Asset Building in Rural Communities: The Experience of Individual Development Accounts*

Transcription

1 Rural Sociology 72(1), 2007, pp Copyright E 2007 by the Rural Sociological Society Asset Building in Rural Communities: The Experience of Individual Development Accounts* Michal Grinstein-Weiss School of Social Work University of North Carolina at Chapel Hill Jami Curley School of Social Work St. Louis University Pajarita Charles School of Social Work University of North Carolina at Chapel Hill ABSTRACT This study examines the unique experiences of low-income rural participants in an asset building program the Individual Development Account. Using data from the American Dream Demonstration, this study addresses three main questions: (1) What are the individual characteristics associated with saving outcomes among rural IDA participants? (2) What are the program characteristics associated with savings among rural participants? (3) What are the policy implications for supporting asset building in rural areas? To answer these questions we conduct an Ordinary Least Squares regression analysis. The results suggest that low-income rural participants have the ability and willingness to save toward the accumulation of assets in IDAs. Looking at individual characteristics, home ownership appears to be an important predictor of savings. In addition, this study suggests that program characteristics (financial education, peer group meetings, match rate, direct deposit, and monthly saving target), not merely individual characteristics, are important in explaining saving performance for this group. Similar to the beginning of many new eras, the dawning of the 21 st century has brought new opportunities as well as new challenges to the stability of our economy. New technology offers more efficient methods of production while the continuing influence of globalization increases * The authors wish to thank the foundation funders of ADD, especially the Ford Foundation, Charles Stewart Mott Foundation, FB Heron Foundation, and Metropolitan Life Foundation for funding ADD research; the IDA program staff at the 14 research sites of the American Dream Demonstration (ADD); the Corporation for Enterprise Development for implementing ADD; Lissa Johnson, ADD research project manager; Margaret Clancy, for ensuring quality of the monitoring data; Mark Schreiner for data management and preparation; and Michael Sherraden and Karen Edwards for their support and leadership. Please direct all correspondence to: Michal Grinstein-Weiss, School of Social Work University of North Carolina at Chapel Hill 301 Pittsboro Street - CB 3550 Chapel Hill, NC , (V) , (F) , (E) michalgw@ .unc.edu.

2 26 Rural Sociology, Vol. 72, No. 1, March 2007 market availability for our goods. Yet, when a region has a hard time transitioning to a new economy, the challenges produced by these changes are often overwhelming and can create hardship. Rural America is currently facing many of the difficulties associated with these changing economies, thus affecting their current economic sustainability and development. The industry base change from manufacturing to service in the late half of the 20 th century produced high unemployment rates from the loss of factory jobs. Although the new service base created jobs, most were low-wage with minimal or no benefits (Falk and Lobao 2003). These conditions only added to the rising poverty rates for rural areas. While the rural poverty rate, as well as the national poverty rate, began to decline after reaching a high of 17.3 percent in 1993, it has since begun to rise again and in 2002 stood at 14.2 percent (U.S. Census Bureau 2004). Furthermore, rural poverty rates have historically been higher than urban poverty rates, leaving rural communities at even more of a disadvantage (Economic Research Service n.d.). One policy approach for social and economic development being discussed in current dialogues is wealth creation (asset building). Some researchers have suggested that asset building in rural areas might be a viable solution to help reduce poverty and increase economic assets in these regions (Curley and Grinstein-Weiss 2003; Dorward et al. 2001). The purpose of this study is to examine the performance of rural participants in an assets building program the Individual Development Account (IDA). IDAs are matched savings accounts for lowincome households, where the savings are used for specific purposes including home purchase, post-secondary education, and microenterprise. Literature Review Asset Building Theory The concept of asset-building evolved over the last several decades out of discussions regarding alternative methods for social and economic development of disadvantaged populations and geographic regions. Policymakers were specifically interested in exploring asset accumulation strategies to reduce poverty that were outside the realm of customary means-tested social programs such as welfare (Sherraden, Curley, and Grinstein-Weiss 2003). The rationale for building assets through mechanisms other than income support stems in part from what Sen (1985, 1993, 1999) identifies as strengthening human and economic capabilities. Asset-building policy was developed to influence

3 Asset Building in Rural Communities Grinstein-Weiss et al. 27 and improve many aspects of individual and household welfare including knowledge, resources, and functioning skills (Sherraden, Schreiner, and Beverly 2003). This asset-building approach, referred to as capacity building, can be examined in a variety of dimensions. One of the most important aspects stems from the concept of human assets or capital. According to Becker (1964), human capital is the range of personal assets and resources belonging to an individual, such as skills, education, and intellectual ability that influence future money and psychological outcomes. He maintains that human capital represents an estimated 75 percent of total wealth. Still another dimension of capacity building is through the growth of tangible and financial capital. Sherraden (1988, 1991), whose work has been instrumental in advancing this concept, proposes that building financial assets has far-reaching effects on the current well-being of individuals, in addition to the well-being of future generations. Based on these ideas, Sherraden (1991) put forth a welfare-based asset policy designed to increase the tangible and financial assets of low-income households. The rationale for doing so developed out of the premise that the poor do not participate in asset-based policies and programs (e.g., 401(k)s, 403(b)s, individual retirement accounts (IRA), and educational savings accounts) and that asset-based policies operate primarily out of a tax system that is highly regressive and generally not beneficial to the poor (Sherraden 2001). Moreover, strict asset limits in means-tested programs such as Temporary Assistance for Needy Families, Supplemental Security Income, Medicaid, and food stamps serve as strong disincentives for the poor to accumulate assets (Powers 1998; Sherraden 2005; Ziliak 1999). As participants in these programs reach or exceed asset limits, their benefits are either reduced or eliminated altogether (Lowe and Weisner 2004). Thus, even the most modest asset accumulation can jeopardize a low-income family s eligibility for assistance with meeting basic needs. Sherraden proposed to make already existing asset-building policies and programs more inclusive by bringing the poor into the system and by providing subsidies to low-income households to assist and encourage participation in these programs. One such program is the Individual Development Account (IDA), an interest-bearing, tax-benefited, longterm, restricted account designated for specific expenditure purposes (Sherraden 2001). IDAs are saving programs that provide financial incentives and an institutional structure that promote saving behavior. Individual account holders accumulate their own savings, in addition to

4 28 Rural Sociology, Vol. 72, No. 1, March 2007 matched savings provided by public and/or private institutions, in order to promote their self-sufficiency and long-term well-being through homeownership, education, and microenterprise. The mainstream economic theory, the life cycle hypothesis (LCH), assumes that consumption and savings patterns represent an individual s age or stage within the life cycle, with a majority of saving occurring in the middle years (Modigliani and Ando 1957). The limitation of this theory is that it is biased toward middle and upper income groups (Beverly 1997). The U shape that reflects the age-savings patterns, which are predicted by the life cycle hypothesis and characterized by higher income households, failed to hold for lowincome households (Ziliak 1999). The institutional saving model first developed by Beverly and Sherraden (1999) and later modified by Sherraden et al. (2003) and again by Sherraden and Barr (2005) seeks to explain saving behaviors among low income populations in a way that is different than previous saving behavior theories. The institutional saving model identifies seven components of institutional characteristics that influence saving performance beyond flows of income and personal preferences. These include (1) access, (2) information, (3) incentives, (4) facilitation, (5) expectations, (6) limits, and (7) security. (1) Access to an institutionalized saving opportunity is hypothesized to promote saving by calling attention to the need for and benefits of saving. (2) Financial information provided through education can positively influence saving participation and levels of contribution by increasing participants understanding of their economic vulnerabilities and future return to savings. (3) Incentives can motivate individuals to save because they provide participants with new opportunities and enticements. (4) Facilitation, a key feature of most saving programs, refers to the practice of assisting people with their saving process. Those provided with such assistance tend to save more. (5) Individuals with a targeted saving expectation or goal are more likely to save than those without such goals. (6) Restrictions (i.e., match caps) and withdrawal penalties are also likely to positively affect savings behavior. Rules regarding the use of and access to savings (similar to rules restricting access to 401(k) funds until retirement) tend to act as controls against the use of savings for reasons other than those initially desired. For example, without restrictions a family member could request funds for short-term consumption rather than allowing the savings to accumulate for future investment in a house. (7) Finally, having a safe place to accumulate money can increase the feeling of security and comfort regarding saving and positively influence saving behavior.

5 Asset Building in Rural Communities Grinstein-Weiss et al. 29 Asset Building and IDAs in Rural Communities Poverty in rural America is a persistent problem that has captured the attention of rural social scientists and policy makers for many years. The concern is driven largely by the fact that rural communities tend to be more disadvantaged than urban areas and are generally characterized by lower household income and higher poverty rates (Economic Research Service 2005). In 2003, the poverty rate was 14.2 percent for rural residents compared to 12.1 percent for urban residents (U.S. Census Bureau 2004). Children living in rural areas are considerably more likely to live in poverty (20.8%) than children living in urban areas (16.9%) (Lichter, Roscigno, and Condron 2003). Stark differences by family structure and race also exist. Poverty in families headed by a woman with no husband present in the home is higher in rural communities (34.1%) than in urban areas (26.8%). Minorities in rural areas also have higher poverty rates than their urban counterparts. Moreover, there are racial disparities within rural areas where the black and Hispanic poverty rate (30.2% and 25.4%, respectively) is more than twice that of the white poverty rate (12.5%) (Economic Research Service 2005). Education is also central to understanding poverty in rural communities. Those with a bachelor s degree or more experience have far lower rates of poverty (3.5%) than do those with less than a high school degree (22.2%) (Jensen, McLaughlin, and Slack 2003). The characteristics of rural America today suggest a compelling need for policies that help families build wealth and economic selfsufficiency. Individual Development Accounts is one such policy. IDAs may be a particular useful way to build capacity, reduce poverty, and improve long-term welfare among poor and working poor rural residents for several reasons. First, poor rural families are less likely to receive cash assistance and tend to rely more heavily on earnings and more on jobs in the informal economy than other families (Brown and Lichter 2004; Rural Sociological Society 2006). With higher than average poverty rates and less cash assistance, participation in a matched saving program may afford rural families an opportunity to improve their human capital and economic well-being that would otherwise be unavailable. Second, the structural characteristics of rural areas make competition for good jobs and external investments difficult to attract and retain. Compared to urban areas, rural communities have fewer economic opportunities, lower earnings, fewer high quality jobs, and fewer educational and training opportunities (Zimmerman and Hirschl 2003). Communities need assistance in building human, social, and

6 30 Rural Sociology, Vol. 72, No. 1, March 2007 financial resources in order to develop structural and institutional opportunities (Jensen, McLaughlin, and Slack 2003) and to help families save. IDAs have the potential to do this and specifically address the policy needs of rural communities. Moreover, nonagricultural rural policy has been described as suffering from institutional neglect (Brown 2001). Swanson and Brown (2003) suggest that rural policy is sometimes misguided and developed with too narrow a focus because of overpowering agricultural interests and policy assumptions based on urban and suburban conditions. They call for a revamping of nonagricultural institutions (e.g., social welfare and education) in order to generate appropriate policy that fits the needs of a diverse rural America. One answer [to the problem of institutional neglect] is to formulate policies that enhance the capacity of rural people and communities to help themselves. The assumption underlying this approach is not that people will make a positive difference in their local communities if given the tools; rather, the assumption is that they should be given real opportunities to do so. (Swanson and Brown 2003:399) IDAs represent one such opportunity. Sherraden et al. (2003) suggest that the use of IDAs in rural communities may have several challenges however. Lack of accessibility and convenience may make IDAs less user-friendly and less attractive to individuals in rural areas as compared to those in urban areas. For example, population dispersion could lead to infrequent and inconsistent financial education and peer-mentoring meetings, leaving participants mostly on their own to gather important saving information. Even when participants are able to travel the distances required to attend classes or meetings, a second challenge is the higher transaction cost in time and money (e.g., more gas, time away from work, and automobile maintenance) associated with lengthy trips. The lack of a strong economic infrastructure in many rural areas is an additional possible disadvantage. Rural areas with few resources may have more difficulty providing strong operational program support, well informed program staff, complementary financial services, and adequate matching funds. Despite their potential challenges, IDAs may serve as an important mechanism for economic support in rural communities by cultivating capacity building through investments in both infrastructure and residents. IDAs, therefore, may positively affect asset-building and longterm economic self-sufficiency in rural areas.

7 Asset Building in Rural Communities Grinstein-Weiss et al. 31 The purpose of this study is to examine the performance of rural participants in IDA saving programs. Today, for rural policy to be effective, it has to address the diverse needs of the communities and their residents by investing in the people, infrastructure, and economy. Asset building, through increases in tangible and financial assets, is one policy option that could address the needs of many rural areas and could benefit both individual households and whole communities. This study seeks to answer the following research questions: (a) What are the individual characteristics associated with saving outcomes among rural IDA participants? (b) What are the program characteristics associated with saving outcomes among rural IDA participants? (c) What are the program and policy implications for supporting assetbuilding in rural areas? Program and Individual Characteristics Previous research and theory suggest that both institutional and individual factors are associated with saving. As previously discussed, the institutional model of saving posits that institutional factors other than flows of income and personal preferences affect saving behavior (Sherraden et al. 2003). This model suggests that without institutional incentives and opportunities, poor households will save less (Sherraden 1991). As such, we hypothesize that five institutional features (measured in terms of program characteristics) may be associated with saving level and the frequency with which participants make deposits. These include direct deposit (as a measure of facilitation), the match rate (as a measure of incentive), monthly saving target (as a measure of expectation), and financial education and peer group meetings (as a measure of information). Individual factors associated with saving include demographic and financial characteristics. Age, gender, marital status, race, education, employment, and household composition have been found to be associated with saving performance. According to the life-cycle model, the young and the old do worse in terms of wealth and home ownership than do middle-aged earners (Modigliani and Ando 1957; Wolff 2001). Family structure also affects wealth and asset accumulation. Cohabiting individuals and single-mother families, for example, are generally at a disadvantage when it comes to long-term wealth accumulation and rarely generate similar levels of wealth compared to that of married couples (Hao 1996; Wilmoth and Koso 2002). Married individuals in a study of saving performance among low-income participants in an

8 32 Rural Sociology, Vol. 72, No. 1, March 2007 IDA savings program had higher levels of home ownership and car ownership than unmarried participants by as much as 20 percent (Grinstein-Weiss, Zhan, and Sherraden 2006). Studies have also shown that minorities have less mean wealth than whites by as much as 20 percent (Wolff 2001), and blacks have been found to have lower saving levels than whites (Grinstein-Weiss and Sherraden 2006). Higher levels of education are associated with increased savings and wealth (Bernheim and Garrett 1996; Zhan and Pandey 2004), as is employment status (Beverly 1997). Childless families are also found to be wealthier than families with children (Grinstein-Weiss, Wagner, and Ssewamala 2006; Wolff 2001). Financial characteristics, such as income and asset ownership, are also associated with saving and wealth. We include measures of total income (Acs and Danzinger 1993) and various forms of asset ownership and participation, such as owning a home or a car and having a bank account. Methods Data and Sample The data come from the American Dream Policy Demonstration (ADD), the first large-scale test of IDAs designed to study the merits of IDAs as a community development and public policy tool. Beginning in 1997, ADD research followed more than 2,000 participants at 14 community-based IDA program sites across the United States for four years ( ). IDA programs in ADD were operating in community-based organizations that were working together with financial institutions. The organizations in ADD are a diverse group of community development corporations, social service agencies, and for-profit and not-for profit organizations (see a detailed description by Schreiner et al. 2001). IDA savings are used for specific purposes, usually home purchase, post-secondary education, and microenterprise. The accounts are similar to other defined contribution plans such as 401(k) retirement plans. IDAs offer a monetary incentive for participation by matching dollars deposited by participants. The level of matching in IDAs differs by program and sometimes even within programs based on the goal of the savings (i.e., homeownership, education, and microenterprise). In most cases, the matches are either 1:1, 2:1, or 3:1. Participants in the programs enroll in mandatory financial education classes. Participants work with a case worker and develop peer relationships with other program participants during peer group meetings.

9 Asset Building in Rural Communities Grinstein-Weiss et al. 33 ADD employed a multi-method research design to gather information on many aspects of IDA programs and participants. This study used data created from monitoring all the participants savings transactions. Program staff collected both program and participant data with the Management Information System for Individual Development Accounts (MIS IDA). The data were checked for data entry errors, outliers, missing cases, and inconsistencies using the MIS IDA quality control software. This may be the best available data set on savings patterns among low-income families to date (Sherraden 2002). The MIS IDA data are complemented by an additional data set gathered from the 14 ADD sites through a program survey. The survey data were collected using face-to-face and telephone interviews with administrative personnel at the 14 ADD sites. The survey instrument was designed based on constructs offered by institutional theory (Ssewamala and Sherraden 2004). Participants in this study consist of the 315 IDA enrollees from rural areas only, including those who dropped out of the program without a matched withdrawal. Since the sample was restricted to the 315 rural participants, only the 8 programs in which they participated were analyzed. The regression analyses use the participants characteristics that were recorded at time of enrollment to avoid issues of two-way causation between income and savings. Measurement Dependent variables. Two dependent variables, the Average Monthly Net Deposit (AMND) and deposit frequency, are used to measure savings in IDA programs in order to capture two major aspects of savings: amount and regularity. These variables were constructed and used in previous reports on ADD programs (Schreiner, Clancy, and Sherraden 2002). AMND is defined as net deposits per month and is calculated as deposits plus interest minus unmatched withdrawals, divided by the number of months of participation. Thus, AMND controls for the length of participation in the program. The variable net deposits that is used to calculate AMND is defined as deposits plus interest (net of fees) minus unmatched withdrawals. Net deposits include matched withdrawals, but exclude deposits in excess of the match cap (maximum amount that can be matched) or after the time cap. Excess deposits, late deposits, and unmatched withdrawals are savings in IDA accounts, but they cannot be matched and therefore are not considered net deposits. AMND is the key measure of savings outcomes in this study because greater AMND implies greater savings

10 34 Rural Sociology, Vol. 72, No. 1, March 2007 Table 1. Descriptive Statistics for the Dependent Variables (N5247) Dependent Variables Mean SD Range AMND Deposit Frequency and asset accumulation (Schreiner et al. 2001). As seen in Table 1, the mean AMND in this study was $19.30 (SD 18.29), with a high and low of $2.52 and $76.92 respectively. Deposit frequency is defined as the number of months with a deposit divided by the number of months of participation. Deposit frequency indicates how regularly a participant saves. This variable is an important indicator of whether rural IDA participants are acquiring the habit of saving. It may even be an indicator of whether participants continue to save after graduating from the program. The mean deposit frequency for this study was 48 percent (SD.27). Individual control variables. The independent variables include participant and program characteristics. Participant demographics include gender (1 5 female, 0 5 male); age (in years); a set of dummies that measure marital status: single, divorced/separated and married (reference); number of children (under 18 years); and number of adults (18 years and older) in the household. We also include a set of dummy variables for race: African American, Latino/ Hispanic, other ethnicity, or Caucasian (reference). Another set of dummies measure educational attainment: no high-school diploma (reference), high-school graduate, some college, and college graduate. Finally, employment status is measured by whether the participant was employed full-time ($ 35 hours per week), part-time (, 35 hours per week), was unemployed (reference), or was a working student. Participants financial characteristics include monthly household income; car ownership (1 5 yes, 0 5 no); home ownership (1 5 yes, 0 5 no); and having either a checking or savings account (1 5 yes, 0 5 no). For the purpose of interpretation, we divide the household income by 100 for the regression analyses. Program variables. Direct deposit, match rate, financial education, monthly savings target, and peer group meetings are the program variables included in the analysis. While all of these variables capture program characteristics, they are subject to the rules and restrictions set by the different programs. All of the variables (with the exception of peer group meetings) are measured at the individual level. The first program variable is direct deposit. This variable measures whether or not an individual is using direct deposit (1 5 yes, 0 5 no). The second program variable, match rate, is measured with four dummy variables

11 Asset Building in Rural Communities Grinstein-Weiss et al. 35 that account for different match rates the participants received: 1:1 (reference), 2:1, 3:1, and 4:1 to 7:1. While match rate is a program characteristic, it is measured at the individual level; individual programs usually offered different match rates to different participants based on the purpose of their savings. For example, program 1 offers a match rate of 1:1 to people who save for education and 2:1 to people who save for homeownership. Hours of financial education are also measured at the individual level. These mandatory financial education classes included material on financial management and saving strategies, and incorporated information on how to create a budget, how to manage money, and how to fix and establish credit records. We include a measure of general financial education, which depicts the number of financial education hours a participant has taken. Each program in the study had the flexibility to determine its own requirements resulting in different levels of service utilization both across and within programs. The fourth program variable, monthly savings target, is the total match cap (or the limit on the amount of deposits that can be matched), divided by the time cap (or the number of months after opening an account in which a participant may make matchable deposits). This variable, also measured at the individual level, is adjusted for each participant because of variation in the duration of enrollment across programs. The final program variable, the peer group meeting, indicates whether programs offered informal peer group meetings to IDA participants. This is the only variable that is measured at the program level (1 5 yes, 0 5 no). Analysis This study focuses on the experiences of IDA participants living in rural areas (n5315). Due to missing cases primarily on the direct deposit and household income variables, the sample was reduced by 69 cases resulting in a final analytic sample of n5247. Household income information is a sensitive issue especially among the low-income population and as such is harder to obtain. Data on direct deposit was collected in a later phase, so it was harder to get full responses. In the analysis phase, descriptive statistics are produced to characterize the sample. Then, in order to answer the first question, What individual characteristics are associated with saving performance for rural IDA participants? and the second question, What program characteristics are associated with saving performance for this group? an Ordinary Least Squares (OLS) regression analysis is conducted.

12 36 Rural Sociology, Vol. 72, No. 1, March 2007 Table 2. Descriptive Statistics for Individual Characteristics of the Sample (N5247) Independent Variables Mean SD Range Gender (1 5 Female) Age Marital Status Single Divorce/Separated/Widowed Married Household Composition Number of Children Number of Adults Race/ Ethnicity African American Latino/ Hispanic Other Ethnicity Caucasian Education No high school Completed high school Attended some college Graduated from college Employment Unemployed Working student Employed part-time Employed full-time Monthly household income 1, ,400 Asset Ownership Home ownership Car ownership Bank account Results Individual Characteristics As seen in Table 2, most of the participants were female (79%). Ages ranged from 13 to 70 years, with a mean age of 34 years, and a standard deviation of The majority of the participants were single (40%), 32 percent were divorced, separated, or widowed, and 28 percent were married. The average number of children in the household was 1.8, and the average number of adults in the household was 1.6. The majority of the participants were Caucasian (81%), 13 percent were African American, 1 percent were Latino, and 6 percent were from another ethnic group.

13 Asset Building in Rural Communities Grinstein-Weiss et al. 37 Table 3. Descriptive Statistics for Program Characteristics of the Sample (N5247) Independent Variables Mean SD Range Direct Deposit Match Rate 1: : : :1 to 7: Financial Education Monthly Saving Target Peer Group Meetings Approximately 20 percent of the participants did not complete high school, 28 percent had a high school degree, 30 percent attended some college but did not graduate, and 22 percent had a college degree (either 2 year or 4 year). Fifty-two percent of the sample was employed full-time (35 hours per week or more), while 30 percent worked parttime. Nine percent were unemployed or not working, and 9 percent were students. The mean monthly household income was $1,260, and the median was $1,200. In annual terms, the average income was $15,120 a year. The majority (64%) of the participants lived below 100 percent of the poverty level. Twenty five percent of the participants were between 100 percent and 150 percent of the poverty level, and nine percent lived between 151 percent and 200 percent of the poverty level. Only three participants earned more than 200 percent of the poverty level. There were no significant differences in poverty level by marital status (F 5 373, p 5.689). The number of children in the household is negatively related to the poverty level (r 52.25, p 5.000). The majority (81%) of the rural participants had either a checking or savings account (other than their IDA). Thirty-eight percent owned a home, and 78 percent owned a car. Program Characteristics As seen in Table 3, only 6 percent of the participants had direct deposit. Eighteen percent of the rural participants had a match rate of 1:1. Twenty-eight percent had a match rate of 2:1, another 32 percent had a 3:1 match rate, and 19 percent had a match rate between 4:1 and 7:1.

14 38 Rural Sociology, Vol. 72, No. 1, March 2007 IDA participants were required to attend free financial education and asset-specific classes as part of the program. Rural IDA participants received, on average, 13 hours of general financial education. Monthly savings target is defined as the amount which, if saved each month and not removed in unmatched withdrawals, will be matched. The average monthly saving target per participant was $32.87, meaning that each participant had to save on average $33 a month in order to take full advantage of the match. Almost half of the programs (49%) offered peer group meetings. Saving Outcomes AMND is defined as net deposits per month of participation. As seen in Table 1, the AMND was $19.30 (SD ) with a high of $76.92 and a low of $2.52. This variable is slightly positively skewed (skewness ). This skewed distribution is probably due to a few cases with AMND of greater than $50. These cases were spread across several programs. Most of them were participants who joined the program late. It is likely that they made larger than average deposits in attempting to meet the saving goal in a relatively short time. Twenty nine percent of the participants had zero savings. This number captures participants who dropped out of the program without a matched withdrawal. Deposit frequency is defined as the number of months with a deposit divided by the number of months of participation. The mean deposit frequency for this group was 48 percent (SD 5.27). The typical IDA participant made a deposit in six of twelve months. Deposit frequency is considered symmetric (skewness ). Saving Performance of Rural Participants The results of the OLS regression analyses when AMND is regressed on individual characteristics and on measured program characteristics are significant [F(26, 220) , p 5.000] and explain approximately 40 percent of the variance in AMND (R , Adjusted R ). Likewise, significant results appear when deposit frequency was regressed on the individual characteristics and measured program characteristics [F(26, 220) , p 5.000] and explained approximately 39 percent of the variance in AMND (R , Adjusted R ) (see Table 4). The regression results indicate that two individual variables and all of the program variables are associated with savings performances for rural IDA participants. First, marital status is statistically associated with deposit frequency. Specifically, compared with married participants,

15 Asset Building in Rural Communities Grinstein-Weiss et al. 39 single participants are associated with 11 percentage points lower deposit frequency. Second, home ownership is associated with both AMND and deposit frequency for rural IDA participants. Specifically, rural participants who are homeowners are associated with an $8.21 higher AMND, and 9 percentage points higher deposit frequency than rural participants who are not homeowners. Turning to program characteristics, direct deposit is statistically associated with deposit frequency. Compared to participants who do not have direct deposit, having direct deposit is associated with a 23 percentage point increase in deposit frequency. Match rate is statistically associated with deposit frequency. IDA participants with a match rate of 3:1 are associated with 19 percentage points higher deposit frequency, and participants with a match rate of 4:1 to 7:1 are associated with 13 percentage points higher deposit frequency compared to the participants with a 1:1 match rate. Hours of financial education attended by IDA participants is statistically related to both AMND and deposit frequency. Each additional hour of financial education is associated with a $0.63 increase in AMND, and one percentage point increase in deposit frequency. Monthly saving target is significantly related to both AMND and deposit frequency. Each additional dollar in the monthly saving target is associated with a $0.23 increase in AMND. In addition, a dollar increase in the monthly saving target is associated with a one percentage point increase in deposit frequency. Finally, peer group meetings are statistically associated with the two measures of savings: AMND and deposit frequency. Participants in programs that offer peer group meetings in addition to regular financial education meetings are associated with a $16.53 higher AMND, and 9 percentage points higher deposit frequency compared with participants in programs that do not offer these additional peer group meetings. Limitations Some limitations of this study are important to note. First, the data analysis phase uses individual characteristics that were collected on the participants at time of enrollment in the IDA programs. Some individual characteristics may have changed during the time an individual spent in the program. This change might impact the saving outcome; however, these changes have not been recorded (Ssewamala 2003). Second, this study assumes that deposits in IDAs come from

16 40 Rural Sociology, Vol. 72, No. 1, March 2007 Table 4. Regression Analysis: Individual and Program Characteristics and Saving Performance AMND Deposit Frequency Individual Variables b S.E b S.E Gender Female (Male) Age Marital Status Single * 0.05 Divorce/Separated/Widowed (Married) Household Composition Number of Children Number of Adults Race/ Ethnicity African American Latino/ Hispanic Other Ethnicity (Caucasian) Education (No high school) Completed high school Attended some college Graduated from college Employment (Unemployed) Working student Employed part-time Employed full-time Household Income Asset Ownership Home ownership 8.21** * 0.04 Car ownership Bank account Direct Deposit *** 0.06 Match Rate (1:1) 2: : *** :1 to 7: * 0.06 Financial Education 0.63*** * Monthly Saving Target 0.23* *** Peer Group Meetings 16.53*** * 0.04

17 Asset Building in Rural Communities Grinstein-Weiss et al. 41 Table 4. Continued Individual Variables AMND Deposit Frequency b S.E b S.E R F N *p #.05; **p #.01; ***p #.001. new savings. However, it may be the case that some participants in IDAs are transferring money from other assets they have, and as a result, deposits are coming from assets that have been shifted and not from new savings (Schreiner et al. 2001; Zhan, Sherraden, and Schreiner 2002). This seems unlikely given participants are lowincome and do not have many assets to redirect. Finally, since the ADD data were not collected using randomized assignment techniques, one cannot equivocally attribute the significant findings to program participation. It is difficult to determine how the participants would have saved if they had not participated in IDAs. The experimental design in ADD will be able to test this; however, the data are not available yet. Discussion and Implications This study examines the unique experiences of low-income rural participants in a matched saving program IDA. IDAs provide institutional mechanisms to rural participants to save and accumulate assets and may promote social and economic development within rural communities. Results from this study indicate that low-income rural participants have the willingness and the ability to save toward the accumulation of assets when provided with institutional opportunities, such as those provided by IDA programs. Examination of the factors associated with successful performance of IDA participants in rural areas suggest that both individual characteristics and institutional characteristics (i.e., program characteristics) are important in explaining saving behaviors. Moreover, it seems that program characteristics may be more important than individual characteristics in explaining saving behaviors. While only two individual variables (marital status and home ownership) had association with saving outcomes, all five program characteristics (direct deposit, match rate, financial education, monthly saving target, and peer group meetings) were associated with saving outcomes.

18 42 Rural Sociology, Vol. 72, No. 1, March 2007 Home ownership appears to be an important predictor of savings among rural IDA participants. Home ownership may suggest that participants already have some experience with savings. In addition, this study finds that single participants are saving less frequently compared with married participants. These results are congruent with other studies that have examined the impact of marital status on savings and family wealth accumulation that found that marriage can enhance wealth accumulation (Hao 1996; Lupton and Smith 2003; Siegel 1993). For example, through the analysis of data from the Health and Retirement Survey and the Panel Study of Income Dynamics, Lupton and Smith (2003) find that married couples save significantly more than other household types. Additional ways to help single rural participants to save in IDAs should be explored. Turning to program characteristics, this study uses the institutional theory that claims that institutional (program) characteristics in addition to individual characteristics play an important role in explaining and promoting savings (Beverly and Sherraden 1999; Sherraden 1991; Sherraden et al. 2003). Five of the components of the institutional model suggested by Sherraden and Barr (2005) were found to be important predictors of saving outcomes among rural participants. Specifically, information measured by financial education and peer group meetings, incentives measured by match rate, facilitation measured by direct deposit, and expectations measured by monthly saving target were all important predictors of program success. These results support the argument that institutions have an important role in shaping savings behavior and may explain at least a part of the variance in personal savings, thus implying that policies and program design can have a positive effect on savings among low-income rural participants. Financial education classes in IDAs teach low-income participants different aspects of financial literacy such as how to establish credit, how to cut down on expenses, how to set goals for saving, and how to overcome barriers to saving. This study suggests that financial education is an important predictor of saving performances for rural IDA participants. Receiving more hours of financial education is associated with higher savings. It might be the case that IDA classes are not only successful because of the class content, but also due to additional aspects of the workshops. For example, classes are taught by people from local community organizations who can relate on a more personal basis and have a better understanding of the population they are teaching. Another aspect is the sharing of experiences that may take place among the participants. Finally,

19 Asset Building in Rural Communities Grinstein-Weiss et al. 43 counseling is provided, and empowerment and self-sufficiency are constantly reinforced. Direct deposit use in the program is associated with more frequent deposits among rural IDA participants. This result supports the proposition suggested by the institutional theory that argues that individuals who receive some kind of saving facilitation which makes saving more manageable and convenient will increase their willingness to save (Beverly and Sherraden 1999; Sherraden et al. 2003). By utilizing direct deposit and transferring money from one account into another, individuals are more likely to save and less likely to use the money for consumption (Beverly and Sherraden 1999). However, only 6 percent of the rural participants have direct deposit. Based on these results, program administrators should not only make direct deposit available to account holders, but they should also encourage participants to use it. Match rate is also associated with more frequent deposits. Rural IDA participants with a match rate of 3:1 saved more frequently than those with a match rate of 1:1. Similarly, rural IDA participants with a match rate of 4:1 to 7:1 saved more frequently than those with a match rate of 1:1. A match rate of 3:1 is associated with higher increase in deposit frequency than a match rate of 4:1 to 7:1. It might be that a match rate of 3:1 is the optimal match rate for this group, but further studies are needed to verify this conclusion. Match rate, however, is not associated with the savings amount. This may indicate that higher match rates may encourage people to save more frequently, but might not affect the amount they save. Monthly savings target is defined as the amount that, if saved each month and not removed in unmatched withdrawals, will be matched. The monthly saving target is viewed in this study as a measure of expectation of the IDA programs from their participants regarding the saving amount. The results suggest that the monthly saving target is an important predictor of saving performance and that a higher monthly saving target leads to higher and more frequent savings. Therefore, it is suggested that program administrators should raise the limits on matchable deposits for rural IDA participants. Finally, peer group meetings, other than the financial education meeting, are another way for rural participants to share information, tips, encouragement, and ideas among themselves. This study finds that peer group meetings appear to be an important predictor of savings performance for this group. Therefore, it is suggested that more IDA programs establish peer groups meetings. This study suggests that IDAs may be an effective tool to help lowincome people in rural areas save and accumulate assets. Many rural

20 44 Rural Sociology, Vol. 72, No. 1, March 2007 areas today are facing economic hardships and are struggling to survive. In order to be successful, these communities need to identify their common resources and create new approaches that will encourage social and economic development. Current public policy does not properly address the needs or the challenges rural regions are experiencing. One way future policy can help promote rural growth and expansion is through more asset-based initiatives focusing on local areas. IDAs may be used as one piece of a larger plan that could help support and advance asset building as a social and economic catalyst for many rural regions. References Acs, G. and S. Danzinger Educational Attainment, Industrial Structure, and Male Earnings through the 1980s. Journal of Human Resources 28: Becker, G Human Capital. New York: Bureau of Economic Research. Bernheim, B.D. and D. Garrett The Determinants and Consequences of Financial Education in the Workplace: Evidence from a Survey of Households. Cambridge, MA: National Bureau of Economics. Beverly, S How Can the Poor Save? Theory and Evidence on Saving in Low-Income Households. Center for Social Development, Washington University, St. Louis. Beverly, S. and M. Sherraden Institutional Determinants of Saving: Implications for Low-Income Households and Public Policy. Journal of Socio-Economics 28: Brown, J.B. and D.T. Lichter Poverty, Welfare, and the Livelihood Strategies of Nonmetropolitan Single Mothers. Rural Sociology 69: Brown, W.P The Failure of National Rural Policy: Institutions and Interests. Washington, D.C.: Georgetown University Press. Curley, J. and M. Grinstein-Weiss A Comparative Analysis of Rural and Urban Saving Performance in Individual Development Accounts. Social Development Issues 25: Dorward, A.R., S. Anderson, S. Clark, B. Keane, and J. Moguel Asset Functions and Livelihood Strategies: A Framework for Pro-Poor Analysis, Policy and Practice. Presented at the 74th EAAE Seminar on Livelihoods and Rural Poverty. Imperial College, September 2001, Wye, UK. Economic Research Service Rural America at a Glance, Retrieved May 15, 2006 ( U.S. Department of Agriculture.. n.d. A History of American Agriculture Retrieved September 24, 2002 ( Falk, W.W. and L.M. Lobao Lessons from the Past, Challenges for the Future. Pp in Challenges for Rural America in the Twenty-First Century, edited by D.L. Brown and L.E. Swanson. University Park, PA: Pennsylvania State University Press. Grinstein-Weiss, M. and M. Sherraden Racial Differences in Performances in a Matched Saving Program. Journal of Income Distribution 13: Grinstein-Weiss, M., K. Wagner, and F. Ssewamala Saving and Asset Accumulation among Low-income Families with Children. Children and Youth Services Review 28: Grinstein-Weiss, M., M. Zhan, and M. Sherraden Saving Performance in Individual Development Accounts: Does Marital Status Matter? Journal of Marriage and Family 68: Hao, L Family Structure, Private Transfers, and the Economic Well-Being of Families with Children. Social Forces 75: