New Page 1

In July of 2003, the Legislature of Pennsylvania took official notice of the growing mortgage foreclosure rate in Pennsylvania and passed House Resolution No. 364, which called upon the Secretary of Banking “to study residential lending practices in Pennsylvania, to identify trends in foreclosures and to document lending practices which are disadvantageous to Pennsylvania’s consumers and submit a report to the General Assembly.” 1

To understand what was driving the growing foreclosure rate in Pennsylvania, the Secretary of Banking hired e Reinvestment Fund to gather as much data as possible about foreclosures in Pennsylvania and systematically analyze trends and potential causes. Simultaneously, the Secretary convened an Advisory Group of representatives from the mortgage industry, legal services, advocacy groups and state government to provide the Secretary with guidance during the course of the study.

e fundamental goals of this study presented are to understand how Pennsylvania foreclosure trends compare to other places; develop a set of facts regarding overall foreclosure trends in Pennsylvania; and conduct a detailed analysis of foreclosure activity in communities across Pennsylvania upon which the Secretary of Banking can rely to make (or propose) requisite changes to law, regulation and policy. To do this, TRF:

n Conducted literature reviews of issues related to foreclosures, including traditional triggers of mortgage foreclosure, abusive lending, loss mitigation, and efficacy of housing counseling.

n Analyzed how traditional economic indicators affect foreclosure rates in Pennsylvania and across the nation.

n Collected and analyzed recorded information regarding the last four years of mortgage foreclosure filings in Pennsylvania.

n Conducted one-on-one interviews and focus groups with industry representatives, representatives of relevant county offices and housing assistance providers.

DATA SOURCES AND METHODOLOGY

U.S. Census Data & Census Estimates 1980, 1990, 2000, 2003 - e U.S. Census Bureau 1990 and 2000 Summary Files 1 and 3 data permit a categorization of counties in terms of any number of relevant social, demographic and economic characteristics (e.g., income level, housing value, owner occupancy rates, etc.). Census data is analyzed using statistical software known as e Statistical Package for the Social Sciences (“SPSS”) and GIS software known as ArcView. e U.S. Census also provides estimates of some of the characteristics for 2003.

U.S. Census 1980, 1990 and 2000 5-Percent Public Use Microdata Sample (PUMS) for Pennsylvania - PUMS is a compilation of a sample of individual Census forms collected by the Census. is form of the data allows for a very detailed and customized examination of the data.

Property Specific Sale and Mortgage Data- TRF developed a methodology employed in studies of both Philadelphia and Monroe County foreclosure activity. Unlike methods used in most studies of loans in foreclosure which look only at the loan in foreclosure, this technique allows TRF to trace a foreclosure filing back to the originating loan and lender and review the transactional history of a property in foreclosure. is distinction is important as most loans – particularly subprime loans - are not held by originating lenders anymore, but sold oftentimes more than once, in the secondary market.

To obtain this data, TRF queried each foreclosure property studied in the First American Real Estate Solutions (FARES) database. is database allows TRF – where property information is available – to document the transaction history on each property. e database records when a property was sold, at what price, and to whom; recorded liens; lenders or mortgage lenders involved; and the assessed value given the property by the County. is data allows TRF to determine the types of lenders involved in originating loans now in foreclosure, if the owner owes more on the house than it is likely worth, how long people in foreclosure have lived in their home, how much of the foreclosure activity is associated with loans to purchase a house or refinance an existing loan(s) and the geographic concentration of foreclosures.

A fuller study of Pennsylvania’s foreclosure activity is constrained, however, by the number of counties reflected in the FARES database. At this time, 14 of Pennsylvania’s 67 counties have property specific information in the database (see Map 1 for the locations of the 14-county study area). ese counties do, however, represent all or part of the larger metro areas in the Commonwealth and together account for almost 60% of the occupied housing units (homeowners) in the Commonwealth.

1) Philadelphia, Bucks, Chester, Montgomery and Delaware Counties (Southeastern PA)

2) Allegheny and Washington Counties (Southwestern PA)

3) Erie County (Northwestern PA)

4) Berks, Lehigh and Northampton Counties (Southeasten PA)

5) Dauphin and Lancaster Counties (Southcentral PA)

6) Monroe County (Northeastern PA) (TRF recently completed a study of foreclosure activity in Monroe. A copy of that study can be downloaded from the Department of Banking’s web site at http://www.banking.state.pa.us)

Foreclosure Filings - For the 14 areas with FARES coverage, TRF obtained listings of all mortgage foreclosure filings for the four year period from 2000 through 2003, inclusive. is represented a time consuming process as there is no centrally located office in the Commonwealth that collects filing data - instead, Prothonotary Offices in each county do. Complicating the effect is the fact that different counties capture different pieces of information regarding filings and, while some counties were able to provide filing data electronically to TRF, others have paper-based record keeping processes. e sheer volume of paper required TRF to systematically analyze only a sample of filings in some of these counties.

Home Mortgage Disclosure Act (HMDA) - TRF analyzed HMDA data for the period 1998-2002. ese data, together with the Census data, allowed an examination of the types of mortgage loans made and the characteristics of areas in which they are made.

Homeowners’ Emergency Mortgage Assistance Program Applicant Data - TRF acquired a full state set of data on applicants to the Commonwealth’s Homeowners’ Emergency Mortgage Assistance Program (hereafter, “HEMAP”) covering the last several years. Each record contains a date of application as well as the foreclosing lender and disposition by HEMAP. TRF analyzed all HEMAP applications, spatially and statistically, to reveal trends, geographic concentrations within the state and reason for each homeowner’s mortgage crisis.

Focus Groups and One-on-One Interviews - All TRF studies use interview or focus group results to inform and complement findings revealed by data analysis. Data findings can many times be confirmed or better understood by learning from practitioners and those who are experiencing first-hand what we observe statistically. TRF held a number of focus groups and interviews for this project including those with the following:

n Prothonotary and Sheriff Offices

n Mortgage industry representatives (local and national)

n Pennsylvania realtors

n Housing counselors

n Attorneys representing lenders and consumers

n Consumers over the last two years for this and related studies

Literature Review - TRF conducted a literature review to outline what is working in other parts of the country or in other industries to address problem trends identified in the analysis. Some of the following areas are discussed in this report:

n Causes of foreclosure

n Efficacy of housing counseling

n Mortgage loss mitigation programs

n Predatory lending laws and other laws designed to address abusive lending practices

Analysis - TRF collected and analyzed all of the above data – statistically and spatially – in an effort to produce a set of findings about why households are facing foreclosure in Pennsylvania, and where residents have been affected. e result of these analyses is the subject of this report and includes:

n Maps of foreclosures and HMDA analysis;

n Complete database of information regarding the 4-year foreclosure list;

n Written analysis of findings;

n Fact-based findings upon which the Commonwealth can act;

n Legislative, administrative and legal remedies revealed by the literature reviews, focus groups, and task force recommendations that relate to the causes of foreclosure in Pennsylvania.

Over the past few years, a variety of data began to signal expanding financial trouble for homeowners in Pennsylvania. Foreclosure rates were growing, Sheriff Offices around the Commonwealth were overwhelmed by the volume of transactions, the number of applications to the state for HEMAP assistance was increasing, bankruptcy rates were rising and more owners were cost-burdened (paying more than 30% of their income for housing costs) than ever before.

FORECLOSURE RATES STATEWIDE

Pennsylvania’s foreclosure rates continue to grow. In 2003, Pennsylvania had the 9th highest foreclosure rate among prime loans; and the 4th highest rate among subprime loans.

As a general matter, a mortgage foreclosure is a legal action that is defined in part as:

e process by which a mortgagor of real or personal property, or other owner of property subject to a lien, is deprived of his interest therein. A proceeding in equity whereby a mortgagee either takes title to or forces the sale of the mortgagor’s property in satisfaction of a debt. Black’s Law Dictionary (6th ed. 1990)

At the most basic level, a mortgage foreclosure action is usually started after an individual has stopped making payments on their mortgage (voluntarily or involuntarily). Unless those payments begin again, an arrangement is made with the lender, a consumer seeks and receives bankruptcy protection, or some other extraordinary event occurs, the individual is going to lose their home. e loss of a home through foreclosure adversely affects the homeowner and community in which they live along with the lender or investor who holds the loan.

Data obtained from the Mortgage Bankers Association of America (MBAA) show that the trend in foreclosures for the Commonwealth of Pennsylvania has been on the rise. Looking back to 1979, the typical quarterly percent of conventional loans in foreclosure was less than one-half of one percent. at figure rose steadily during the decades of the 80s and 90s. Since the year 2000, the percent of conventional loans in foreclosure rose from about one percent to one and one-half percent (see Figure III-1: Trend in Conventional Loan Foreclosures).

According to the MBAA, Pennsylvania’s percent of prime loans in foreclosure was .85% in 2003 and ranked the state as having the 9th highest prime foreclosure rate in the nation (see Figure III-2: Prime Foreclosure Rate by State). By comparison, though, Pennsylvania’s percent of subprime loans in foreclosure – orders of magnitude higher than the percent of prime loans in foreclosure at 11.9% - ranks it among the top four states in the country (see Figure III-3: Subprime Foreclosure Rate by State). FHA-insured mortgages had a foreclosure rate in 2003 of 4.5% (see Maps 2, 3 and 4: National Maps of Foreclosure Rates by State).

SHERIFF SALES

In just the last three years, an estimated 55,163 homes have been lost to Sheriff Sale in Pennsylvania.

Sheriff Offices report increases in the numbers of properties being exposed to and sold at Sheriff Sales across the Commonwealth. Data from 43 of the Commonwealth’s 67 counties indicates that 35,980 properties were sold at sheriff sale during the last three years. is represents an increase of over 14% during this three year period.

By comparing the total number of housing units in each reporting county with their actual number of properties sold at Sheriff Sale, TRF estimated the number of properties likely sold in the non-reporting counties and found that a total of 55,163 properties were likely sold in all 67 counties during this time period. is exceeds by some 20% the number of households in the City of Allentown (Pennsylvania’s 3rd largest city below Philadelphia and Pittsburgh).

An even greater number of properties have been exposed to sheriff sale, but not actually sold. Sheriff Offices report that many are simply not equipped to handle the volume of properties being brought to auction (see Maps 5,6 and 7: Sheriff Sales Completed).

FORECLOSURE FILINGS BYCOUNTY

In just the last four years, the number of foreclosure filings in the 14-county study area grew from 15,610 in 2000 to 20,737 in 2003 – an increase of 33%.

TRF collected foreclosure filing information from Prothonotary offices in 14 counties throughout the Commonwealth for the four year period, 2000 through 2003. (While foreclosure filings are known to have been rising for years before 2000, this four year period was the point of time analyzed for this study.)

Filing numbers indicate that, without exception, the number of foreclosure filings has risen in each county during the four-year period. Foreclosure filing increases were most dramatic in Erie, Washington and Allegheny counties where the number of filings grew by over 60% (see Figure III-4: Filings by County).

HEMAP APPLICATIONS

In 1996, almost 5,700 households applied to the Commonwealth for mortgage assistance. In 2003, 8,881 did so.

e Pennsylvania Housing Finance Agency (PHFA) operates a program called the Homeowner’s Emergency Assistance Program (HEMAP). Designed to help homeowners keep up with mortgage payments during periods of unemployment or illness, the HEMAP program is one of the most lauded housing programs in the state.2 Applications for mortgage assistance have almost doubled since 1996 and growth has been relatively even across the Commonwealth. e four maps shown here are standardized to show applications per 1,000 owner occupied housing units. (see Maps 8, 9, 10 and 11: HEMAP Applications Over Time).

Additionally, between 2000 and 2003, 4,222 households were approved for HEMAP assistance. Simply put, the presence of HEMAP saved (or postponed) 4,222 households from being subject to foreclosure and likely underestimates the true numbers of households in serious mortgage- related distress in the Commonwealth.

As the chart here shows, 2,527 residents within the 14-county study area and 4,222 residents in the Commonwealth as a whole remained homeowners as a result of HEMAP (see Figure III-5: HEMAP Approvals by Year).

e success of HEMAP can contribute to an understatement of the total number of failing loans within the Commonwealth. If a resident with a delinquent mortgage is approved for HEMAP, a foreclosure will not be filed. e number of foreclosure filings reported in this study would actually be higher were it not for the HEMAP program.

BANKRUPTCY FILINGS

Bankruptcy filings in Pennsylvania have grown at one of the fastest rates in the country over the last 10 years.

Research suggests that changes in the personal bankruptcy filing rate tend to mirror changes in the risk of default to a household. e logic is that the same set of consumer burdens which make a household more likely to default on a mortgage loan are the same burdens which make them more likely to declare bankruptcy. Personal bankruptcy filings nationally continue to be uniquely high and economists are debating its cause. One side of that debate believes a significant share of the rise in filings is due to the current law and to a lessening of the stigma associated with filing for bankruptcy.

e other side of the debate argues that the rise primarily reflects an increase in financial distress within the consumer sector.3

Bankruptcy filings in Pennsylvania have grown exponentially. Between 1990 and 2001, the number of bankruptcy filings in Pennsylvania grew by over 200% - the 5th fastest rate in the country (see Figure III-6: Change in the Number of Bankruptcy Filings).

HOUSING COST BURDENS

More homeowners are paying over 30% of their income for housing costs.

According to the U.S. Census Bureau, 16.7% of all homeowners were paying more than they could afford to own a home in 1990; by 2000 20.8% were. 7.3% of all homeowners are most severely burdened (paying over 50% of income) in 2000. Lower income households are far more likely to face housing cost burdens than any other income group.

Together, these data suggest that more homeowners are financially distressed. More are applying to the state’s HEMAP program for help in paying their mortgage, more are filing for bankruptcy, more are facing foreclosure and more are losing their home to Sheriff Sale. Understanding the nature and potential causes of this problem in Pennsylvania is the subject of this report.

Finding 1: Traditional factors alone do not explain Pennsylvania’s high foreclosure rate. If they did, the subprime foreclosure rate, in particular, would be at least 3 percentage points lower.

TRF reviewed a significant amount of the research conducted to date on foreclosures across the country and found a broad set of economic indicators that researchers suggest affect mortgage delinquency and subsequent foreclosure rates. Early foreclosure studies focused on two sets of factors that play key roles in mortgage default and foreclosure: the risk attached to the asset underlying the interest secured by the mortgage (home value), and the risk exhibited by the mortgagor. While these two factors continue to be primary focuses of much research, more recent work includes a look at homeownership rates, home appreciation, loan-to-value ratios, securitization in the mortgage industry, consumer debt, subprime activity, mortgage terms, share of FHA originations and financial distress caused by certain “trigger events” such as divorce, unemployment and medical catastrophes. (A bibliography of the research reviewed for this study is included in the Appendix.)

A look at the indicators in Pennsylvania reveals a rather murky picture. Loan to value ratios are increasing, home prices are appreciating slowly, and Pennsylvania’s already high homeownership rate has risen even further. At the same time, Pennsylvanian’s save more than their state counterparts, carry a higher average credit score, are less likely to be unemployed than their national counterparts, and divorce less frequently. In the end, some of these indicators may suggest reasons why the rate is growing – but they do not fully explain why Pennsylvania’s rate is so much higher than other states around the nation.

A REVIEW OF THE INDICATORS

Loan-to-value ratios are increasing in Pennsylvania.

Loan-to-value ratios are likely one factor in the Commonwealth’s growing foreclosure rate. Data from the Federal Housing Finance Board indicates that in 2003, the average conventional single family mortgage loan in Pennsylvania had a loan-to-price ratio of 76.9%.4 is was somewhat higher than the nation (73.5%). More importantly, the loan-to-price ratio in Pennsylvania has traditionally been lower than that in the nation. In 1994, however, the ratio in Pennsylvania began to exceed that of the nation and has remained higher. is is likely the result of the huge growth in subprime originations which began in the early nineties.

TRF’s research review suggests that higher loan-to-value ratios increase the likelihood of default because borrowers have less invested in the property. It is also true that high loan-to-value ratios increase the likelihood that the borrower will be in a negative equity situation during the early years of the loan. Higher loan-to-value ratios also suggest that borrowers have less savings. Were more savings available, they would likely have put more money down at closing in order to obtain better loan products and pricing. ese borrowers may lack the necessary cushion to keep up mortgage payments during a time of financial crisis, further increasing the likelihood of foreclosure.

Some studies have further suggested that the higher percentage of loans with over 90% loan-to-value ratios, the higher the rate of foreclosure. Nationally, the Federal Housing Finance Board estimates that 20% of all conventional loans in 2003 had an LTV of 90% or more.5 In Pennsylvania, 23% exceeded 90%. at said, almost half of the states in the nation had higher percentages than Pennsylvania.

Generally, Pennsylvania home values are appreciating at slower rates than the nation.

Our research review suggests that non-appreciating real estate markets tend to have higher foreclosure rates for two reasons. When faced with an economic hardship, a borrower living in an appreciating market may: 1) tap the equity in their home until the financial hardship passes; or 2) sell the home and walk away with some money instead of losing their home to foreclosure. Either of these may be a rational economic choice in an appreciating market. For the borrower in a non-appreciating market, without equity, the “choice” is foreclosure.

According to the U.S. Census, the median home value in the nation in 2000 was $119,600 and represented an 18% increase since 1990, after adjusting for inflation. In Pennsylvania, the median home value was $97,000 in 2000 – an appreciation of only 5.2% since 1990. Compared to neighboring states, Pennsylvania had the lowest home value in 2000 (see Figure IV-1: Home Values).

More recent data from the Federal Housing Finance Board indicates that the trend has continued through 2003. e median home sale price of all conventional mortgages made in 2003 in Pennsylvania was $179,900. All of Pennsylvania’s neighboring states had higher sale prices – as did the nation ($197,900).

Not surprisingly, sales have not been robust statewide. Data from the National Association of Realtors indicate that, in Pennsylvania, home sales for the year 2003 increased 10.6% since 2001. During the same time period, the sale volume across the nation increased by 15.8%; Pennsylvania had the 11th slowest growing sale volume in the nation.

Pennsylvania has one of the highest homeownership rates in the nation.

Research suggests that much of the rise in the nation’s homeownership rate is due to the expansion of credit within the mortgage markets. In particular, the expansion of the subprime mortgage market has allowed borrowers who would normally have been turned down by traditional lenders to obtain credit. ese borrowers tend to be lower-income and many are minorities. So, in one sense the expansion has resulted in a growing number of households now poised to accumulate wealth as a result of owning their own home. e negative side effect of this growth, however, is the associated rise in delinquencies and foreclosures that arise from riskier subprime products.

“Overall homeownership rates have gone from 64 percent to more than 68 percent over this period. Nearly 9 million more households own their home now than just nine years ago. A major portion of this expansion in homeownership seems clearly attributable to the increased access to credit afforded by expansions in prime and subprime mortgage lending.”6

Pennsylvania’s homeownership rate is particularly high. 71.3% of all households owned their own home in Pennsylvania in 2000; compared to 68 % nationwide.

Unemployment in Pennsylvania is lower than it is in the nation and is currently increasing at a slower rate than most other states. Nonetheless, unemployment is the most often cited cause of financial distress by homeowners at risk of losing homes to foreclosure.

Unemployment or job loss, logically, makes it more difficult for households to afford their mortgage payments and increases the likelihood of default and foreclosure. In Pennsylvania, the unemployment rate in 2003 was 5.6% - a rise of 1 percentage point since 1998. Nationally, however, the rate was 6% - a rise of 1.5% since 1998.

While Pennsylvania’s unemployment rate continues to be lower than the national average, it is rising. As a result, it is likely a contributing factor to Pennsylvania’s rising foreclosure rate. Although the unemployment rate does not explain why the rate is so high, it may at least partially explain why the foreclosure rate is growing. For example, of the applicants who were approved for HEMAP assistance in 2000, 53.3% cited “layoff-loss of income” as their cause for financial distress; by 2003, 62.8% had.

e average credit score in Pennsylvania is 696 and is one of the highest in the nation.

Credit scores generally range from 300 to 800. Lower scores generally represent higher risk consumers who carry a greater possibility of default and foreclosure. According to Experian, a repository of consumer credit information, the average credit score7 in Pennsylvania in December 2004 was 696 – the 11th highest score among the 50 states in the nation. Unlike Pennsylvania, the other 10 states with high credit scores do not have high foreclosure rates. at consumers in Pennsylvania have a higher than average credit score and a high foreclosure rate is counterintuitive. (See Figure IV-2: States with Highest Credit Scores.)

Pennsylvanians appear to be saving more than the average household nationwide.

“Personal saving, as measured by the Commerce Department’s Bureau of Economic Analysis, is essentially the amount of after-tax income left after household bills are paid. From the end of the Second World War until the early 1980s, the personal saving rate--personal saving expressed as a percentage of disposable income--gradually trended up. To be sure, the saving rate showed considerable volatility from year to year, and in some periods, such as the second half of the 1970s, its upward drift stalled for a time. But overall, the picture was that of a fairly steady rise in the personal saving rate, from about 7-1/2 percent in the early 1950s to around 10-1/2 percent in the early 1980s. Since that time, however, the household saving rate has declined precipitously and, in the last couple of years, it has averaged only about 1-1/2 percent.”8

Personal savings rate data is not available at the state level, so TRF used Current Population Survey (CPS) data to calculate an estimate of non-housing wealth for households in each state and nationwide. CPS data are collected monthly by the Census Bureau and Bureau of Labor Statistics and is available nationally for a sample of households. e March survey (called the Annual Dem-ographic Survey) contains information on the various sources of income received by each household. Among the various sources are income received from interest and dividends. TRF’s procedure to estimate “wealth” was to assume a typical savings interest rate for the period and multiply that times the amount of interest received during the year. at provided an estimate of the amount of savings available to the household.

Similarly, TRF applied the same procedure to dividends received using a typical S&P dividend yield for the appropriate time period. at too provided an estimate of the value of securities available to the household. ese are obviously imperfect estimates, but taken together provide a guide as to the relative amount of savings and non-housing investments households have available. Moreover, for purposes of comparing Pennsylvania to other states, there is no reason to suspect that the measure is more perfect (or imperfect) in any state(s). And as most other studies of household wealth suggest, these estimates too are related to household income such that higher income households tend to have greater amounts of wealth.

is data reveals that Pennsylvania households may, in fact, be saving more than their national and in some cases, regional counterparts. In 2004, for example, the data suggests that 44.1% of all households in the nation lacked any non-housing “wealth”. By comparison, only 38.8% of Pennsylvania’s lacked wealth. Similarly, only 24.8% of households nationwide have estimated wealth exceeding $50,000; in Pennsylvania, 27.6% do (see Figure IV-3: Wealth Estimates by State and Nation).

Research suggests that lower savings rates can be a factor in rising foreclosure rates as borrowers lack the financial cushion when a financial hardship occurs. Given Pennsylvania’s favorable position compared to the nation and others, this may not be a substantial factor in Pennsylvania’s rising foreclosure rate.

Mortgage terms in Pennsylvania tend to be similar to states across the nation.

Some studies suggest that those states where average mortgage rates, fees and terms are higher may have higher foreclosure rates as the loans are more expensive to the borrower. Data from the Federal Housing Finance Board, however, suggest that rates in Pennsylvania do not differ much from those around the nation. In 2003, the average mortgage interest rate nationwide was 5.67%; in Pennsylvania it was slightly higher at 5.84%. In 2003, the average term to maturity of mortgage loans was 26.8 years; it was the same for Pennsylvania. In 2003, an estimated 18% of all mortgages had adjustable rate mortgages; in Pennsylvania only 9% did.

Originations in Pennsylvania are not disproportionately FHA.

Some studies argue that part of the rise in foreclosure activity is due to the growing concentration of FHA loans in a state. ey contend that FHA loans are comprised of riskier borrowers who tend to be first-time homebuyers, lower income and have little money for down payments - characteristics which make these loans more likely to go into foreclosure.

A review of FHA origination data, indicate, however, that this is likely not a factor in Pennsylvania’s rising foreclosure activity. As will be discussed later, many of the foreclosure filings analyzed for this study are FHA, but originations in Pennsylvania are not disproportionately FHA. In 2002, for example, an analysis of HMDA data reveal that 15.3% of all conventional, purchase money mortgages in the nation were FHA loans. In Pennsylvania, only 13.2% were. By comparison, FHA purchase originations in some states are close to or over one-quarter of all originations. ese include: Indiana (23.2%), Arkansas (27.6%), Colorado (25.4%), Utah (28.8%) and Puerto Rico (32.6%).

e FHA share of refinance origination is much smaller, both nationally (3.9%) and in Pennsylvania (2.5%).

Pennsylvania continues to have one of the lowest divorce rates in the nation.

One of the most often cited triggers of foreclosure is divorce. Studies suggest that divorce, with the accompanying loss of income, increases the likelihood of foreclosure. In 2001, however, the divorce rate in Pennsylvania was 3.2%. is rate is lower than the national rate of 4%, is one of the lowest in the nation, and has declined since 1990 when the rate was 3.3%. (e rate is reported annually by the Division of Vital Statistics, National Center for Health Statistics, CDC). (See Figure IV-4: Divorce Rate by State).

In the end, Pennsylvania’s subprime foreclosure rate, compared to other states, is likely 3.2 percentage points too high.

TRF performed a multiple regression analysis as part of an exploratory analysis to see how Pennsylvania’s conventional prime and subprime foreclosure rates compare to other states. TRF found that: 1) both the prime and subprime foreclosure rates in Pennsylvania are substantially higher than one might expect statistically and 2) traditional economic indicators are more predictive of the prime foreclosure rate than of the subprime foreclosure rate. ese suggest that something else is at work with regard to the subprime rate.

Each of the following characteristics were associated with both the prime and subprime foreclosure rates for the 50 states and the District of Columbia:

1. Percent of the occupied housing units that were owner occupied in 2000;

2. Real (i.e., inflation-adjusted) appreciation in home values between 1990 and 2000;

3. Median household income;

4. Median housing value in 2000;

5. Average loan-to-price ratio in 2003;

6. Average credit score in 2004;

7. Unemployment rate in 2003;

8. Type of foreclosure process (i.e., judicial v. non-judicial);

9. Percent change in population from 1990-2000;

10. Percent of the population in 2000 that is aged 65 and older.

e multiple regression analysis determines not only if each is related, but if each characteristic is related independent of the others, and how well all of the indicators taken together explain the foreclosure rate.9

With respect to the prime foreclosure rate, the regression analysis reveals:

• States with higher rates of owner occupancy have higher prime foreclosure rates;

• States with real estate that is appreciating at higher rates tend to have higher prime foreclosure rates;

• States with lower median household incomes tend to have higher prime foreclosure rates;

• States with lower loan-to-price ratios tend to have higher prime foreclosure rates;

• States with lower average credit scores have higher prime foreclosure rates;10

• States with a judicial foreclosure process have higher prime foreclosure rates;

• States with a lower percent of the population aged 65 and older had higher prime foreclosure rates;

• e calculated impact of home values, unemployment rates and proportionate population change have little independent impact on the prime foreclosure rate;

• Taken together, the variables explain a statistically significant and substantial portion (R2 = .595) of the variation observed across states with respect to the prime foreclosure rate.

With respect to the subprime foreclosure rate, the regression analysis reveals:

• States with higher rates of owner occupancy have higher subprime foreclosure rates;

• States with appreciating home values tend to have higher subprime foreclosure rates;

• States with lower median household incomes tend to have higher subprime foreclosure rates;

• States with lower average credit scores have higher subprime foreclosure rates;

• States with a judicial foreclosure process have higher subprime foreclosure rates;

• States that have experienced higher (positive) population change have lower subprime foreclosure rates;

• e calculated impact of median home value, loan-to-price ratio, unemployment rate and percent aged 65 and over have little independent impact on the subprime foreclosure rate;

• Taken together, the variables explain a statistically significant and substantial portion (R2 = .453) of the variation observed across states with respect to the subprime foreclosure rate.

By statistically weighting each of the relevant characteristics in each state, TRF estimated what the prime and subprime foreclosure rates should be based on its unique constellation of characteristics. Both the actual prime and subprime foreclosure rates are higher in Pennsylvania than the regression-predicted rate would project. In fact, Pennsylvania’s prime foreclosure rate of .85 is approximately .19 higher than we would expect based on the above referenced indicators. Pennsylvania’s subprime foreclosure rate of 11.94 is approximately 3.89 higher than its expected value of 8.05. There is something different about the foreclosure situation in Pennsylvania above and beyond these economic factors that is producing a statistically unusual (i.e., higher) rate of home mortgage foreclosures (see Maps 13 and 14: Standardized Differences Between Observed and Predicted Rates of Foreclosure).

Additionally, the explanatory power of this basic set of characteristics is greater for the prime foreclosure rate than it is for the subprime foreclosure rate. is means that the prime foreclosure rate is more predictable using these economic factors than is the subprime foreclosure rate. is gives rise to the notion that the prime mortgage market follows a more traditional approach to borrower risk (and loss) than does the subprime market.

Finding 2: Sixty to seventy-five percent of all sampled loans in foreclosure were originated by subprime lenders. is is disproportionate to the percentage of subprime loans actually originated in Pennsylvania. In 2002, 9.9% of all loans originated in Pennsylvania were made by subprime lenders.

Subprime lending has grown dramatically both nationally and in Pennsylvania. “Subprime mortgage loan originations rose by the whopping rate of 25 percent per year over the 1994-2003 period, nearly a ten-fold increase in just nine years. Even prime mortgage lending grew by the strong annual rate of nearly 18 percent, reflecting many of the same trends” 11 (see Figure IV-5: Subprime Originations).

e market distinction between prime and subprime lending is one that has taken on enhanced importance since the early to middle 1990’s. Practically, it is reasonable to understand the distinction between these two categories of borrowers as reflecting a market estimation that subprime borrowers represent a greater loss risk than prime borrowers. Accordingly, subprime borrowers pay a higher price to borrow money and that price is generally considered to be commensurate with the enhanced risk.

In comparison to prime borrowers, borrowers with subprime loans tend to have one or more of the following traits: lower-income;12 FICO (Fair Isaac Corporation) scores below 620 – 660; high loan-to-value ratios;13 collateral property that fails to meet one or more critical appraisal standard; incomplete or unverifiable documentation of income, savings, down payment sources and/or employment; housing and other debt that exceeds 45% of monthly gross income.14 Borrowers with subprime loans are also more likely than borrowers with prime loans to have loan provisions that penalize refinancing, to end up in foreclosure and to be brought to default faster.15 Demographically, subprime borrowers are disproportionately minority, lower income, older, less well-educated, less financially sophisticated and less likely to shop interest rates.16

Data from the Home Mortgage Disclosure Act (HMDA) for 2002 indicate that 9.9% of all loans originated in Pennsylvania were made by subprime lenders. HMDA data can similarly indicate what percentage of loans originated in each of the 14-county study area were prime and subprime (see Figure IV-6: Originations by County, Prime vs. Subprime).

By tracking back to the originating lenders for loans on the foreclosure filing list and identifying those lenders as either prime, subprime or both, TRF was able to compare the types of loans originated to the types of loans in foreclosure in each county.

Given the varying nature of data available in each county, TRF was able to track back to the originating lender for almost all of the loans in foreclosure between 2000 and 2003 for Monroe, Montgomery and Philadelphia counties; and for a sample of loans in Allegheny, Berks, Bucks, Delaware, Lancaster, Lehigh, and Dauphin. For the remaining counties, TRF was unable to use the data provided from the Prothonotary offices to track back to originating lender.

For the nine counties with originating lender data, a comparison of the types of lenders originating by county and the types of lenders foreclosing reveals:

• Without exception, every county had a significantly higher percentage of subprime loans in foreclosures than subprime originations.

• Berks, Allegheny, Lehigh and Bucks had the highest shares of subprime foreclosures, each with more than 70%.

• Monroe, Philadelphia, Allegheny and Berks counties, however, had the highest rates of subprime originations.

Methodology for Identifying Originating Lender and Characterizing its “Type”: e method used by TRF to identify and characterize the originating lenders of loans in foreclosure is a two-step process:

1) Depending on the data provided by the Prothonotary in each of the 14-counties studied, TRF searched the RealQuest database for the owner’s name, address or tax-id of the house in foreclosure. If found in RealQuest, transaction information about that property was downloaded and analyzed. If enough transaction information was available on a property, TRF was able to identify when the loan was made, for what amount, and who the originating lender was. TRF did this for all filings in Philadelphia and Montgomery counties and for a statistically significant sample in the remaining study area counties.

2) TRF then characterized the originating lender as either one that predominantly does prime or subprime business or that does a mix of both. Ideally, TRF would prefer to make absolute determinations about whether a particular loan, not lender, was either prime or subprime. Given the inability to view loan documentation for every loan in foreclosure, TRF employed a commonly used methodology that characterizes the lender as one that generally makes prime or subprime loans.

e U.S. Department of Housing and Urban Development (HUD) publishes, annually, a list of lenders it identifies as those that specialize in subprime lending (see: www.huduser.org/datasets/manu.html). TRF recognizes and acknowledges the potential flaws in characterizing lenders rather than loans. Most notably, there are lenders that have a full array of loan products and on the list prepared by HUD end up being characterized as either prime or subprime. Moreover, all the loans originated by that lender regardless of whether they are prime or subprime, get characterized in whatever way the lender has been characterized. To ameliorate this, TRF modified the list published by HUD to identify some lenders, that do both prime and subprime business, as “both”. TRF also modified the list to include lenders that were involved in loans in foreclosure but were not on the HUD list. at said, the methodology employed here is a standard and widely accepted approach utilized and reported by the Federal Reserve, Harvard University’s Joint Center for Housing Studies, and highly respected scholars at prestigious universities publishing in professionally refereed journals and economic research commissioned by the National Home Equity Mortgage Association.17

PHILADELPHIA AND MONTGOMERY COUNTIES – A CLOSER LOOK

Prothonotary offices in Montgomery and Philadelphia counties document substantial information about foreclosure filings in their jurisdictions. TRF was able to obtain this information electronically and, as a result, was able to track back to many more foreclosure filings. In Philadelphia County, 15,592 filings (of 23,742) were connected back to original loan information, including lender names, loan amounts, loan types and origination date. In Montgomery County, 4,240 foreclosure filings (of 4,950) were connected to original loan information. A closer look at the types of lenders and their involvement in the foreclosure filing list reveals:

Almost 40% of subprime loans originated in Philadelphia in 1998 were in foreclosure at some point between 2000 and 2003. In Montgomery County, 20% were.

TRF was able to identify when most of the loans in foreclosure between 2000 and 2003 were originated and by whom. TRF was also able to obtain the number of loans originated by type of lender in each county from HMDA starting in 1998. Comparing the types of loans originated in 1998 with the types of loans in foreclosure that were made in 1998 reveals that almost 40% of all subprime loans originated in Philadelphia in 1998 were in foreclosure during the study period of 2000 and 2003. In Montgomery County, 20% were. Similarly, in 1999, 40.9% of all subprime loans originated in Philadelphia were in foreclosure during the study period; 19.7% in Montgomery County were. (These numbers are a statistical estimate of the lending health in a county and exclude FHA and VA originations.) In contrast, 2.8% of all prime loans originated in Philadelphia in 1998 were in foreclosure during the study period; 2.58% in 1999. In Montgomery County, .3% of 1998 prime originations were in foreclosure; in 1999 and .38% were.18

Subprime loans originated post 1994 appear to be driving the rising numbers of foreclosure filings.

The two charts below track 2000-2003 foreclosures back to the year of their origination, and classify the loans by the “type” of lender: prime, subprime or lenders that do both types of lending. As the charts demonstrate, loans in foreclosure between 2000 and 2003 that were made before 1994 were originated by an even mix of prime and subprime lenders. By 1994/1995, the number of originated subprime loans in foreclosure began to outnumber prime loans in foreclosure. Since that time, subprime foreclosures have outnumbered prime foreclosures in both counties (see Figure IV-7: Philadelphia Foreclosure Filings and Figure IV-8: Montgomery County Foreclosure Filings).

Refinancing patterns in Philadelphia vary by the type of loan being refinanced and the value of the home.

One of the benefits often told to borrowers of a subprime loan is that if their credit is flawed, they can get a subprime loan, make payments, and then refinance that loan into a less expensive prime loan. Based on a random sample of properties in the City of Philadelphia, it appears that this is somewhat true. Among those properties that started out with a prime loan and had that loan refinanced, 66.6% refinanced into a prime loan; 27.9% refinanced into a subprime loan; 5.6% refinanced with a lender that could not readily be identifiable as either prime or subprime (i.e., these lenders did both sorts of lending). Among those properties that started with a subprime loan and refinanced, 29.0% ended up with a prime lenders, 66.7% ended up with a subprime lender and 4.3% ended up with a lender that did both sorts of lending.

The refinance pattern is very different depending upon the value of homes in the surrounding area. For example, the likelihood of a prime to subprime refinance in a lower price area (i.e., 34.8%) is substantially higher than in higher price range areas (12.5%). The refinance pattern reflective of “credit repair” (i.e., subprime to prime loan refinances) is far more prevalent in higher value areas (42.9%) than in lower value areas (20.7%) (see Figure IV-9: Refinance Patterns by Census Tract and Figure IV-10 Refinance Patterns by Area).

Finding 3: Growing foreclosure filings do not appear to be simply the result of an expanding mortgage market, as filings are outpacing any gains in homeownership.

Some have argued that the growth in foreclosure filings is outweighed by the growth in the number of people who now own their own home. In an effort to discern whether the rise in foreclosure filings is disproportionate to this increased homeownership activity, TRF used data from the U.S. Census Bureau to standardize the number of foreclosure filings against the number of housing units in each of the 14 counties. To that end, the U.S. Census Bureau reports, on a county-by-county basis, the estimated number of housing units for 2003 (the most recent year for which data are available); these same data are available for the year 2000 based upon tabulations from the decennial Census. Using the county percentage of owner occupied properties in the year 2000 (the most recent year for which this particular piece of information is available), TRF then estimated the size of the owner occupied housing stock in each county in 2003.

For the years 2000 through 2003, TRF collected the number of foreclosure filings for each of the 14 counties studied. The formula used to standardize the foreclosure filings in each county is:

# foreclosures in year y

___________________________* 100

# of estimated owner occupied housing units in year y

These data show:

• In all counties sampled, the number of foreclosure filings rose faster than the estimated number of owner-occupied housing units (homeowners);

• The rate is most extreme in Monroe County – where a separate investigation, in fact, took place earlier this year;

• After Monroe, counties with the highest rates in 2003 include Philadelphia, Dauphin and Allegheny; and

• Erie County experienced the greatest increase in its rate between 2000 and 2003.

(See Figure IV-11: Estimated Foreclosure Filings per 1,000 Owner Occupied Housing Units and Figure IV-12: Estimated Rate of Foreclosure Filings per 1,000 Owner Occupied Units.)

To be clear, these are not foreclosure rates like those reported by the MBAA and are, therefore, not directly comparable to the MBAA figures reported previously in this report. The rates reported by the MBAA require data on the totality of existing mortgages – data that is not available to TRF.

Finding 4: Foreclosures are typically concentrated in areas with lower than average housing values, lower than average family incomes, higher than average Black or African American households and higher than average percentages of Hispanic households.

Data on each foreclosure filing for which TRF was able to obtain data (representing the universe of filings in some instances and samples in others) were geocoded with information related to the Census Block Group in which the property in foreclosure was located. The foreclosures were then aggregated (or combined) so that for each Block Group a total number of foreclosures was identified. In each county there were varying numbers of Block Groups identified by the Census (e.g., as few as 71 in Monroe County and as many as 1,816 in Philadelphia); many of the Block Groups had no foreclosures, some had many.

From the Census Bureau’s Summary File 3 (2000), TRF extracted a few fields of information related to each Block Group’s:

• median home value in 1999;

• median family income in 1999;

• percent of households headed by a Black or African American person;

• percent of households headed by a person of Hispanic origin.

The foreclosure filing information was then combined with the Census information. This approach allowed TRF to determine whether places with no (or few) foreclosures look different with respect to this Census information than places where foreclosures were more densely clustered.

Overall, the pattern of differences across counties is fairly consistent – areas with more highly clustered foreclosures tend to be areas with lower than average housing values, lower than average family incomes, higher than average percentages Black or African American and higher than average percentages Hispanic. Although the magnitude of differences changes quite a bit based on the widely differing demographics of the counties identified, the pattern is consistent. The exception to this pattern is Monroe County which, as noted earlier, tended to be experiencing a somewhat different set of foreclosure dynamics.

The samples sizes in Philadelphia and Montgomery Counties were large enough for TRF to geographically distinguish whether the loan that went into foreclosure was likely of a prime or subprime character. This is because much more complete information was available for each filing. Looking at all foreclosures without differentiating prime and subprime, the general pattern observed for the other counties held (i.e., that clustered foreclosures occurred in areas that were of lower home values and family incomes than average and higher percentages Black or African American and Hispanic). However, the pattern for where there were clustered foreclosures that were prime versus subprime was quite different.

In Montgomery County, there are no Block Groups with 11 or more prime foreclosures. Additionally, where the concentration of foreclosures is greatest, the subprime clusters tend to occur in areas with even lower median home values and median family income, higher percentages Black or African American and higher percentages Hispanic than areas with the more densely clustered prime foreclosure filings.

In Philadelphia County, that pattern is partially reversed. Areas with clusters of subprime foreclosures in comparison to areas with clusters of prime foreclosure tend to be areas with higher average median family incomes and home values. Conversely, areas with clusters of subprime foreclosures in comparison to areas with clusters of prime foreclosures are higher in average percent Black or African American but lower in percent Hispanic (see Figure IV-13A&B: Characteristics of Block Groups by Number of Foreclosures).

Finding 5: Loans in foreclosure are an even mix of purchase money mortgages and refinances. e typical homeowner in foreclosure between 2000 and 2003 in Pennsylvania is not a long-term homeowner. ey tend to have purchased their home in the mid-to-late 1990s and took out the loan currently in foreclosure just a few years earlier.

As mentioned earlier, by linking Prothonotary lists to information contained in the First American Real Estate Solutions, Inc. database, TRF was able to collect, analyze and map a significant amount of information regarding all of the loans in foreclosure between 2000 and 2003 for Monroe, Montgomery and Philadelphia counties; and for a sample of loans in Allegheny, Berks, Bucks, Delaware, Erie, Lancaster, Lehigh, and Northampton. In Dauphin County, only a sample of loans in foreclosure from 2003 was available for analysis. No detailed information was available for Chester and Washington counties.

In all, information for approximately 22,979 loans on the foreclosure filing lists was analyzed. e purpose of that analysis was to answer the following questions:

• How have foreclosure filings increased in each county?

• What types of lenders originated the loans in foreclosure?

• Were the loans in foreclosure mortgages that were used to purchase a home or were they refinance/home equity loans?

• How long ago did the borrowers in foreclosure purchase their home? Are borrowers on the foreclosure filing list long-term homeowners or did they recently buy a home?

• For those loans in foreclosure that were purchase money mortgages, how long ago did the homeowner buy their home and for how much?

• How long ago did the borrower take out the loan that is in foreclosure?

Where the data was available, TRF was able to answer each of these questions on a county-by-county basis. e most striking observation about the county-by-county analysis that follows is how similar the characteristics of the loans in foreclosure are across counties.

ALLEGHENY COUNTY

Foreclosure Filing Numbers: Between 2000 and 2003, 13,887 mortgage foreclosures were filed in Allegheny County. is represents an increase of 60.3% during the four year period – one of the highest increases among counties studied for this report. In 2000, 2,567 foreclosures were filed; in 2003, 4,115 were. e “rate” per 1,000 owner occupied housing units also rose, from 7.13 per 1,000 owner occupied housing units in 2000 to 11.42 per 1,000 owner occupied housing units in 2003.

Types of Lenders: Foreclosure filings between 2000 and 2003 are disproportionately subprime. In 2002, 12% of all conventional loans originated in Allegheny County were considered subprime. Yet, 71% of the loans in foreclosure sampled in this study were subprime

Length of Ownership: e typical household in foreclosure bought their home in 1995. Of all households in foreclosure between 2000 and 2003, 68% bought their home after 1991.

Time from Origination to Filing: e typical household took out the loan in foreclosure 2.8 years prior to the foreclosure filing.19

Purchase Money Mortgage vs. Refinance Loans:

e majority of loans in foreclosure in Allegheny County (60%) were refinance loans. Of the 40% that were purchase money mortgages, the typical homebuyer bought their home in 1999 and paid $54,000.

BERKS COUNTY

Foreclosure Filing Numbers: Between 2000 and 2003, 3,924 mortgage foreclosures were filed in Berks County. is represents an increase of 26.4% during the four year period. In 2000, 855 foreclosures were filed; in 2003, 1,081 were. e “rate” per 1,000 owner occupied housing units also rose, from 8.2 per 1,000 owner occupied housing units in 2000 to 10 per 1,000 owner occupied housing units in 2003.

Types of Lenders: Foreclosure filings between 2000 and 2003 are disproportionately subprime. In 2002, 12% of all conventional loans originated in Berks County were considered subprime. Yet, 78% of the loans in foreclosure sampled in this county were originated by subprime lenders. Of all counties studied, Berks County has the highest share of loans in foreclosure originated by subprime lenders.

Length of Ownership: e typical household in foreclosure bought their home in 1997. Of all households in foreclosure between 2000 and 2003, 89% bought their home after 1991.

Time from Origination to Filing: First American Real Estate Solutions, Inc. only started recording mortgage transactions in Berks County in 1999. As a result, TRF cannot make this calculation.

Purchase Money Mortgage vs. Refinance Loans: e majority of loans in foreclosure in Berks County (62%) were purchase loans. Of these, the typical homebuyer bought their home in 2000 and paid $58,000.

BUCKS COUNTY

Foreclosure Filing Numbers: Between 2000 and 2003, 4,018 mortgage foreclosures were filed in Bucks County. is represents an increase of 13% during the four year period – the lowest increase during this four year period among counties studied for this report. In 2000, 886 foreclosures were filed; in 2003, 1,002 were. e “rate” per 1,000 owner occupied housing units also rose, from 5.2 per 1,000 owner occupied housing units in 2000 to 5.7 per 1,000 owner occupied housing units in 2003. is is the second lowest rate among all counties studied.

Types of Lenders: Foreclosure filings between 2000 and 2003 are disproportionately subprime. In 2002, 9% of all conventional loans originated in Bucks County were considered subprime. Yet, 70% of the loans in foreclosure sampled in the county were originated by subprime lenders.

Length of Ownership: e typical household in foreclosure bought their home in 1995. Of all households in foreclosure between 2000 and 2003, 66% bought their home after 1991.

Time from Origination to Filing: e typical household took out the loan in foreclosure 3.3 years prior to the foreclosure filing.

Purchase Money Mortgage vs. Refinance Loans: e majority of loans in foreclosure in Bucks County (73%) were purchase loans – the highest percentage of purchase loans among counties studied. Of theses purchase loans, the typical homebuyer bought their home in 1998 and paid $112,000.

CHESTER COUNTY

Foreclosure Filing Numbers: Between 2000 and 2003, 2,481 mortgage foreclosures were filed in Chester County. In 2000, 515 foreclosures were filed; in 2003, 668 were. e “rate” per 1,000 owner occupied housing units also rose, from 4.3 per 1,000 owner occupied housing units in 2000 to 5.2 per 1,000 owner occupied housing units in 2003.

Data was not available to conduct additional analysis on the nature of foreclosure filings in Chester County.

DAUPHIN COUNTY

Foreclosure Filing Numbers: Between 2002 and 2003, 1,569 mortgage foreclosures were filed in Dauphin County. In 2002, 776 foreclosures were filed; in 2003, 793 were filed. e “rate” per 1,000 owner occupied housing was 11.6 in 2003 – relatively the same as in 2002.

Types of Lenders: Foreclosure filings between 2000 and 2003 are disproportionately subprime. In 2002, 11 % of all conventional loans originated in Dauphin County were considered subprime. Yet, 64% of the loans in foreclosure sampled in the county were originated by subprime lenders.

Time from Origination to Filing: Of the loans sample in 2003, the typical household in foreclosure took out the loan in foreclosure 3 years prior to the foreclosure filing.

DELAWARE COUNTY

Foreclosure Filing Numbers: Between 2000 and 2003, 6,022 mortgage foreclosures were filed in Delaware County. is represents an increase of 36.4% during the four year period. In 2000, 1,246 foreclosures were filed; in 2003, 1,700 were. e “rate” per 1,000 owner occupied housing units also rose, from 8.4 per 1,000 owner occupied housing units in 2000 to 11.3 per 1,000 owner occupied housing units in 2003.

Types of Lenders: Foreclosure filings between 2000 and 2003 are disproportionately subprime. In 2002, 10% of all conventional loans originated in Delaware County were considered subprime. Yet, 64% of the loans in foreclosure sampled in the county

were originated by subprime lenders.

Length of Ownership: e typical household in foreclosure bought their home in 1995. Of all households in foreclosure between 2000 and 2003, 71% bought their home after 1991.

Time from Origination to Filing: e typical household took out the loan in foreclosure 3.8 years prior to the foreclosure filing.

Purchase Money Mortgage vs. Refinance Loans: e majority of loans in foreclosure in Delaware County (60%) were purchase loans. Of these purchase loans, the typical homebuyer bought their home in 1998 and paid $69,950.

ERIE COUNTY

Foreclosure Filing Numbers: Between 2000 and 2003, 2,218 mortgage foreclosures were filed in Erie County. is represents an increase of 115% during the four year period. In 2000, 361 foreclosures were filed; in 2003, 778 were filed. Erie experienced the greatest relative increase in filings of all counties studied. e “rate” per 1,000 owner occupied housing units also rose, from 4.9 per 1,000 owner occupied housing units in 2000 to 9.1 per 1,000 owner occupied housing units in 2003.

Types of Lenders: Foreclosure filings between 2000 and 2003 are disproportionately subprime. In 2002, 9% of all conventional loans originated in Erie County were considered subprime. Yet, 64% of the loans in foreclosure sampled in the county were originated by subprime lenders.

Length of Ownership: e typical household in foreclosure bought their home in 1995. Of all households in foreclosure between 2000 and 2003, 75% bought their home after 1991.

Time from Origination to Filing: e typical household took out the loan in foreclosure 3 years prior to the foreclosure filing.

Purchase Money Mortgage vs. Refinance Loans: e majority of loans in foreclosure in Erie County (52%) were refinance loans.

LANCASTER COUNTY

Foreclosure Filing Numbers: Between 2000 and 2003, 3,437 mortgage foreclosures were filed in Lancaster County. is represents an increase of 30.5% during the four year period. In 2000, 734 foreclosures were filed; in 2003, 958 were filed. e “rate” per 1,000 owner occupied housing units also rose, from 6 per 1,000 owner occupied housing units in 2000 to 7.6 per 1,000 owner occupied housing units in 2003.

Types of Lenders: Foreclosure filings between 2000 and 2003 are disproportionately subprime. In 2002, 7 % of all conventional loans originated in Lancaster County were considered subprime. Yet, 64% of the loans in foreclosure sampled in the county were originated by subprime lenders.

Length of Ownership: e typical household in foreclosure bought their home in 1997. Of all households in foreclosure between 2000 and 2003, 89% bought their home after 1991.

Time from Origination to Filing: e typical household took out the loan in foreclosure 2.5 years prior to the foreclosure filing.

Purchase Money Mortgage vs. Refinance Loans: e majority of loans in foreclosure in Lancaster County (66%) were purchase loans. Of these purchase loans, the typical homebuyer bought their home in 1999 and paid $81,000.

LEHIGH COUNTY

Foreclosure Filing Numbers: Between 2000 and 2003, 3,219 mortgage foreclosures were filed in Lehigh County. is represents an increase of 17.5% during the four year period. In 2000, 702 foreclosures were filed; in 2003, 825 were filed. e “rate” per 1,000 owner occupied housing units also rose, from 8.4 per 1,000 owner occupied housing units in 2000 to 9.5 per 1,000 owner occupied housing units in 2003.

Types of Lenders: Foreclosure filings between 2000 and 2003 are disproportionately subprime. In 2002, 8% of all conventional loans originated in Lehigh County were considered subprime. Yet, 71% of the loans in foreclosure sampled in the county were originated by subprime lenders.

Length of Ownership: e typical household in foreclosure bought their home in 1997. Of all households in foreclosure between 2000 and 2003, 74% bought their home after 1991.

Time from Origination to Filing: e typical household took out the loan in foreclosure 2.8 years prior to the foreclosure filing.

Purchase Money Mortgage vs. Refinance Loans: e majority of loans in foreclosure in Lehigh County (68%) were purchase loans. Of these purchase loans, the typical homebuyer bought their home in 1999 and paid $66,000.

MONROE COUNTY

In August of 2004, TRF conducted an in-depth study of rising foreclosure in Monroe County for the Pennsylvania Department of Banking. at study concluded:

e growth in foreclosure filings is real, outpaces housing unit growth in the County and is disproportionate to other counties in the Commonwealth.

Monroe County’s rapid growth was fueled by the migration of families from New York and New Jersey.

Loans in foreclosure in Monroe County from 2000 through 2003 have different characteristics than the typical loan in Monroe County.

Foreclosure filings are concentrated in five townships and twelve subdivisions.

e full study can be viewed from the Pennsylvania Department of Banking’s web site at www.banking.state.pa.us.

Foreclosure Filing Numbers: Between 2000 and 2003, 3,443 mortgage foreclosures were filed in Monroe County. is represents an increase of 34.5% during the four year period. In 2000, 699 foreclosures were filed; in 2003, 940 were filed. e “rate” per 1,000 owner occupied housing units also rose, from 18 per 1,000 owner occupied housing units in 2000 to 22.7 per 1,000 owner occupied housing units in 2003. Monroe County has the highest “rate” of foreclosure of any county studied for this report.

Types of Lenders: Foreclosure filings between 2000 and 2003 are disproportionately subprime. In 2002, 16% of all conventional loans originated in Monroe County were considered subprime. Yet, 67% of the loans in foreclosure sampled in the county were originated by subprime lenders.

Monroe County has the highest level of subprime originations of any county studied for this report.

Time from Origination to Filing: e typical household took out the loan in foreclosure 2.8 years prior to the foreclosure filing.

Purchase Money Mortgage vs. Refinance Loans: e majority of loans in foreclosure in Monroe County (64%) were purchase loans.

MONTGOMERY COUNTY

Foreclosure Filing Numbers: Between 2000 and 2003, 4,950 mortgage foreclosures were filed in Montgomery County. is represents an increase of 14.6% during the four year period. In 2000, 1,142 foreclosures were filed; in 2003, 1,309 were. e “rate” per 1,000 owner occupied housing units also rose, from 5.4 per 1,000 owner occupied housing units in 2000 to 6.0 per 1,000 owner occupied housing units in 2003 – the third lowest rate among counties studied.

Types of Lenders: Foreclosure filings between 2000 and 2003 are disproportionately subprime. In 2002, 8% of all conventional loans originated in Montgomery County were considered subprime. Yet, 61% of the loans in foreclosure sampled in the county were originated by subprime lenders. Montgomery County has the lowest percent of subprime foreclosures among all counties studied.

Length of Ownership: e typical household in foreclosure bought their home in 1995. Of all households in foreclosure between 2000 and 2003, 67% bought their home after 1991.

Time from Origination to Filing: e typical household took out the loan in foreclosure 4 years prior to the foreclosure filing.

Purchase Money Mortgage vs. Refinance Loans: e majority of loans in foreclosure in Montgomery County (60%) were purchase loans. Of these purchase loans, the typical homebuyer bought their home in 1997.

NORTHAMPTON COUNTY

Foreclosure Filing Numbers: Between 2000 and 2003, 2,109 mortgage foreclosures were filed in Northampton County. is represents an increase of 32.2% during the four year period. In 2000, 438 foreclosures were filed; in 2003, 579 were filed. e “rate” per 1,000 owner occupied housing units also rose, from 5.86 per 1,000 owner occupied housing units in 2000 to 7.47 per 1,000 owner occupied housing units in 2003.

Types of Lenders: Foreclosure filings between 2000 and 2003 are disproportionately subprime. In 2002, 8% of all conventional loans originated in Northampton County were considered subprime. Yet, 67% of the loans in foreclosure sampled in the county were originated by subprime lenders.

Length of Ownership: Data is not sufficient to make this calculation.

Time from Origination to Filing: Data is not sufficient to make this calculation.

Purchase Money Mortgage vs. Refinance Loans: e majority of loans in foreclosure in Northampton County (66%) were purchase loans. Of these purchase loans, the typical homebuyer bought their home in 2000 and paid $84,590.

PHILADELPHIA COUNTY

Foreclosure Filing Numbers: Between 2000 and 2003, 23,742 mortgage foreclosures were filed in Philadelphia County. is represents an increase of 23.1% during the four year period. In 2000, 5,112 foreclosures were filed; in 2003, 6,292 were. e “rate” per 1,000 owner occupied housing units also rose, from 14.6 per 1,000 owner occupied housing units in 2000 to 18.1 per 1,000 owner occupied housing units in 2003 – the second highest rate among counties studied.

Types of Lenders: Foreclosure filings between 2000 and 2003 are disproportionately subprime. In 2002, 16% of all conventional loans originated in Philadelphia County were considered subprime. Yet, 72% of the loans in foreclosure sampled in the county were originated by subprime lenders. Philadelphia has the second highest percent of subprime originations, and second highest foreclosure rate among all counties studied.

Length of Ownership: e typical household in foreclosure bought their home in 1995. Of all households in foreclosure between 2000 and 2003, 66% bought their home after 1991.

Time from Origination to Filing: e typical household took out the loan in foreclosure 4 years prior to the foreclosure filing.

Purchase Money Mortgage vs. Refinance Loans: e majority of loans in foreclosure in Philadelphia County (60%) were purchase loans. Of these purchase loans, the typical homebuyer bought their home in 1997 and paid $45,900.

WASHINGTON COUNTY

Foreclosure Filing Numbers: Between 2000 and 2003, 1,933 mortgage foreclosures were filed in Washington County. In 2000, 353 foreclosures were filed; in 2003, 587 were. e “rate” per 1,000 owner occupied housing units also rose, from 5.63 per 1,000 owner occupied housing units in 2000 to 9.12 per 1,000 owner occupied housing units in 2003.

Data was not available to conduct additional analysis on the nature of foreclosure filings in Washington County.

NATIONAL FACTORS

Cause 1: Securitization of the residential mortgage market – particularly the subprime market.

Securitization is generally understood to be “…the transformation of mortgage loans into securities that are issued and traded in the capital markets.” 20

e prime mortgage market has long been an example of a market that performs quite well in a traditional economic sense. ere is much competition amongst lenders, giving borrowers choices, and keeping costs and fees to a minimum. Default rates are generally low, and millions of Americans regularly participate in the market. Fannie Mae and Freddie Mac provide standardization and liquidity to the market making it work very efficiently. Economic rents, or profits, are held to a minimum because borrowers have the ability to refinance if they see rates advertised at lower rates than they currently pay. e prospect of borrower pre-payment, which is generally accepted to be extremely undesirable for lenders, provides lenders with a strong incentive to provide credit at as low a rate as financially feasible. Compared to the prime market, in which the nation’s GSEs are major purchasers of mortgages, subprime loans are not purchased as widely by Fannie Mae or Freddie Mac. e volume of these subprime securitizations rose from $202.56 million in all of 2003 to $291.6 million for the first nine months of 2004 (see Table V-1).

Interviews with subject-matter experts suggest that instead of thinking of the subprime market as a traditional supply of and demand for residential credit, the subprime market must instead be looked at in terms of the creation of a demand for packaged, securitized debt. In this market scenario, the true customers are not the homeowners, but the mutual funds and investors who purchase the securities. If this market scenario is used, a relatively high default rate is expected, but it can be priced as the loans are packaged and sold. is type of market, where there is a mortgage investor seeking a borrower, rather than a borrower seeking a mortgage, has produced a high rate of default and is thus likely a contributing cause of the rising foreclosure rate.

Cause 2: Increased consumer access to mortgage products which allow for lower down payments, lower savings balances, higher loan-to-value ratios, and lower credit scores to buy a home may make long-term homeownership unsustainable.

e very programs that the mortgage industry and government created to help lower-income people buy homes may be the very programs that increase the likelihood that they will lose their homes to foreclosure. Whether the cause is an unaffordable mortgage or a steep rise in property taxes, the problem is the same. People may be getting into homeownership when they are too close to the financial margin so that any extraneous shock can cause a loss. is is compounded by the belief among many of those interviewed that cost-burdened homeowners with any equity in their home are using their home like a credit card. In essence, they are refinancing or taking out home equity loans to pay for everyday expenses that their incomes cannot support.

Lower Down Payments and Higher Loan-to-Value Ratios: In the early to mid-1990s, the mortgage industry evolved from a Rules Based System to a Risk Based System. Loan-to-value ratios (LTVs) needed to be much lower and down payments much higher under the Rules Based system. e switch to the Risk Based System allows LTVs to be as high as 95% - sometimes even higher. Interviewees suggest that when people put very little down, as is reflected in a higher loan-to-value ratio, they do so because they have very little saved and, as a result, are less able to afford their monthly mortgage payment when financial problems or crises occur. Since households have less invested in the property they may be more easily inclined to simply walk away when the payments become unmanageable.

Qualification for Mortgages that Allow More Relaxed Credit Scores: Interviewees suggest that as it stands now, there is rarely a credit score that can’t find a home mortgage. e fees and interest rate may be quite high, but the loan product is out there. To a financially unsavvy household, with little income and savings and a poor credit score, holding off the purchase of a home to some point in the future may be a wiser decision than deciding to purchase now with the costly loan. But, for households determined to buy a house, they can immediately – and perhaps too easily.

In its analysis of the 2003/2004 FHA loan portfolio for HUD’s Office of the Inspector General, KPMG made observations about defaults among FHA insured mortgages. Although focused on FHA lending, their observations bear relevance to the conventional market as well. ey state:

“We recognize that economic factors such as home purchase price appreciation and increased unemployment rates have an impact on the default rate; however, changes in underwriting policies may also contribute to the higher default rates.” e sort of underwriting changes that KPMG and HUD, in its response, highlighted were designed to make credit more readily available to persons who were lower income and/or had credit issues.”21

Cause 3: A lack of financial education among borrowers ranging from understanding the economics of interest rates to the importance of paying bills on time.

As detailed earlier, many of the borrowers facing foreclosure purchased their home within the last 10 years and may likely have been first-time homebuyers. Interviewees suggest that they were likely ill-equipped to handle the mortgage process. is includes:

Limited Experience with Credit: Interviewees cited the lack of a complete and correct understanding among many about credit. Interviewees suggest that people tend not to understand the extent to which a low credit score can impact every aspect of their financial health (not only future borrowing costs, but insurance premiums and job eligibility).

Lack of Financial Literacy in School: e school curriculum in Pennsylvania does not provide financial literacy education to replace what young people lack in terms of understanding regarding savings, credit, income, and household budgeting. In the end, the six hours of homeownership counseling provided to many of these newer buyers - regardless of how good those classes may be - are not enough to counteract a lifetime of poor financial habits.

Lack of Resources for Consumers in Foreclosure: Consumers in foreclosure do not always have the benefit of a lawyer, and housing counselors are not uniformly effective. In 2004, HUD awarded $36.1 million in housing counseling grants. Of that, approximately $4.5 million went to Pennsylvania. In 2003, HUD awarded $4.4 million statewide. Beyond these amounts, the PHFA funded housing counseling activities and many of the municipalities that receive Community Development Block Grant (CDBG) dedicated a portion to housing counseling. HUD also awarded $7.75 million to the Neighborhood Reinvestment Corporation in 2004 to both standardize housing counseling services and train HUD-approved counselors. Given these resources, the efficacy of counseling in foreclosure prevention is uneven, but seems to have the best possibility of working where the method of delivery is most personalized (and most expensive).

Cause 4: A lack of information among borrowers and communities about alternatives to high-cost loans.

Fannie Mae’s 2002 National Housing Survey makes it quite clear: a substantial portion of the adult population does not have a complete and accurate picture of the mortgage and home-buying process. e following table, reproduced in part from that National Housing Survey, shows that there are some very critical gaps in information not only generalized across the population, but acutely so among minority adults.

e lack of knowledge is not evenly distributed across mortgage products. Evidence suggests that those borrowers who end up with some of the more expensive (and oftentimes complex loan products) are least able to understand and appreciate them. Lax, et al., for example, identify certain very critical differences between prime and subprime borrowers with respect to knowledge, experience and search behavior for their mortgages.22 In general, subprime borrowers were less knowledgeable, less well-prepared and did substantially less searching for the best loans available. In fact, subprime borrowers were more likely to respond to telephone calls or other advertisements.

Exacerbating the problem of knowledge, experience and behavior is the transparency of the various markets. Unlike the prime mortgage market where products and procedures are fairly standard, the subprime market operates in a less clear manner. White states: “Mortgage interest rates and points in the prime market are widely advertised…By contrast, current subprime mortgage rates 23at the retail level are secret. No newspaper lists them in the real estate section. e rate tables used by wholesale subprime lenders are available only to brokers and are sometimes regarded as trade secrets.”

TRF’s own interviews with industry representatives, housing counselors, brokers and attorneys suggest that it is not only a lack of information about alternatives to the higher cost loans that people often end up with, but misinformation can often lead people into transactions which they later regret. Moreover, where reasonable alternatives (i.e., low interest home improvement loans that exist in a number of municipalities across the Commonwealth) exist to the higher cost loans that people get, they tend to be less aggressively marketed. ese lower cost alternatives are generally made available; the higher cost products are targeted and push marketed. In marketing their products, lenders and brokers are making extensive use of information technology which allows them to identify borrowers before they even know they want to be borrowers. Interviews suggest that brokers will oftentimes search public record information for people who already (likely) have high-cost loan products (e.g., have liens against their home by finance companies) because they have a demonstrated history of borrowing high-cost funds; others were purchasing lists of people who manifest certain demographic, neighborhood, and/or consumer behavior (e.g., use credit cards). is ready and inexpensive (pennies a prospect) access to potential borrowers makes the less knowledgeable, less experienced and less aggressive shoppers more likely to end up with less advantageous loan products. And it is these less advantageous products that are associated with greater rates of failure (i.e., foreclosure).

PENNSYLVANIA FACTORS

Cause 5: Regulations in Pennsylvania are not protecting homeowners as originally intended.

Act 6 - In 1974, when Act 6 was enacted, virtually all homeowners benefited from its protections. Today, the failure to index the original $50,000 threshold has limited the benefits of Act 6 to a minority of homeowners in most counties studied.

At its most basic, the consumer protections of Act 6, enacted in 1974, were designed to help keep Pennsylvania residents in their homes whenever possible. e law has a number of “protective provisions” that provide protections for those borrowers with a “residential mortgage” in default. One provision of the Act states that the borrower must be given 30 days notice prior to a foreclosure action. Another provision states that the borrower may cure his default up to “one hour prior to the commencement of bidding at a sheriff sale,” and that curing this default “restores the residential mortgage debtor to the same position as if the 24default had not occurred.” A third provision prohibits the use of prepayment penalties on Act 6 qualified mortgages. A fourth provision limits attorneys’ fees to a “reasonable” level. Additional provisions combine with those above in an attempt to create a legal environment that ensures Commonwealth homeowners have a reasonable chance to stay in their home and recover from a mortgage delinquency.

Act 6 defines a “residential mortgage” as a loan that is not more than $50,000 in size. In 1980, close to the period when Act 6 was enacted, 66.3% of homes in the Commonwealth fell under the $50,000 threshold. By the year 2000, 17.8% of homes were valued at or below $50,000. As a result, this provision renders Act 6 protections to a minority of homes and loans in foreclosure in all but Philadelphia County. If the $50,000 threshold had risen with inflation, Act 6 would currently protect all those mortgages below $197,000. As the chart demonstrates, adjusting the original Act 6 threshold would protect over 90 percent of homeowners in foreclosure in each county studied (see Figure V-1: Act 6 Chart, and Figure V-2: Act 6 Table).

Interviews with attorneys suggested that the Act 6 provisions limiting legal fees can be an effective means to keep costs down for borrowers going into foreclosure. In order to determine whether this observation was born out statistically, TRF compared the difference in fees for Act 6 and non-Act 6 prime mortgage foreclosures within Philadelphia and found that the difference was significant and real.25

In terms of a ratio between fees and the principal of the loan still due, TRF found that Act 6 protected loans in fact had a somewhat larger proportion of fees to loan size. Because each foreclosure filing has a set of basic, threshold costs attached to it, it is logical that even with Act 6 protections, lower balance loans would be paying a higher proportion of fees. at is, if a basic foreclosure filing carries with it $2,000 of costs and fees, a resident will pay a higher ratio of fees on a $40,000 versus a $60,000 loan.

If a borrower with a smaller Act 6 qualified loan inevitably pays a higher ratio, then the raw amount of fees that a non-Act 6 foreclosure carries should not be much higher than an Act 6 protected loan. However, in a comparison of means, the average Act 6 protected loan had $4,341 in associated costs and fees. e average non-Act 6 protected loan had $7,758 in associated costs and fees. 26 is difference, $3,417, is dramatic. If the majority of foreclosures involve the same basic level of work and therefore similar costs, these raw amounts should remain relatively constant. Instead, they suggest that lawyers may be charging fees in relationship to the size of the loans in foreclosure, not the cost of executing a foreclosure action.

Case law suggest that the “reasonable” level of fees for Act 6 protected loans is tied to the actual amount of work performed for a foreclosure filing.27 If true, this indicates that if the Act 6 threshold is raised and the “reasonable” fees provision of Act 6 was applied to loans between $50,000 and $200,000, a large number of Pennsylvania residents will experience a significant drop in the cost of curing their mortgage delinquency, as their fees more closely reflect the amount of work involved in their foreclosure filing. Undoubtedly, this will save some homeowners from the Sheriff Sale – the exact numbers of which can’t be known at this time.

Act 55 - Due to the fact that Act 55 uses identical trigger levels as HOEPA, Act 55 provides limited additional protections for Pennsylvania homeowners. Additionally, the reduced remedies for consumers further diminishes the impact of the law.

To understand the scope of Pennsylvania’s lending law that, when passed in 2001, was described as an attempt to deal with predatory lending, it is imperative to first understand the law that it is modeled after: the federal Homeownership and Equity Protection Act, or HOEPA. Congress enacted HOEPA in 1994 in order to address the problem of “reverse redlining” - a practice in which lenders and home improvement dealers, among others, target the residents of poor and minority communities for “credit on unfair terms, . . . peddling high-rate, high-fee home equity loans to cash-poor homeowners.”28 HOEPA, which Congress inserted into the Truth in Lending Act, was designed to give certain additional protections to consumers whose loans meet the statutory definition of a “high rate.”

A “HOEPA loan” is defined by Congress as a loan that contains “points and fees” that are more than 8 percent of the loan principal or that carries a rate more than 8 points above the comparable Treasury rate. Once either of these HOEPA thresholds are triggered, the law provides for additional disclosures and consumer protections in the transaction. Among the protections in HOEPA are limitations placed on pre-payment penalties and balloon mortgages. HOEPA also outlaws negative amortization loans and interest rate penalties that kick in after a default. Finally, HOEPA forbids lenders from engaging in a pattern or practice of making collateral-based loans to consumers who lack the ability to repay.

Although there are no good statistical numbers to measure HOEPA’s exact impact, there is near unanimous agreement that it has had a very real effect on the subprime lending market. Interviews with consumer lawyers as well as lenders suggest that HOEPA changed the typical subprime refinance loan from one that had extraordinarily high fees and rates, to loans that fall just below the HOEPA threshold. Fannie Mae and Freddie Mac will not purchase HOEPA loans; nor, for that matter, will many other secondary mortgage market purchasers or originators of subprime loans.

White hypothesizes that using the most basic economic pricing assumptions, HOEPA should have limited available credit to those consumers who were receiving loans above the HOEPA threshold. However, there has not been a single anecdote that HOPEA has limited access to subprime credit. HOEPA has, however, changed the characteristics of these loans being originated. He states that studies have shown that “high-priced loans above the given cutoffs (and/or the sale of single-premium credit insurance) declined substantially over time, at least as a share of first mortgages. is coincided with substantial growth in overall subprime loan volume.”29

Given worries about predatory lending, many states and cities passed additional legislation to protect their consumers from predatory lending practices. In this environment, the City of Philadelphia passed such a bill in the Spring of 2001.30 Within a few months, the Commonwealth passed the Consumer Equity Protection Act, or Act 55.31 is Act preempted Philadelphia’s law and was to impose statewide restrictions on predatory lending practices in Pennsylvania. While the effectiveness of Act 55 has never been studied, an analysis of the law and interviews suggest that if the law was intended to provide Pennsylvania homeowners with additional protections in the lending market, it has fallen short of its goals.

A commonly voiced criticism of Act 55 is that the rate and “points and fees” triggers set within the law are identical to that of HOEPA, giving Commonwealth residents no substantial additional protections. In fact, the remedies for consumers suing over an Act 55 violation are severely reduced versus those available in HOEPA. Act 55 does provide a few additional protections for consumers, such as a prohibition of balloons shorter than ten years, as opposed to the five years limited by HOEPA. However, there is little evidence to suggest that the limited additional provisions set out in Act 55 have resulted in any substantive changes in the Pennsylvania lending market, other than to ensure a uniform lending standard in the Commonwealth that relies on the limits set out by federal legislation.

Many argue that if Pennsylvania were to pass a law that lowered the HOEPA threshold, credit to low-income, high-cost borrowers would necessarily be reduced as well as become more expensive in nature. To best answer this question, it is useful to consider the experience of first state in the nation to enact comprehensive anti-predatory lending legislation: North Carolina. In 1999, North Carolina took official notice of reports of predatory lending, and passed a law that included a variety of anti-predatory lending provisions. Chief among the provisions of the North Caroline act was to effectively lower the HOEPA threshold from eight points to five.

In 2004, Quercia et al., summarized each of the studies of the impact of the North Carolina bill thus far, and concluded that virtually all research points to a small, measured decline in the amount of subprime lending in North Carolina. However, using access to a large, private loan database, they analyzed not only the changes in general subprime lending volume, but changes in specific types of loans and loan terms that the North Carolina bill attempted to curb.32

Quercia et al., come to three conclusions. First, although North Carolina’s subprime lending fell overall, this was concentrated solely in the refinance market, where most classically recognized predatory lending practices occur. Second, using terms of loans that the North Carolina law specifically banned, (extended pre-payment penalties, loans with balloons payments and loans with a loan-to-value ratios of 110% or more) they found that 90 percent of the decline in originations were loans that would be defined as having one or more characteristics of a predatory loan. A lending industry official interviewed by TRF for this study confirmed these statistics generally, stating that as a direct result of North Carolina’s legislation, he did not believe predatory lending generally existed any longer in the state, nor were North Carolina banks harmed in any way. Quercia, et al., also determined that contrary to statements made otherwise, the North Carolina lending bill did not result in an increase in the cost of subprime credit.33

North Carolina Commissioner of Banks Joseph Smith, in speeches, papers and an interview for this study, said there are as many subprime lenders within North Carolina today as there were in 1999, prior to the enactment of the law. Smith states that while his office receives thousands of complaints abut credit transactions each year, his office rarely hears of a North Carolina resident who wants credit and cannot acquire it. Smith hypothesizes that after the law was passed most lenders simply adjusted to the lower HOEPA-like threshold, and continued to do business in the State. ose lenders who relied solely on very high cost loans may have left, but there is still an abundant supply of subprime credit to purchase and refinance homes in the state.

Nationally, Morgan Stanley came to similar conclusions about predatory lending laws. While expecting to find that new lending laws were “crimping growth,” Morgan Stanley instead found that new laws were not hurting growth in the subprime market. Instead, they concluded that new lending laws may have, in fact, increased originations, as consumers have become more comfortable with the loan process. ey also found that consumer credit remained stable, and that brokers do not seem to be having trouble funding loans. Further, they concluded that the demand for subprime lending was fairly inelastic, meaning that tougher laws and changing lending practices should not greatly effect its growth.

Regionally, New Jersey passed an anti-Predatory Lending law in November of 2003. Wall Street agencies and industry officials initially expressed concern about the law, but after subsequent amendments and clarifications, Moody’s, S&P, Fitch and Fannie Mae continue to do business in the New Jersey subprime market. Robert Levy, Counsel to the Mortgage Bankers Association of both New Jersey and Pennsylvania, (as well as an advisory committee member for this report), stated that although he thought there was an initial number of lenders who left New Jersey, lenders “…have since come back into the market” after an amendment was passed that, among other things, lowered the threshold of a high cost loan to 4.5 percent while eliminating a provision on “flipping.” Since the law took effect, predatory lending complaints in New Jersey dropped 33 percent.

Act 91 - HEMAP, a unique program to Pennsylvania has successfully saved 4,222 homeowners from foreclosure over the last four years. However, the program excludes some of the most vulnerable Pennsylvania homeowners.

Act 91, which created HEMAP, was passed in 1983 as the reorganization of the steel industry in Western Pennsylvania precipitated a foreclosure crisis within the Commonwealth. e foreclosure rate, although acute at that time, was only half what it is today. Act 91 mandates that notice be given to virtually any non-federally insured mortgage thirty days prior to the filing of a foreclosure. e notice lets the borrower know that they are delinquent and of the intent to file a foreclosure. To be eligible for HEMAP, a homeowner must meet with a PHFA approved housing counselor within thirty-three days of the Act 91 postmark. is meeting immediately stops the foreclosure action for another thirty days, while the homeowner is given the chance to apply for HEMAP. If the homeowner does so, the foreclosure is stopped for up to 60 more days as PHFA decides whether to approve or decline the application for assistance.

e approval process for HEMAP focuses on whether the circumstances of the delinquency were beyond the borrower’s control, and whether or not the borrower will realistically be able to begin payments on the loan within two years. For residents accepted into the program, PHFA pays the amount of the mortgage in delinquency, and helps cover the monthly mortgage payment for up to two years.34

HEMAP was initially modeled after the assignment program run by the Department of Housing and Urban Development (HUD) for FHA mortgages. At the time Act 91 was enacted, the HUD Program covered all FHA loans, so Act 91 specifically excluded these loans from eligibility for assistance. Act 91 was to supplement HUD’s assignment program. HUD, however, has since ended the assignment program, but Act 91 still prohibits Pennsylvania residents with FHA loans from receiving assistance. As the chart indicates, depending on the county, anywhere from 9% to 32% of loans in foreclosure between 2000 and 2003 were FHA (see figure V-3: HEMAP and FHA).

Cause 6: Abusive lending practices are evident in segments of the mortgage industry.

In Pennsylvania, as in other states, there have been a number of documented and officially alleged abuses perpetrated by brokers, contractors, builders / developers, appraisers, lenders and loan servicers against consumers in the making of a purchase or refinance mortgage, or related service. ese abuses, taken together, have impacted transactions numbering (conservatively) in the thousands. e inventory of practices that follows is not simply a set of anecdotes. e inventory represents systemic practices – illegal and unethical - endemic to a variety of industries operating in Pennsylvania, all of which contribute directly or indirectly to Pennsylvania’s foreclosure rate. e sources of these practices include, but are not limited to:

• Case filings and settlements by the US Attorney for the Eastern District of PA;

• Case filings by the Attorney General for the Commonwealth of PA;

• Case filings and settlements by the Federal Trade Commission;

• Case filings and settlements by the US Department of HUD;

• Court decisions in cases filed by private attorneys;

• Declarations by ratings agencies regarding their systematic observations on pools of mortgages that they have rated;

• Policy statements by the nation’s GSEs (Fannie Mae and Freddie Mac);

• Congressional testimony by representatives of national trade associations regarding widespread practices impacting their industry.

Practices have been categorized in terms of where they are most likely to occur. e list of demonstrable abusive practices includes, but is not limited to:

Lenders

Abuses attributed to lenders generally revolve around actions (or higher level organizational failures to act) associated with the loan origination process. It is at this stage of the process that we observe the elevated “…risk of misrepresentation, fraud, and adverse selection.”35 Among the practices that have been demonstrated in various cases are:

• Downgrading borrower’s credit profile in order to charge higher interest rates and fees;

• Forcing borrowers who are in trouble with their mortgages to sign “deeds in lieu of foreclosure” along with forbearance agreements – those deeds-in-lieu would be exercised by the lender in the event of a future breach of the forbearance agreement;

• Failing to escrow funds paid by the borrower where escrow is appropriate and/or required by law;

• Refinancing low (or no) interest rate loans into higher rate loans;

• Splitting loans, placing the fees for one loan into another (usually higher interest rate and fee) loan;

• Misrepresenting critical loan terms (e.g., interest rate, fees, prepayment penalties);

• Failing to provide appropriate notification – in a timely manner - when a loan is a high-cost / HOEPA loan;

• Failing to provide appropriate ECOA notices when the loan for which an applicant actually applied is not the loan that they received; and

• Loan officers improperly demanding cash payments that were improper and not disclosed on any settlement documents.

Brokers

Mortgage brokers serve as the conduit of business to mortgage lenders. In fact, although independent information is not readily available, the Pennsylvania Association of Mortgage Brokers reports that brokers are involved in two-thirds of loan originations nationwide.36 us, brokers are a significant part of the mortgage origination market. eir relationship to the borrower, although legally clear (i.e., that although it is the borrower who is paying the broker for services, the broker has no fiduciary responsibility to that borrower), is not clear to the borrower. Moreover, based on the usual compensation structure for brokers (i.e., a percentage of the total loan amount and/or a yield-spread-premium) the broker has an economic incentive to create a transaction that is most advantageous to the broker, not the borrower. In some instances, that incentive leads to the selling of a loan that places the borrower at a distinct disadvantage thereby increasing the probability of mortgage default. Among the practices attributed to the acts of brokers in Pennsylvania are:

• Inducing homeowners to take a (large) first mortgage refinance loan rather than a (small) home improvement loan commensurate with their financial need and circumstance;

• Charging / collecting fees for services not performed;

• Selling mortgage products that are inappropriate (or unsuitable) given the minimal borrower’s unique circumstance in order to get the borrower “qualified”;

• Adjustable rate mortgages that can only adjust upward (especially when interest rates are at a historically low point);

• Balloon mortgages when the borrower’s prospects for being able to pay (or comprehend) the balloon are minimal;

• Broker-initiated successive refinancing of a mortgage without any demonstrable benefit to the borrower;

• Falsifying bank statements, W-2 forms and employment verifications;

• Misrepresenting fake gifts as down payments;

• Representing a borrower’s mortgage payment would be less than it actually was; and

• Making loans either without regard for the borrower’s ability to pay, or falsely inflating the borrower’s ability to pay.

Servicers

In today’s mortgage market, the individual (or entity) that made the loan to the borrower is oftentimes not the entity responsible for collecting the monthly payments. at role is played by the mortgage servicer. In general, the responsibilities of the servicer include, but may not necessarily be limited to: (1) collecting the borrower’s monthly payments and crediting the borrower’s account; (2) if an escrow is established, paying property taxes and/or property insurance premiums; (3) sending payments to the investor that actually owns the loan on which the borrower is paying; (4) engaging in appropriate loss mitigation activities including counseling, forbearance agreements, recasting of the loan and where all that fails, filing the foreclosure action. Assuming that the borrower is fulfilling their mortgage obligation, failure on the part of the servicer in any of these roles may result in the borrower ending up in foreclosure. e specific issues associated with servicing include:

• Making it difficult (if not impossible) for the borrower to accurately comprehend their true indebtedness or cure a default if necessary;

• Posting payments made on-time late to the borrower’s account;

• Failing to respond to qualified written requests for an accounting on a borrower’s loan;

• Force-placing property insurance on properties that have existing policies;

• Charging fees not permitted (or justified); and

• Failing to engage in proper loss mitigation strategies (which are likely dictated by the investor or in the case of FHA loans, the US Department of HUD).

Contractors

Ordinarily the contractor is not part of the official lending process. In some instances though, the contractor will refer the homeowner to a broker or a lender. Where this may become abusive is where the contractor steps out of that contracting role and into other parts of the transaction, or where they contract to do work which they do not do, yet receive full compensation for that work.

• Builders / home improvement contractors acting as (unlicensed) mortgage brokers; and

• Charging for work that was never performed.

Builders / Developers

e true market value of a home is both important and difficult to establish. at said, each year between 2000 and 2004, more than 40,000 new housing units were authorized annually. ere are no figures available on the extent to which these new homes are reasonably valued. e Pennsylvania Attorney General has however alleged the following against a limited number of builders / developers operating in Monroe County, Pennsylvania:

• Falsely inflating the value of a property in order to give the appearance later that concessions were made on the price of a home;

Appraisers

As stated by the Appraisal Institute in some of its public information material:

e role of the appraiser is to provide objective, impartial and unbiased opinions about the value of real property— providing assistance to those who own, manage, sell, invest in and/or lend money on the security of real estate.

Pressure may be placed on the appraiser to “hit” a particular value; that pressure may be exerted by the broker, lender or perhaps even the consumer. at sort of pressure impinges on the appraiser’s ability to produce a fair and impartial estimate of market value. In Pennsylvania, appraisers are licensed by the Pennsylvania Department of State. When the appraiser succumbs to that pressure, the likelihood of a transaction ending up in foreclosure is elevated.37 Specific actions attributed to appraisers include:38

• Yielding to pressure from brokers / lenders / consumers to produce inflated appraisal values in an effort to support larger loans;

• Voluntarily cooperating with builders / developers to estimate market inflated values in an effort to support higher sale prices.

Beyond the aforementioned practices, the nation’s GSEs have made certain declarations about characteristics of loans that they feel would be abusive and will not purchase loans with those characteristics. Generally, Fannie Mae and Freddie Mac have established similar policies with respect to loan purchases. Among the characteristics that preclude their purchase of a loan are:

• HOEPA loans or loans with fees exceeding 5 percent of the loan, except where that would make the transaction unprofitable;

• Prepaid single premium credit life insurance;

• Prepayment penalties, except under limited circumstances;

• Servicers that do not regularly report credit information to the credit bureaus;

• Servicers that do not escrow, except under limited circumstances;

• Mandatory arbitration clauses.

Cause 7: e costs of homeownership in Pennsylvania including those costs associated with maintaining an older housing stock, property taxes and energy costs have risen.

• Aging Population and Housing Stock

Pennsylvania has the fifth largest senior population in the nation. Over 28% of homeowners in Pennsylvania are aged 65 or older compared to 25% nationally. Additionally, while 14% of all homes in the nation were built prior to 1939 – 29% were in Pennsylvania. Most striking, is the frequency with which older home owners in Pennsylvania live in these older homes. Compared to national rates, a homeowner aged 65 or older in Pennsylvania is twice as likely to live in a home built before 1939. In Pennsylvania, 36% of all homeowners over the age of 65 lived in a home built before 1939; in the nation, only 18% of elderly homeowners lived in such homes (see Figure V-IV: Age of House).

is may be raising Pennsylvania’s foreclosure rate as older homeowners are more likely to be living on fixed incomes; their older homes may need more repair than they can afford.

• Property Taxes

Pennsylvania’s per capita property taxes increased 70% between 1989 and 1999 – faster than the 53% national increase. Pennsylvania’s tax burden relative to income also increased in this 10 year period, while national trends decreased.39 As a result, rising taxes may contribute to growing homeowner cost-burdens in some communities (i.e. Allegheny County).

Exacerbating the problem of high taxes, interviewees suggest that many subprime lenders do not escrow for property tax payments in order to keep the servicing costs of the loan down. Others suggest that some lenders do it in order to make the monthly mortgage payment look lower than it actually would be if taxes were being escrowed. While the second practice could also be considered deceptive, both may catch the borrower unaware when property tax payments come due to the County. If the borrower did not save enough each month to pay taxes, the result can be financially devastating.

• Energy Costs

While electricity costs have varied year-to-year and went up slightly between 1990 and 2000, natural gas costs have risen substantially – increasing the annual home heating bills of residents across Pennsylvania. According to the Energy Information Agency at the Department of Energy, natural gas prices rose by 29% between 1990 ($6.37 per million btus) and 2000 ($8.20 per million btus). Electricity prices rose from $27.03 per million btus in 1990 to $27.94 per million btus in 2000.

Cause 8: Consumer expenditures on health care costs have risen faster than the growth in incomes.

Nationally, between 1997 and 2002, consumer spending on health care increased by 27%; while incomes only rose by 24%.40 Of the 4,222 homeowners approved for HEMAP assistance in Pennsylvania between 2000 and 2003, 40.3% cited medical costs as the reason for their financial distress.

Members of the Pennsylvania Realtors Association interviewed for this study uniformly indicated that a growing number of homeowners, when faced with these health care costs, either pay their medical bills and forgo mortgage payments or utilize built-up equity in their homes.

CONCLUSION

TRF believes these causes to be the primary contributing factors to Pennsylvania’s foreclosure problem. While many of the causes reflect more macroeconomic and market forces beyond the control of the Pennsylvania Legislature or the Secretary of Banking, many are not. To successfully address these causes, Pennsylvania will need to undertake an affirmative approach to legislating, law enforcement and policy design while remaining sensitive to the needs of a vital home mortgage market.

Ackelsberg, Irv, and George Gould. “Mortgage Foreclosure,” Pa. Consumer Law, ed. C. Carter. Bisel, 2003.

ACORN. “Stolen Dreams: Predatory Lending in San Jose.” ACORN. San Jose, CA: 2003.

ACORN. “A Dream Deferred: Predatory Lending in Colorado.” ACORN. 2002.

ACORN Fair Housing. “Predatory Lending in South Central Pennsylvania: A Review of Rising Foreclosure Filings and the Relationship to Predatory Lending.” For the South Central Assembly for Effective Governance. Baltimore, MD: 2003.

Alanna L. Curry, et al., Plaintiffs, vs. Fairbanks Capital Corporation, Defendant. United States District Court for the District of Massachusetts. Civil Action No. 03-10895-DP

Apgar, William, Allegra Calder and Gary Fauth. “Credit Capital and Communities: the implications of the changing mortgage banking industry for community based organizations.” Report CCC04-1.

Appraisal Institute. “Some Commonly Asked Questions About Real Estate Appraisers and Appraisals.” Appraisal Institute. Chicago, IL. 2002.

Appraisal Institute. “What is Client Pressure?,” AppraisalInstitute.org. 2002. <http://www.appraisalinstitute.org/govtaffairs/issues/client_pressr.asp>

American Society of Appraisers. Appraisal Societies Urge Cooperation With Mortgage Industry Leaders, Government Officials and Consumer Groups to End Improper Pressure on Appraisers. Herndon: 2004. <http://www.appraisers.org/news/announcedetail.cfm?announcementId=152>

Baxter, Vern and Mickey Lauria. “Residential Mortgage Foreclosure and Neighborhood Change.” Housing Policy Debate 11.3 (2000).

Benson, Clea. “Bankers take aim at Phila.’s lender law,” Philadelphia Inquirer 20 June 2001: B11.

Berkovec, Jim and Peter Zorn. “How Complete is HMDA?: HMDA Coverage of Freddie Mac Purchases.” Journal of Real Estate Research 11 (1996): 39-55.

Bond Market Association, e. e Secondary Market for Subprime Mortgages: A Common Sense Approach to Addressing Assignee Liability through Federal Legislation.” March 2004. < http://www.bondmarket.com/Legislative/Subprime_Lending_Whitepaper_032904.pdf>

Bourassa, Steven C. “Predatory Lending in Jefferson County: A Report to the Louisville Urban League.” Urban Studies Institute, School of Urban and Public Affairs, University of Louisville. Louisville, KY: 2003

Bunce, Harold L. et al. “Subprime Foreclosures: e Smoking Gun of Predatory Lending?” Housing Policy in the New Millennium Department of Housing and Urban Development. Washington, DC: 2000.

Burnett, Kimberly, Christopher E Herbert and Bulbul Kaul. “Subprime Originations and Foreclosures in New York State: A Case Study of Nassau, Suffolk and Westchester Counties.” Abt Associates. Cambridge: 2002

Bradford, Calvin. “Risk or Race? Racial Disparities and the Subprime Refinance Market.” Center for Community Change. Washington, DC: 2002.

Calem, Paul S, Kevin Gillen and Susan Wachter. “e Neighborhood Distribution of Subprime Lending.” University of Pennsylvania Law School, Institute for Law and Economics Research Paper No. 03-39U. 2004.

Carr, James H and Lopa Kolluri. “Predatory Lending: An Overview.” In Financial Services in Distressed Communities: Issues and Answers. Fannie Mae Foundation. Washington, DC: August 2001.

Case, Karl E. “Real Estate and the Macroeconomy.” Brookings Papers on Economic Activity 2 (2000): 119-162. Brookings Institute. Washington, DC.

Center for Statistical Research. “Subprime Lending in Arizona: An Empirical Analysis.” Released by Arizona Financial Services Association. Alexandria, VA: 2004.

Courchane, Marsha J., Brian Surette and Peter Zorn. “Subprime Borrowers: Mortgage Transitions and Outcomes.” Journal of Real Estate Finance and Economics 35.9 (2004): 365-392.

Cutts, Amy Crews and Robert A. VanOrder. “On the Economics of Subprime Lending.” Freddie Mac Working Paper Series. 04-01. Freddie Mac (2004).

Davies, Paul. “Lending Bill Becomes Law,” Philadelphia Daily News 20 Apr., 2001: 5.

Dietz, Richard and Ramon Garcia. “Examining the Rising Foreclosure Rate.” e Regional Economy of Upstate New York. Federal Reserve Bank of New York, Buffalo Branch. Buffalo: 2003.

Eissner, Kelsch & Pendley. “Reviewing Residential Mortgage Originators.”

Fitch IBCA Residential Mortgage Special Report. July 2001. < http://www.cswv.com/research/mortgages/>

Elmer, Peter J and Steven A Seelig. e Rising Long-Term Trend of Single Family Mortgage Foreclosure Rates. FDIC, Division of Research and Statistics. Washington, DC: 1998.

Evans-Cowley, Jennifer. “Buy Now, Pay Later: Risks of Aggressive Financing.” Mortgage Lending 1625 (2003).

Engel, Kathleen C. and Patricia A. McCoy. “A Tale of ree Markets: e Law and Economics of Predatory Lending.” Texas Law Review 80.6 (2002):1255-1381.

Fannie Mae. Expanding Responsible Lending: Increasing Access to Suitable Mortgage Credit. Washington, DC: 2003

Fannie Mae. Frequently Asked Questions: Predatory Lending Practices. Fannie Mae, Washington: 2004. <http://www.fanniemae.com/faq/231001q.jhtml?p=FAQ>

Fannie Mae. “e Growing Demand for Housing.” 2002 Fannie Mae National Housing Survey. Washington, DC: 2002.

Fannie Mae. “Understanding America’s Homeownership Gaps.” Fannie Mae National Housing Survey. Washington, DC: 2004.

Federal Trade Commission. “Fairbanks Capital Settles FTC and HUD Charges.” Washington: 12 Nov 2003. <http://www.ftc.gov/opa/2003/11/fairbanks.htm>

Ferguson, Jr., Vice Chairman Roger W., Federal Reserve Board. Remarks to the National Bankers Association. Nashville, TN, 6 October, 2004.

Fitch IBCA, Duff & Phelps. “Rating U.S. Residential Subprime Mortgage Securities.” Residential Mortgage Criteria Report. 2001. <http://www.fitchratings.com>

Fitch IBCA, Duff & Phelps. “Reviewing Residential Mortgage Originators.” Residential Mortgage Special Report. 2001. <http://www.fitchratings.com>

Fleishman, Sandra. “Sarbanes Reviving Effort to Crack Down on Predatory Lending.” e Washington Post May 2002: E.

Freddie Mac. “Key Corporate Initiatives: Anti-Predatory Lending.” FreddieMac.com. 2004. <http://www.freddiemac.com/corporate/initiatives/protection/predlend.html>

Freddie Mac. Combat Predatory Lending. 2004. Available from the World Wide Web: <http://www.frediemac.com/singlefamily/anti-predatory-faq.html>

Goldstein, Deborah. “Protecting Consumers from Predatory Lenders: Defining the Problem and Moving Towards Workable Solutions.” Harvard Civil Rights–Civil Liberties Law Review winter (2003).

Goldstein, Ira. “Bringing Subprime Mortgages to Market and the Effects on Lower-Income Borrowers.” Joint Center for Housing Studies Working Paper Series. BABC04-7. Harvard University, (2004).

Gramlich, Governor Edward M. “Subprime Mortgage Lending: Benefits, Costs and Challenges,” Remarks to the Financial Services Roundtable Annual Housing Policy Meeting, Chicago, IL, 21 May, 2004.

Greenberger, Scott S. “Study says blacks squeezed on mortgages.” Boston Globe May 2002.

Gruenstein, Debbie and Christopher E. Herbert. “Analyzing Trends in Subprime Originations and Foreclosures: A Case Study of the Atlanta Metro Area.” Abt Associates. Cambridge, MA: 2000.

Gruenstein, Debbie and Christopher E. Herbert. “Analyzing Trends in Subprime Originations and Foreclosures: A Case Study of the Boston Metro Area.” Abt Associates Cambridge, MA: 2000.

Harney, Kenneth. “Lenders Pressing for Inflated Values, Appraisers Say: “Inflated appraisals could hurt consumers if economy dips.” Washington Report, Federal Managers Association, Alexandria: November 2000. <http://www.newcomerinfo.com/news_stories/inflated_appraisals.htm>

Hummel, Alan E. Testimony delivered to U.S. House Subcommittee on Housing and Community Opportunity. 25 Feb 2003.

<http://www.appraisalinstitute.org/govtaffairs/downloads/ltrs_tstmny/1Cngrsnl_Tstimny.pdf>

Immergluck, Dan and Geoff Smith. “Risky Business – An Econometric Analysis of the Relationship Between Subprime Lending and Neighborhood Foreclosures.” Woodstock Institute. Chicago: 2004.

Inside Mortgage Finance. Inside B&C Lending: An In-Depth Report on the Dynamic Subprime Mortgage Business. 9.1 (2004).

Inside Mortgage Finance. Inside B&C Lending: An In-Depth Report on the Dynamic Subprime Mortgage Business. 9.24. (2004)

Issues PA, “Pennsylvania’s Property Taxes – Too High?” 2002.

Joint Center for Housing Studies of Harvard University. e State of the Nation’s Housing 2004. Cambridge: 2004.

Lax, Howard, Michael Manti, Paul Raca and Peter Zorn, “Subprime Lending: An Investigation of Economic Efficiency.” Housing Policy Debate, 15.3 (2004): 533-571.

Lampe, Donald C. “Trigger Happy: Enactment and Aftereffects of North Carolina’s “Predatory Lending” Law.” e Federalist Society Conference on Predatory Lending. Washington, DC. 24 July 2003.

Leonhardt, David. “Wide Racial Disparities Found in Cost of Mortgages,” e New York Times CLI.52,105 May 2002.

Lucrecia L. Taylor, et al., Plaintiffs vs. McGlawn & McGlawn, et. al., Defendants. Case Nos. 200027668 & 200201787. Pennsylvania Human Relations Commission. 26 Oct 2004.

Mansfield, Cathy Lesser and Alan White. “Literacy and Contract.” Stanford Law & Policy Review 13.2 (2002)

Margaret Newton, et al., Plaintiffs, v. United Companies Financial Corp., et al., Defendants No. CIV. A. 97-5400 United States District Court, E.D. Pennsylvania Nov. 5, 1998. p.444-464 24 Federal Supplement, 2d Series

Marsico, Richard D. “Patterns of Lending to Low Income and Minority Persons and Neighborhoods: e 1999 New York Metropolitan Area Mortgage Scorecard.” New York Law School Journal of Human Rights 27 (2001).

Morgan Stanley. “Channel Check: Surprisingly Strong Subprime Growth.” MorganStanley.com, 1 Aug, 2002.

Munoz, Ric and David Blatt. “Stealing the American Dream: Predatory Lending in Oklahoma.” Community Action Project. Oklahoma: 2003.

National Association of Attorneys General. Multistate Actions: States Settle With Household Finance: up to $484 Million Consumers. Washington: 11 Oct 2002. <http://www.naag.org/issues/20021011-multi-household.php>

National Association of REALTORS. “Rising Foreclosure Rates in Indiana: An Explanatory Analysis of Contributing Factors.” National Association of REALTORS, Research Division. 2003.

National Training and Information Center. “Analysis of Chicago Foreclosures.” Chicago: 2002.

National Training and Information Center. “Slash and Burn Financing: A Study of CitiFinancial’s Recent Lending in Chicago.” Chicago: 2001.

Neighborhood Reinvestment Corporation. “Analyzing Trends in Subprime Originations: A Case Study of Connecticut.” NeighborWorks Campaign For Home Ownership. Washington, DC: 2002

Office of the Comptroller of Currency. Economic Issues in Predatory Lending. Washington, DC: 2003.

Office of the Inspector General for the U.S. Department of Housing and Urban Development. Audit of the Federal Housing Administration’s Financial Statements for Fiscal Years 2004 and 2003. Washington, DC: 2004.

Option One Mortgage. Fair Lending. 2004. <http://www.oomc.com/corp/corp_bestpract_fair_protection.asp>

Litan, Robert. “Unintended Consequences: e Risks of Premature State Regulation.” Prepared for the American Bankers Association. <http://aba.com/Industry+Issues/PredatoryLendingMenu.htm>

PA Act No.6, January 30, 1974, 41 P.S., Sections 404-406

Pennington-Cross, Anthony. “Subprime Lending in the Primary and Secondary Markets.” Journal of Housing Research. 13.1 (2002).

Pennington-Cross, Anthony. Subprime and Prime Mortgages: Loss Distributions. Office of Federal Housing Enterprise Oversight. Washington, D.C: 2003.

Pennsylvania Office of Attorney General. AG Fisher Announces landmark Agreement With Household International; Approximately $30 Million Slated for PA Consumers Harrisburg: 11 Oct 2002. <http://www.attorneygeneral.gov/press/release>

Pennsylvania Department of Banking. State-Chartered Financial Institutions and Licensees Activity Quarterly Report. Harrisburg: 2004.

Pitkin, Bill and Neal Richman. “Subprime Lending and Neighborhood Conditions in the City of Los Angeles.” UCLA Advanced Policy Institute. Los Angeles: 2001

Pittsburgh Community Reinvestment Group. “Disparities in Lending: A Study of Subprime Lending in Pittsburgh and Allegheny County,” prepared by Christine Arriola. Pittsburgh, PA: 2004.

Pugh, Tony. “Refinancing Rates Higher for Minorities,” Philadelphia Inquirer May 2002: C1.

Quercia, Roberto, Michael Stegman and Walter Davis. “Assessing the Impact of North Carolina’s Predatory Lending Law.” Housing Policy Debate, 15.3 (2004): 573-601.

Reckard, E. Scott. “Sub-Prime Loans Going to More Minorities,” Los Angeles Times May 2002: C5.

Ruggles, Steven, Matthew Sobek, Trent Alexander, Catherine A. Fitch, Ronald Goeken, Patricia Kelly Hall, Miriam King, and Chad Ronnander. Integrated Public Use Microdata Series: Version 3.0 (Machine-readable database). Minnesota Population Center (producer and distributor), Minneapolis: 2004. (5 Percent Public Use Microdata Samples for Pennsylvania: 1980-2000)

Scheesele, Randall M. HUD Subprime and Manufactured Lender List. Department of Housing and Urban Development. <www.huduser.org/datasets/manu.html>

Smith, Jim. 2002. “Olney, Real Estate Dealer Admits HUD, Bank Fraud.” Philadelphia Daily News 27 June 2002. <http://www.dfw.com/mld/dailynews/2002/06/27/news/local/3545581.htm?1c>

Smith, Joseph. “Financial Literacy, Regulation and Consumer Welfare,” North Carolina Banking Institute Journal 8 (2004): 77-97 <http://www.nccob.org/NCCOB/Researchers/CommissionersPage/>

Smith, Joseph. “North Carolina’s Predatory Lending Law: Its Adoption and Implementation,” Paper Presented to National Conference of State Legislatures Annual Meeting, Denver, 26 July, 2002. <http://www.nccob.org/NCCOB/Researchers/CommissionersPage/>

SMR Research Group. “e Subprime Lending Industry and Allegations of Predatory Practices.” National Home Equity Mortgage Association. 2000. <http://www.butera-andrews.com/legislative-updates/directory/Background-Reports/FINAL%20SMR%20Report.pdf>

Sopko, Kate, et al. “Home Insecurity: Foreclosure Growth in Ohio” Policy Matters Ohio. Cleveland, OH: 2000.

Stock, Richard D. 2001. “Predation in the Subprime Lending Market: Montgomery County, OH.” Dayton, OH: Center for Business and Economic Research, University of Dayton.

e Reinvestment Fund and the Department of Banking, Commonwealth of Pennsylvania. A Study of Mortgage Foreclosures in Monroe County and e Commonwealth’s Response. Harrisburg: 2004. <http://www.banking.state.pa.us>

United States Attorney’s Office Eastern District of Pennsylvania. Allentown Real Estate Agent and Mortgage Broker Charged in Predatory Lending Scheme. Philadelphia: 2004. <http://www.usdoj.gov/usao/pae/Documents/balf.html>

United States Attorney’s Office Eastern District of Pennsylvania. U.S. Reaches Agreement with Mortgage Lender to End Alleged Predatory Lending Practices. Philadelphia: 2003. <http://www.usdoj.gov.usao/pae/News/Pr/2003/dec/jointa.html>

United States Attorney’s Office Eastern District of Pennsylvania. U.S. Reaches Settlement with Mortgage Lending Company. Philadelphia: 2004. <http://www.usdoj.gov/usao/pae/News/Pr/2004/aug/ms.html

United States Department of Labor, Bureau of Labor Statistics. Consumer Expenditures, 1997-2002.

Weicher, John C. e Home Equity Lending Industry: Refinancing Mortgages for Borrowers with Impaired Credit. Indianapolis: Hudson Institute, 1997.

White, Alan. “Risk-Based Mortgage Pricing: Present and Future Research.” Housing Policy Debate 15.3 (2004): 503-531.