Community Stabilization Index

Methodology1 and Applications

Prepared by the Community Development Department2

Please send questions or comments to

The Community Stabilization Index (CSI) is a composite index that aims to provide community leaders with a relative measure of local housing market conditions, with a particular focus on recovery potential. The index is specific to conditions at the zip code level, and is comparable across all zip codes within a county or larger geographic area. Another helpful feature of the CSI is that periodic recalculations allow leaders to track relative changes in housing market conditions through time at the zip code level.

The CSI synthesizes several variables into a single comparable measure of recovery potential. However beneficial, this simplification should not deemphasize the importance of tracking underlying and other available housing variables, nor preclude understanding the limitations of this tool.

Data sources

The index draws data from Lender Processing Services, Inc. Applied Analytics (LPS) and the Federal Reserve Bank of New York’s Consumer Credit Panel. The LPS dataset includes loan-level servicing data for both securitized loans and loans held in portfolio from the top 10 residential mortgage servicers in the nation and others, covering about 65% of the US mortgage market. Smaller servicers have less representation.  The Consumer Credit Panel is a nationally representative 5% random sample of all individuals with a social security number and a credit report.  The database contains approximately 40 million individuals each quarter and includes household-level credit and debt, including credit cards, auto loans, student loans, mortgages (separately for first and second liens), and other student loans.


Records in LPS include active and inactive loans. The status of active loans can be current, delinquent, or in foreclosure. Inactive loans are those loans on properties that have moved into REO (Real Estate Owned) status, have been transferred to another servicer, or terminated. Only first-lien loans on residential properties are included in the analysis. The index is comprised of six components calculated for each zip code:

  1. Loans in 90-day delinquency: This component represents the percent of active loans that are at least 90 days delinquent in a given month.
  2. Loans in foreclosure: This component represents the percent of active loans that are in foreclosure status in a given month.
  3. REO: This component represents the ratio of Real Estate Owned (REO) properties to the number of active loans in a given month.  Inactive loans related to properties in REO status add to the shadow inventory of the zip code.
  4. Originations-to-shadow-inventory (SI) ratio: This component represents the ratio of originations in a given quarter to the number of REOs, foreclosures, and loans greater than 90 days delinquent in a given month.
  5. Change in median home value: For this component, we calculate the median estimated value of homes in the zip code for two time periods: 2005, the year prices peaked, and the most current full-year available. In the case of a purchase, the value refers to the sales price. If the first-lien loan is originated due to a refinance, the value refers to the appraisal amount. The index tracks the percent change of these two median values.
  6. Non-mortgage credit delinquency: This component represents the total number of individuals with accounts at least 60 days delinquent or in severe derogatory status, divided by the total number of credit holders, for a given month.  Non-mortgage refers to auto loans, credit cards, consumer finance, retail cards, and student loans.

For each zip code, all components are normalized to a scale of zero to one based on the zip codes’ relative level of distress with respect to other zip codes in the county or larger geographic area. Thus, for each of the components, the most distressed zip code—say, the one with the highest foreclosure rate—is assigned a value of one, and the least distressed is assigned a value of zero. The composite index, a simple average of its components, is also normalized to a zero-to-one scale. A higher score on the index indicates a more distressed housing market with fewer signs of recovery.


In general, greater potential for the local housing market and neighborhoods to recover may be observed via:

  1. A decreasing influx of properties entering the high delinquency, foreclosure and REO processes and an increasing outflow of properties from foreclosure/REO back into the market or into the hands of local institutions, such as land banks.
  2. Consumers’ positive expectations of stability in the area as signaled by lower depreciation of home prices and higher level of new mortgage originations relative to the number of properties that are vacant and in REO, foreclosure, or greater than 90 days delinquent.
  3. Decreasing non-mortgage delinquency rates indicating improved household finances, this in turn can lead to fewer mortgage delinquencies and foreclosures.

CSI components

Components 1-3 of the index relate to the inflow-outflow perspective of recovery, while components 4 and 5 relate to the positive expectations interpretation of recovery potential. Finally, component 6 measures non-mortgage signs of household financial stability that could influence housing outcomes. Overall, the composite index aims to reflect the health of the overall housing market across zip codes, given the high negative impact of the mortgage foreclosures crisis.


Following is a series of maps that, for each county, depict the weighted index components individually and then as a composite stabilization index.

FusionCharts will load here


Akron MSA, Ohio

Allegheny County, Pennsylvania

Cuyahoga County, Ohio

Dayton and Springfield, Ohio MSAs

Erie County, Pennsylvania

Fayette County, Kentucky

Franklin County, Ohio

Hamilton County, Ohio

Kenton County, Kentucky

Lucas County, Ohio

Mahoning County, Ohio

Stark County, Ohio

1 Adapted version of the Real Estate Owned (REO) Stabilization Opportunity Score (SOS) developed by Kai-Yan Lee at the Federal Reserve Bank of Boston. An earlier version of this index was posted to our site in 2009. This 2011 version makes use of newly acquired data and revises the methodology.

2 With graduate interns Youngme Seo (2011) and Nick Fritsch (2012)