Stay Informed
FAQs
Why is outcome data missing for some states?
Why are some outcome measures unranked?
How do you calculate the ranks and grades in the Scorecard?
Why aren't states given an overall grade?
How did you select the 12 policy measures?
How did you decide what constitutes a "strong policy"?
How current is the policy data?
Why isn't my state's recent policy success reflected in the Scorecard?
Why do you use the Survey of Income and Program Participation (SIPP) for net worth and asset poverty data in the Scorecard?
Why does the Scorecard data on wealth inequality differ somewhat from other data I’ve seen published on this topic?
Why do the foreclosure data and rankings in the Scorecard differ from other rankings I've seen published by other sources?
Does data in the Scorecard on households and people of color include Native American communities?
How is asset poverty calculated?
What is the difference between asset poverty and liquid asset poverty?
Why does the income poverty rate in the Scorecard differ from the official poverty rate?
Does the homeownership rate data include mobile or manufactured homes?
Can I use the data from the Scorecard in my publications and presentations? How do I cite the Scorecard?
Why is outcome data missing for some states?
It’s true that there are a number of outcome measures for which data are not available for some states. This is mainly the case for states with comparatively small populations where the data source does not have a large enough sample size from that state to produce a reliable statistic. In some states, for example in Arkansas, there are not enough households of color surveyed to provide a sample that can credibly or accurately reflect that population’s economic characteristics. There is a more detailed explanation of how CFED makes decisions about including data for certain states in the methodology section.
Why are some outcome measures unranked?
There are 16 unranked outcome measures in the Scorecard, and there are two reasons that a measure is unranked:
- Insufficient Data: Data may be unavailable because of insufficient sample sizes in the data source, or, for measures from the Survey of Income and Program Participation, because the data are too imprecise to publish or rank (see here for more information). If fewer than 35 states can be ranked for a given outcome measure, that outcome measure is not ranked.
- Too Little Variation between States: Two measures in the 2013 Scorecard (“Homeownership by Gender” and “Four-Year Degree by Gender”) are unranked because there is too little variation in the data between states to produce meaningful ranks. In other words, all of the states are so similar that it would be misleading to, for example, assign a rank of #4 to one state and a rank of #28 to another when the underlying data is virtually the same for both states. Further information can be found in the methodology section.
How do you calculate the ranks and grades in the Scorecard?
To calculate a state’s outcome rank for each issue area, the state’s ranks for each individual measure in that issue area are averaged (all measures are weighted equally). The lower the average, the better the state's outcome rank for that issue area. Using this rank, a grade is assigned to each state. Grades are assigned on a curve: states that rank from 1 to 10 earn an A; from 11 to 20 earn a B; from 21 to 36 earn a C; from 37 to 46 earn a D; and from 47 to 51 earn an F.
The overall outcome rank is calculated by averaging the ranks across the issue areas to generate an overall score (all issue areas are weighted equally). The lower the overall score, the better the state’s overall performance in the Scorecard. The overall score for the states is then ranked from 1 to 51.
Why aren't states given an overall grade?
We try to make the Scorecard as useful for policymakers and advocates as possible; and to help do this, we frequently seek input from end users on how to improve various aspects of the Scorecard. Last year, input from some end users indicated that the overall state grade was not particularly helpful to some because it often masked major issues facing their state. For example, if a state gets an overall “B” but has a “D” in the education issue area, that overall “B” grade makes it harder to build and deliver a powerful message to some target audiences. In addition, an overall rank provides states with more precise information than the overall grade: if a state received a “C”, it could have been ranked anywhere from 21st to 36th.
How did you select the 12 policy priority measures?
We selected 12 policies from the 33 that we include in the Scorecard for several reasons:
- Collectively, the 12 policies reflect a comprehensive financial security and opportunity agenda. Because they are drawn the five issue areas in the Scorecard, they include all elements of a financial security and opportunity policy agenda (Financial Assets & Income, Businesses & Jobs, Housing & Homeownership, Health Care and Education). They also encourage advocates to build coalitions across advocacy networks for policy change.
- The 12 policies are ambitious, yet achievable. All policies have been well-tested on the ground and vetted in state houses across the country. We selected policies that we believe are achievable in the next two to five years, yet may be a reach for some states.
- The policies selected are comparatively straightforward and easy to understand. Policymakers, the media, and others who may be new to asset building can intuitively grasp the logic of and need for these policies.
How did you decide what constitutes a “strong policy”?
CFED defined criteria for what constitutes a strong policy based on extensive research and advice from the leading experts in their respective fields. Criteria vary by policy and are described in detail in the Resource Guides and Policy Briefs for each policy priority.
How current is the policy data?
We strive to include the most to up-to-date policy data available for all of the policies in the Scorecard. Research on all 12 of the policy priorities was conducted and data was collected after state legislative sessions ended in the summer of 2012. With a few exceptions, the 12 policy priorities reflect policies adopted as of the end of September 2012. For the additional policies, we rely on the information that is available from outside sources; the data for these measures reflects the most recent information available. Each policy-specific web page contains complete references documenting the year the data is from and providing links to original sources where relevant.
Why isn’t my state’s recent policy success reflected in the Scorecard?
We do everything possible to ensure that the Scorecard contains the most up-to-date policy information available as of the annual data collection cutoff date (currently September 2012), including working with Assets & Opportunity Network Lead State Organizations to review their state data. Despite our diligence, sometimes we still miss something. If you see information about any of our policy priority data you feel is incorrect or out of date, please drop us a line at Scorecard@cfed.org to ask us about it. We’d appreciate it!
Why do you use the Survey of Income and Program Participation (SIPP) for net worth and asset poverty data in the Scorecard?
The SIPP is considered by most researchers to be one of only three national data sets that contain reputable information about the assets and liabilities of U.S. individuals and households – the other two are the Federal Reserve Board’s Survey of Consumer Finances (SCF) and the Panel Survey of Income Dynamics (PSID) conducted by the Institute for Social Research at the University of Michigan. We consider the SIPP data to be the best of these data sources for purposes of the Scorecard for two reasons:
- The SIPP contains data from over 40,000 households and the data is representative at the state level for every state. In contrast, the total sample sizes for the SCF and PSID of 4,500 and 7,500 respectively. The larger sample size makes it possible to perform state-level analysis of the data in a methodologically rigorous way.
- The larger sample size and the methodology behind the SIPP make it possible to document wealth and income patterns by broad economic and demographic characteristics (e.g., income, gender and race).
For more information on CFED’s methodology in calculating estimates from the SIPP, click here.
Why does the Scorecard data on wealth inequality differ somewhat from other data I’ve seen published on this topic?
As mentioned above, there are several public data sources that contain information on net worth. Each of these sources uses its own set of survey questions and sampling technique. Each is also administered at different moment of time. For example, the SIPP assets & liabilities module that is the source for Scorecard data is usually collected every 18 months by the U.S. Census Bureau, while the Federal Reserve’s Survey of Consumer Finances is fielded every three years. Different time periods for conducting interviews, and different samples of individuals, contribute part of the explanation for differences. What is important is that the data from each of these data sources reflect the same general trends in findings. There are a few other basic factors that contribute to differences between the data in the Scorecard and other estimates:
- The Scorecard provides an estimated asset inequality ratio for white households as compared to all households of color, whereas some other researchers and organizations publish disparities between whites and specific communities of color, i.e., African American or Hispanic or Latino households.
- The Scorecard provides disparities between single female-headed and single male-headed households, instead of disparities between all male- and female-headed households so as not to conflate issues of gender and of family and marital status.
For more information on asset inequality and asset ownership among different demographic groups see the recent research published by the Pew Research Center, and also research published by the Institute for Assets and Social Policy.
Why do the foreclosure data and rankings in the Scorecard differ from other rankings I've seen published by other sources?
The Scorecard’s foreclosure and delinquent mortgage loan data comes from the Mortgage Bankers Association’s National Delinquency Survey. While RealtyTrac is a more frequently cited source of foreclosure data, we believe that the methodological differences behind the MBA estimate better suit their measurement for the purposes of the Scorecard. Where RealtyTrac relies primarily on public records to calculate their foreclosure filing totals, the MBA uses data on outstanding mortgages provided directly by mortgage servicers. In addition, RealtyTrac’s rate measures the percentage of foreclosures per total housing units (which includes some apartments), rather than housing units with a mortgage. The MBA uses data on outstanding mortgages provided directly by mortgage servicers.
Does data in the Scorecard on households and people of color include Native American communities?
Yes, with an important caveat. All of the public data sources that we use in the Scorecard include Native Americans in their sample methodologies. But while most large public data sources are designed to produce representative samples at larger levels of geography (e.g., state or national) or race (e.g., non-white Hispanic, Black, Asian), these sources do not include a sufficient sample size to provide representative and reliable data about specific small geographies (e.g., a county, city or reservation communities), or specific subgroups (e.g., immigrants from Southeast Asia or members of the Lakota Tribe). We know this is frustrating for leaders working to raise awareness about their communities among the public and policymakers. As explained by Tawney Brunsch, executive director of the Lakota Funds in Kyle, SD, “Many of these measures do not reflect reality in reservation communities – in unemployment, in housing, and in financial assets.” While we do not have an immediate solution to the problem, we hope that publishing both what is known and also what is not known from available data will help raise the importance of the issue of developing more robust data and research to understand the specific patterns of financial exclusion for communities of color, including Native communities.
We share the frustration of many Tribal leaders and advocates regarding the inadequacy of most public data sources to provide the kind of granular information that can enable more robust sub-state analysis. If you want to advocate for improvements in public data sources, you can contact Association of Public Data Users.
How is asset poverty calculated?
Asset poverty is defined as having insufficient net worth to subsist at the poverty level for three months in the absence of income. To calculate the rate of asset poverty, the first step is to establish an asset poverty threshold, or the total amount of money necessary to cover costs for three months. In the Assets & Opportunity Scorecard, we use a very conservative estimate of costs, which is equal to what one could cover with a poverty-level income. To get the total amount, we multiply the monthly federal income poverty level, based on household size as the reference point, by three. For example, in 2012, the monthly federal poverty level for a household of three was $1,591, so that household would be asset poor if it had net worth below $4,773. The next step is to determine the relevant assets to include in the calculation. In the Scorecard, we use total household net worth (assets minus liabilities), as measured in the SIPP. Assets included in the SIPP net worth calculation include, but are not limited to, home equity and real estate, retirement, business wealth, financial assets and vehicles. Liabilities included in the SIPP net worth calculation include, but are not limited to, secured debt such as mortgages or vehicle loans as well as unsecured debt such as credit cards or student loans. Once the asset poverty threshold and net worth calculations are established, we calculate the percentage of households that fall below the asset poverty threshold.
What is the difference between asset poverty and liquid asset poverty?
The method for calculating liquid asset poverty is the same as that of asset poverty, but the difference between liquid asset poverty and our traditional measure of asset poverty is the assets that are included in the calculation of resources that are available to the household. Liquid asset poverty only includes financial assets such as bank accounts, stocks and bonds and retirement accounts, i.e., accounts that can be liquidated quickly. It excludes equity in business, vehicles, homes and other real estate, and it also does not take any liabilities into account when calculating the resources a household has on hand. As liquid asset poverty only includes resources that can easily be converted into cash, it is a more realistic measure of the resources families have to meet emergency needs than asset poverty.
Why does the income poverty rate in the Scorecard differ from the official poverty rate?
The Scorecard provides the poverty rate for households in the Scorecard, rather than for individuals as does the official poverty rate released annually by the Census Bureau. We provide data on households so that income poverty is directly comparable to asset poverty and liquid asset poverty, which are also calculated for households. Additionally, our income poverty data comes from the American Community Survey (ACS) rather than the Current Population Survey (the source for the official poverty rate) because the ACS provides a more robust sample at the state level.
Does the homeownership rate data include mobile or manufactured homes?
Yes, the American Community Survey includes owner-occupied manufactured homes in its homeownership data. Manufactured housing is the largest source of unsubsidized affordable housing in the country, but due to the way that manufactured housing is titled and financed as private property instead of real estate, it is all but impossible for millions of manufactured homeowners to have the asset-building opportunities as other homeowners. More information about manufactured housing can be found at CFED's manufactured housing initiative, I'M HOME.
Can I use the data from the Scorecard in my publications and presentations? How do I cite the Scorecard?
Most of the information available in the Assets & Opportunity Scorecard is public information and may be reproduced with appropriate citation. Please cite the data as follows: “2013 Assets & Opportunity Scorecard. CFED. Data Source: … Retrieved Day, Year. The data source may be cited using the reference that appears under the data for each measure on this website under "Source."
For example, a citation for liquid asset poverty would look like this: "2013 Assets & Opportunity Scorecard. CFED. Data Source: Survey of Income and Program Participation, 2008 Panel, Wave 7. Washington, DC: U.S. Department of Commerce, Census Bureau, 2012. Data calculated by the Bay Area Council Economic Institute. Retrieved January 30, 2013.
If you have questions about data use, please contact us at scorecard@cfed.org.
Copyright © 2013 CFED — Corporation for Enterprise Development 
1200 G Street, NW Suite 400 Washington, DC 20005 202.408.9788
Powered by ARCOS