During the last several decades, the combination of growing challenges and declining resources has forced many cities and towns to become more strategic in their approaches toward revitalization. Many have turned to data-based approaches to understanding market conditions and determining appropriate expenditures and investments. These data and analytic efforts have frequently served as the basis for attracting nonprofit and for-profit partners in support of neighborhood and municipal change. The Reinvestment Fund’s (TRF’s) Market Value Analysis (MVA) is one such effort.
Key questions arise in considering how to target limited resources for city revitalization. Which areas should a city prioritize for cleaning up vacant lots, removing abandoned cars, or intensive code enforcement? How should the mix between public acquisition and demolition, and incentivizing private rehabilitation and new construction, vary by market type? Where are “brick-and-mortar” public actions and subsidies generally unnecessary because the market is moving along well on its own? The MVA helps make objective, rigorously analyzed, contemporary market data available to help answer these questions and inform decisions. It starts by assembling a substantial amount of data on an entire city. It then uses a statistical procedure to sort a city’s census block groups into categories or types based on their housing market conditions and offers guidance on the mix of public actions appropriate for each market type. Ultimately, the MVA provides an analytic basis for allocating and prioritizing public, private, and philanthropic resources in service of positive change.
The MVA was first introduced in TRF’s hometown of Philadelphia. Like many older cities with an industrial past, Philadelphia’s population and manufacturing sector peaked decades ago, leaving behind vast expanses of vacant and abandoned homes and factories. Philadelphia’s population in 1950 exceeded 2 million residents but by 1990, the population had declined 25 percent to 1.58 million. At a postwar peak in the 1950s, Philadelphia had more than 350,000 manufacturing jobs; by 1990, that number was fewer than 85,000. The loss of manufacturing jobs was severe and greater than the national decline in the manufacturing sector. Jobs in the service sector replaced the manufacturing jobs but did not match the wages or benefits. Add decades of “machine politics” and poor political decisions to this trend, and Philadelphia was brought to its knees in 1990 because of financial woes. With municipal bond ratings hovering near junk status and a potential bankruptcy hanging over the city, the state legislature created an oversight and financial vehicle that would save the city from bankruptcy and create a structure to begin to repair its finances.
Newly elected mayor Ed Rendell effectively used that tool and others to stabilize Philadelphia’s financial condition and ultimately to begin to turn the corner on decline. However, notwithstanding some extraordinary achievements—particularly in downtown Philadelphia—the neighborhoods of Philadelphia continued to suffer. Citywide, population declined further—albeit at a reduced rate—to 1.52 million in 2000, and those who left took their middle-class incomes with them. Mayor Rendell’s downtown successes helped stabilize Philadelphia economically, but life in Philadelphia’s neighborhoods was not appreciably better outside of the downtown. After two terms as mayor, Ed Rendell went on to serve two terms as governor of Pennsylvania and John Street was elected mayor of Philadelphia. As a long-time district councilperson and the president of city council, Mayor Street had represented a part of Philadelphia that was predominantly African American and had struggled with high levels of poverty, abandonment, and concentrated public housing. Mayor Street’s electoral promise to Philadelphia was that he would work to inject the nascent vitality of the downtown core deeper into the city’s many neighborhoods.
Mayor Street began to develop the Neighborhood Transformation Initiative (NTI), stressing that the effort had to be citywide, market oriented, and data driven in order to systematically address the range of blighting influences resulting from long-term neighborhood disinvestment. He turned to TRF to help build out the plan and devise a data-based framework to guide decision-making within the initiative. NTI ultimately grew into a $290 million effort designed to significantly reduce the number of vacant and dangerous buildings (a number that swelled to more than 25,000) and vacant lots (by 2001, there were an estimated 31,000 vacant lots in the city). TRF was a natural partner because, as a Philadelphia-based community development financial institution (CDFI), it had a long and successful history of investing in Philadelphia. Furthermore, TRF had highly respected leadership and public policy expertise along with capacity to develop action-oriented data analysis tools.
In April 2001, the City of Philadelphia released the first MVA prepared by TRF to a large and receptive gathering at a historic theater in downtown Philadelphia. The theater was packed with investors, politicians, media, and stakeholders ranging from community development corporations (CDCs) to private-sector practitioners. In describing every MVA market type, presenters also offered a set of activities and resources that would be in service of the market type and its residents. A little more than a year later, the city council would support the mayor, voting 16:1 to allocate the NTI funds based on the analysis the MVA provided. Thus, the NTI was born.
TRF’s MVA is a data-based approach to analyzing real estate markets. It is designed using five underlying assumptions, based in the original principles of NTI:
- Public subsidy is scarce and should be treated as a resource to catalyze a market, or clear a path for private investment, but in general subsidy cannot create a market where there is none;
- “Build from strength”—in distressed markets, those investments built on nodes of strength are most likely to be successful;
- All parts of a city (not just downtowns, midtowns, or those parts that are highly distressed) and its residents are “customers” for the programs, services, and resources of that city, and the challenge is to customize investments to the particular needs and capacities that vary across neighborhoods;
- Decisions to invest public, private, or philanthropic funds should be based on objective and rigorous analysis of market data—as should evaluation of the impact of those investments. Accordingly, all MVAs cover an entire jurisdiction, not a particular parcel or neighborhood. MVAs are designed to uncover the full dimensions of both market challenge and market strength;
- MVAs should rely on market data that reflect actual market activity (e.g., residential sales, mortgage foreclosures, new units permitted).
Since 2001, TRF has completed more than 30 MVAs in cities of all sizes across the country. The cities are on different growth trajectories (growing cities such as San Antonio or contracting cities such as Detroit), or are working to reinvent themselves from their industrial past (e.g., Philadelphia, Baltimore, or St. Louis). In several cities (e.g., Baltimore, Pittsburgh, Philadelphia), TRF has created multiple MVAs on a cycle of approximately three years. Each MVA offers a lesson in how to improve the MVA and ensure greater local engagement.
Typically, the MVA relies on a set of indicators obtained from local jurisdictions (i.e., administrative data). In general, an MVA uses (1) real estate sales transactions; (2) variability in the value of those transactions; (3) mortgage foreclosures; (4) owner occupancy; (5) mixture of commercial and residential land uses; (6) vacant land/buildings; (7) new construction/substantial rehabilitation; and (8) subsidized rental stock. Over time, and with the experience of working in various cities, TRF settled on this set of indicators because it symbolizes the sort of market data an investor might consider when evaluating an investment. These indicators are also generally available because they represent the sorts of data cities often track. Last, the field validation of the data demonstrates that, combined into the MVA, the data are effective in creating both quantitative and qualitative market separation. Although these are the indicators typically used in the MVA, we have found in some cities that additional indicators were necessary to properly distinguish markets (e.g., in St. Louis we included bank and investor sales of real estate as an indicator; in Detroit, we incorporated an indicator of sheriff sales resulting from property tax delinquency).
Most of these indicators are acquired at an address level and then aggregated to the census block group. In our experience, the census block group is the correct geographic level because it is large enough to ensure that the data are reasonably stable yet small enough to ensure that the mosaic of a place is revealed; larger geographies (e.g., census tract, neighborhood, ZIP code) obfuscate meaningful differences. The MVA uses administrative data rather than secondary data sources (e.g., census data) for a few reasons. First, administrative data tend to be more up-to-date. Markets can change rapidly. Although the recent waves of the American Community Survey (ACS) conducted by the Census Bureau represent a substantial improvement from the decennial census long-form data of previous decades, the block group level ACS samples are small and subject to large statistical errors. ACS data are also at least 18 months old and as many as six years old when released in the five-year waves. Second, administrative data are preferred because they represent actual conditions, not an estimation (e.g., home value according to the census) or recollection of a condition at a prior time. Third, because the data are mostly at an address level, we are able to generate measures of central tendency (e.g., mean or median) and variability (e.g., coefficient of variance of sale prices) for each block group. Finally, several critical indicators are not available from secondary data sources (e.g., the mixture of commercial and residential land uses, permits for substantial rehabilitation).
Once acquired, we clean and validate each database with two parallel processes. We validate data first by review with local subject-matter experts and then through fieldwork. The latter involves reviewing the data while driving through the streets of a city with a GPS locator. Validation will typically take us through at least 50 percent of the block groups in a city. In most cities where we have worked, local experts accompany us. Those experts could be from a planning or code enforcement department or even a nonprofit community-based organization. Field validation is a critical part of the MVA process and is one of the things that distinguish the MVA from out-of-the-box market analyses. Aside from the aspect of local engagement, which is itself valuable, we uncover data issues that could impede accurate conclusions. For example, in one city we found unusually low-value sales in an area; however, validation revealed the sales were vacant lots not identified as such in the database. In another city, the vacancy measure turned out to be a more accurate depiction of units that were vacant but could be occupied—as opposed to vacant and abandoned (which is what we thought we were measuring). Once we have faith that the data are correct, they are aggregated to census block groups, mapped, and subject to additional validation.
We use a cluster analysis to combine cases (i.e., block groups) based on all of the measured indicators into categories so that cases within categories are more similar with one another than they are with cases in other categories. Stated differently, within each category, block groups are very similar, but each category is very different. We then map and validate the results of the cluster analysis using a similar process as described earlier. (Figure 1) Cluster analysis is a mix of art and science and, therefore, field validation is again important. For example, it is important that the groups be statistically different from one another. The art emerges when inspecting a section of a city to observe whether differences seen on the ground match those on the map.
Once the MVA is complete, TRF works with local stakeholders to identify a subset of indicators to update on a regular basis. For example, many cities have ready access to sale transactions and foreclosures. These indicators can be updated quarterly or annually, giving those stakeholders the ability to understand how an area is changing along these critical dimensions. We have found that stakeholders seeking to evaluate broad market changes related to investment or programmatic activity may need the MVA to be completely reconstructed periodically to accurately capture new data as it is made available.
MVAs in Practice
A variety of organizations—including local governments, state agencies, the federal government (through technical assistance contract intermediaries), and philanthropy—have funded the creation of MVAs. Our recent experience in Milwaukee stands out as an example of an MVA funded by various sources: government, philanthropy, investors and community advocates. Each of the stakeholders engaged in constructing and validating the MVA brought their own perspective to the process, which contributed to the development of collaborative and coordinated evidence-based actions based on the MVA.
Typically, organizations use the MVA to guide key decisions about allocations of programs and resources. Baltimore, Philadelphia, and St. Louis have used the MVA to inform consolidated and comprehensive planning efforts, while Pittsburgh and Houston have used it to decide which projects to support with local, state, or federal incentives. Baltimore used its MVA to target code enforcement, and Milwaukee used its to coordinate funding from government and philanthropic sources. In Detroit, the MVA has been used in a number of ways, from helping the city target the proceeds from a federal civil rights settlement so that the funds could be maximally effective and consistent with the prescriptions of the settlement, to guiding infrastructure investments, to revising the boundaries of previously approved Neighborhood Stabilization Program (NSP) target areas. Pennsylvania and New Jersey used MVAs to help guide NSP plans for nonentitlement communities. Philadelphia and other cities with land banks are developing acquisition and disposition strategies based, at least in part, on the MVA. Finally, TRF uses the MVA on an ongoing basis in cities where we both invest in and develop affordable housing to target our efforts and assess change.
Figure 2 represents a prototypical template for an MVA implementation exercise that clients and stakeholders can use to think about a variety of programs, activities, and resources and how they might be prioritized and coordinated among different markets (represented as types A through I on the top of the chart). The suite of “activities” on the left side of Figure 2 will vary by the resources and programs a particular set of engaged stakeholders wish to prioritize using the MVA. In essence, the MVA facilitates the creation of a logic around matching the objective condition of various market types to the activities that might have the most potential for generating positive outcomes there. When resources contract, a guiding logic that can help diverse participants reach agreement on how to organize and target programs and activities is not an option but an imperative.
Most cities have been open about their MVA and post it publicly on sites they created (e.g., St. Louis, Baltimore, or New Orleans) or on TRF’s PolicyMap (e.g., Philadelphia, Milwaukee, Reading). By whatever means the data and analysis are conveyed to the public, revealing the analysis to a broad public audience lends itself to a more transparent process of decision-making, something we encourage.
Challenges in Developing and Using an MVA
Having conducted many MVAs in a variety of circumstances, we found a limited number of challenges that are similar from place to place. The following challenges fall into three categories: technical, political, and financial. Each category reflects difficulties in implementing data-driven decision-making, but our experiences demonstrate that all can be addressed successfully:
Access to administrative data can be a challenge
Cities have unequal capacity to supply the requisite data. Moreover, administrative offices are sometimes run by elected officials who may not report to a mayor. Therefore, they may not have the same interest in making “their” data available because they believe the MVA will not be relevant to them. This issue has never scuttled an MVA, but it has made the task more difficult.
Data visualization and labels
The MVA map distinguishes markets by shading block groups by color; sometimes those markets are given names. For example, in the original Philadelphia MVA, market types were labeled “regional choice,” “high-value appreciating,” “steady,” “transitional,” “distressed,” and “reclamation.” Although the labels are only meant to serve as shorthand, clients can be reluctant to label a market “distressed,” for example. Similarly, color can evoke emotion rooted in the history of many U.S. cities (e.g., red may be associated with the practice of redlining and coloring an area red may give the impression that resources will not flow to the area). We have dealt with these issues by having the recipients and stakeholders choose labels and colors that minimize potential discomfort. In the end, although clients or other stakeholders may contest the name or the color of a market on a map, they do not contest the way the data describe their area. As John Adams once declared, “Facts are stubborn things.”
Targeting vs. even distribution of resources
A tension recurs between strategies that emphasize an even (and frequently thin) distribution of programs/resources and those in which resources are pegged to eligibility, need, opportunity, and market characteristics. Consider a city with 10 local legislative districts, and the city has historically distributed to each district 10 percent of the community development resources—as scarce as they may be. Now, change that so districts receive an amount commensurate with what the data suggest about needs and opportunities. Even with an effective argument, the political reality is that not all local legislators command the same amount of dollars they did before. Evidence-based targeting of resources to markets, as opposed to an even distribution, is a bitter but necessary pill that can sometimes be easier to swallow if there is acceptance of the analysis and buy-in to the connection of objective data about a place and the resources that are appropriate to address issues in that place.
Sometimes stakeholders fail to take action on the findings presented in a completed MVA. This is most common in locations where TRF prepared the MVA without a champion in local government and a broad base of stakeholders who use the MVA as an organizing vehicle for change. Broad stakeholder support is certainly not a guarantee for success, but it can help. We have had particular success in, for example, Baltimore, Pittsburgh, and Milwaukee, where the MVA was funded by multiple sources and was developed with ongoing input from a set of stakeholders from every relevant sector.
MVAs are high-touch pieces of work that involve organizing people, organizing data, and a substantial amount of fieldwork. All levels of government experience the squeeze of declining resources, and philanthropy is not always comfortable with funding government activities. However, the value proposition of tackling an MVA in a city succeeds when the city and its stakeholders change behavior and make catalytic and effective investments that transform places.
Future Uses of the MVA
Data-based analyses and the resulting tools, such as the MVA, have application beyond simply measuring the real estate market to prioritize housing investment. There is, for example, an increasing awareness of the social determinants of health and how place-based organizing around economic stability, health care, education, social context, and the built environment can enhance the physical and psychological well-being of a community’s residents. The MVA incorporates several of the more frequently cited social determinant measures, in particular economic stability and the built environment. The MVA can help drive investment that will not only enhance the physical environment but also improve the prospects for healthy people and communities.
Social context is often represented by the extent to which there is free and open choice of housing without regard to race, color, or national origin. To that end, there is renewed public and governmental interest in the Affirmatively Furthering Fair Housing (AFFH) provisions of the federal Fair Housing Act (Act). AFFH requires all executive branch agencies to ensure programs are written and executed in a manner that will support the congressionally mandated purposes of the Act. At a minimum, we understand the “purposes of the Act” to include overcoming the legacies of segregation and concentration of racial/ethnic or low-income populations, ensuring equal access to community assets, and addressing people- and placed-based housing needs and disparities. The MVA and the data on which it relies can serve as a resource to facilitate AFFH efforts. When viewed through the additional lenses of racial/ethnic and economic segregation, the MVA points to places of opportunity and inequity where investment can not only transform a place but also address conditions adversely affecting racial and economic equity.
Finally, interest in understanding middle market areas is growing. The MVA is a tool to identify the location and conditions in a city’s middle-market places. Because these areas are not home to large concentrations of very poor people or rampant deterioration of the housing stock, many cities have not directed resources to these places in the past. The middle-market areas are a remarkably important part of the economic and social fabric of any city, and neglecting these places and their residents can have dire effects on the future prospects of cities. As Philadelphia State Senator Dwight Evans said in his remarks supporting NTI, “A neighborhood shouldn’t have to go through the process of becoming completely blighted before it can get help.” Market-based, data-driven analyses, such as the MVA, can help direct attention and provide the market justification for a set of public, private, and philanthropic investments necessary to sustain middle-market areas.
In sum, data are powerful, particularly when they are transparent, ground-truthed, and used as an organizing vehicle for community engagement. With objective, rigorously validated and analyzed data, we can encourage a robust discussion about a future for a place and its people.
 To contextualize the significance of this effort, compare NTI with the federal Neighborhood Stabilization Program (NSP), which uses funds allocated by Congress and distributed by HUD from two stimulus bills. Philadelphia received two NSP awards in 2009 and 2010, totaling slightly more than $60 million dollars. NTI, in 2009 dollars, is valued in excess of $360 million.
 For St. Louis, see City of St. Louis, Missouri, “Residential Market Analysis” (2014), available at http://dynamic.stlouis-mo.gov/mva/. For Baltimore, see City of Baltimore, Maryland, “Planning/Master Plans, Maps & Publications/Housing Market Typology” (2010), available at http://archive.baltimorecity.gov/government/agenciesdepartments/planning/masterplansmapspublications/housingmarkettypology.aspx. For New Orleans, see New Orleans Market Value Analyses 2009−2012,” available at http://nolagis.maps.arcgis.com/apps/OnePane/basicviewer/index.html?appid=623139ce8d3c4f83ade962b79e797164. For TRF, see The Reinvestment Fund, “TRF Policy Map” (Philadelphia, PA: TRF, 2014), available at www.policymap.com.
 See, for example, HealthyPeople.gov, “2020 Topics and Objectives, Social Determinants of Health,” available at www.healthypeople.gov/2020/topicsobjectives2020/overview.aspx?topicid=39#two.
 Philadelphia Daily News, April 19, 2001, p. 12.