely related to the age structure of the population that is on net attracted to the region.
Population size and characteristics will determine whether critical mass is being achieved within the region so that the demand thresholds to support various kinds of economic activity or population service are met. They will also determine whether there is the access to markets required to attract firms to the area and the potential for some urban agglomeration economies to be realized within the area to achieve higher productivity and lower costs of production, including knowledge spillovers. She said these areas are also the appropriate policy targets as they include both the costs and the benefits of infrastructure development and service delivery. Ideally, the population in these areas are both the taxpayers and the recipients of a bundle of goods and services, making it easier to persuade them
to participate in economic development if they are also going to realize the benefits in terms of employment or income-earning opportunities.
Area Classification Requirements
Olfert noted it is essential for researchers to have classifications that can be used to approximate these functional economic areas. Metropolitan statistical areas for the most part achieve this for urban-centered areas, she observed. Commuting sheds are a reasonable approximation. Economic activity and population are increasingly concentrating in metro and urban areas. The best rural development strategy for nearby rural communities is urban-centered growth.
The metropolitan statistical areas rely on commuting sheds, which Olfert argued is a reasonable approximation for functional economic areas. Only commuting flows are being measured, but those commuting flows represent many other things that demonstrate the economic dependency within the commuting sheds around urban cores of metropolitan statistical areas. The larger and more diverse the metro core, the larger will be the population and commuting sheds. That also represents a greater market size, and the fact that these metropolitan statistical areas probably have the highest order of services, such as full-service dentists or lawyers, within the area.
Olfert suggested an advantage to flexibility in outlying areas to be included in terms of the percentage commuting for counties near the borders that are included in the metropolitan statistical areas. As noted earlier, Canada uses a 50 percent commuting rate threshold to attach counties to core areas; the U.S. threshold is 25 percent. Maybe the right threshold lies someplace in between, she suggested, but the information to conduct analysis with either tighter commuting sheds or relatively more generous commuting sheds needs to be available.
Metropolitan areas are similarly useful, Olfert said. Outside the metropolitan statistical areas, these areas are defined by their economic bases and distances from the metropolitan statistical area. Distance from the metropolitan statistical area has been shown by empirical research to be important; for example, even if an area is heavily dependent on mining oil and gas, a company may need an accountant, lawyer, or other service. She said it is probably not only distance to the nearest urban center or a metropolitan area that is important, but also distance to the top of the hierarchy, the largest center.
That is the first level of area classification that is important. Within that, Olfert said, the heterogeneity within the metropolitan statistical areas needs to be recognized and, in that context, where a rural-urban distinction is important. The population density of the counties and the distance
from the core will determine the infrastructure needs and transportation costs. There is a need to have some recognition of that heterogeneity within the metropolitan statistical areas, and she said maybe rural-urban definitions that recognize density and size are important.
Beyond the metropolitan statistical areas, rural and urban distinctions are very useful in terms of population size, density, and distance. She commented that the 2,500-population cutoff was adequate at the time that it was initiated, but asked what population threshold size now represents a size to support the full range of urban activities. The answer will probably help define what is now urban. Referring to Ken Johnson’s presentation about the research need to compare areas over time, she said a current definition can be observed back through time and a historic definition can be moved forward through time. Counties are added to metropolitan statistical areas over time as the definitions change at each decade, she pointed out. At each decade, what seem to be more “rural” populations and counties are becoming dependent upon an urban core for employment. Even though a county may look rural, it may not be in the sense of economic dependency and integration with the urban core.
Summary
Olfert summarized by saying that functionally integrated urban centered regions, differentiated by rural and urban especially at the periphery, are required as spatial units for data analysis. The nonmetro regions should be defined by their economic base. Some current definitions probably approximate what is needed, but there is fine-tuning to consider in terms of threshold commuting percentages and dated definitions of rural and urban. She noted consistency or comparability over time is important for research.
Olfert observed the empirical research to address the research questions and issues raised throughout the workshop are very data intensive. She noted she is impressed with the quantity and the quality of the data available. Requirements for data will likely become more onerous as statistical techniques become more sophisticated.
OPEN DISCUSSION
Waldorf said the presentations demonstrate the difficult task of the workshop. On the one hand, applied researchers want more simplicity and authenticity. Other researchers want more specific details such as recreation and retirement communities. Both groups stressed issues related to suburban, metropolitan, and internal heterogeneity.
Mark Partridge observed that as an outsider, he thinks Congress
needs to write language in legislation that targets their issues of interest. O’Brien commented there should be a continual feedback loop, where every five to seven years previous issues or problems are addressed in the Farm Bill or other legislation. From his experience with rural definitions in the Farm Bill, it is difficult to make changes.
James Fitzsimmons remarked the diversity of classifications, while it causes confusion, is a success in that there is no single classification that can be manipulated to fit everybody’s needs. Fitzsimmons said that maybe it makes more sense to consider program-specific classifications, perhaps with a smaller number of classifications as suggested by Tom Johnson.
O’Brien commented his statement sounded like nothing changes, but changes do occur. That is why, in his view, it is critical that policy makers have the right information when things change. He noted the RD report recommended one definition, but with a way to prioritize projects in different programs based on a set of variables.
David Brown observed that when the workshop steering committee designed this session, members had an idea that researchers, policy makers, and program administrators should be in conversation and should be integrating their efforts. He also asked why none of the eligibility thresholds discussed is based on changes in the number of households. Parker responded that he did not know of any, but it would be interesting. Particularly in rural areas, he has seen families moving in with other families or multiple families in a single household.
John Logan said the session clarified for him the dilemma for ERS to have a reasonable, rational classification system, but one that can be applied to programs. He referred to O’Brien’s point that certain kinds of areas, given their character, have concentrated unmet needs, and they do not have the governmental or fiscal capacity to meet those needs. Logan said it was curious that need and capacity are not measured directly and that the concept of the urban-rural dimension is very loosely associated with those two criteria. Logan also asked about spatial scale, such as counties, towns, or small populated areas. He said as a researcher that would be the first thing he would want to know. He referred to discussion earlier in the workshop about the county level, but asked about the scale for different problems. Until the answer to that question is known, Logan does not think there will be a lot of clarity about how to deal with it.
He said measurement issues are familiar in urban and rural America. The first issue is to measure need and fiscal capacity at the level of governmental units. He suggested counties are probably the relevant government units. For the issue of urban services, he suggested parts of a county are relevant to measure in some places, but municipalities in others. The division of responsibility between government units is very significant, he
stressed. It may be that in rural America, it is usually the county. However towns and townships may matter in terms of governmental capacity in some parts of the country.
Logan said in terms of need, the county seems like a large unit, but subcounty units are very small. For small areas, the only available data are from the American Community Survey (ACS). He said he has become wary of ACS data. He said that if subcounty-level estimates of need are to be used, he would base them on the 2000 Census.
Michael Ratcliffe observed the discussion has centered on geographic areas and specific types of data, but not management of data. He commented on the computing power, database structures, and ability to store vast amounts of metadata about units at all levels of geography that would support building whatever level of complexity and dimensionality is appropriate.
Mark Shucksmith commented on O’Brien’s point about the tension between need and having an impact, which varies greatly among the countries in the European Union. They have developed a series of concepts related to territorial cohesion, territorial potential, and territorial capital to try to get around the direct issue. In terms of political realities, he asked if the classifications are supporting the status quo in terms of political constituents as the constituencies are organized around the classification. To him, this seems to be a circle that prevents change. He suggested that to change the classifications, it is important to consider the constituencies that are not organized and address how they could be organized to allow changes to the classifications. He wondered whether that might be driven by the questions of needs, unmet needs, or impact.
Michael Woods commented on the ERS Natural Amenity Scale1 definitions. He said as an outsider looking in, that classification seems to address issues related to the diversity of rural areas, and economic drivers of difference between disadvantaged and rural areas
Marca Weinberg (ERS) remarked that missing from this discussion is a mention of ERS resource constraints. The agency has three people who work on this and it is only part of their positions. She also said as a statistical agency, ERS does not develop definitions or statistics to serve political purposes. ERS measures its performance by whether or not it is having impact and helping to inform decisions. Weinberg said that ERS is lucky to have a positive relationship with O’Brien, and they appreciate when their work gets used. What she said she would like is a classification system that is statistically reliable, justifiable, based on the best available science, and useful. It is critical for ERS to understand how the system
_______________________
1See http://www.ers.usda.gov/data-products/natural-amenities-scale.aspx [November 2015].
will be used, but its development should not necessarily driven by the constituents and the politics.
O’Brien praised the workshop and his partnership with ERS while at Rural Development in USDA and at the White House Rural Council. He noted the White House Rural Council includes all domestic agencies that do rural work and he offered to be a conduit to the other agencies. If the work on developing a new or revised rural classification scheme moves forward and ERS wants immediate feedback, he could pull together a small group of federal policy implementers.
He added the move toward evaluation-based budgeting and policy making has been a long time coming, but is coming with development of greater data tools. From his experience, rural programs and policy making are slower than others on performance-based budgeting and policy making.
O’Brien concluded that one of the great challenges of government is that resources should go to places that are making an impact or are moving policy in the right way. But in rural places with a lack of capacity to measure or even apply for programs, there is a huge fear of a downward spiral, which he characterized as central to the workshop conversation.
This page intentionally left blank.
