Duration: 10/01/2003 to 09/30/2008

Administrative Advisor(s):

NIFA Reps:

Non-Technical Summary

Statement of Issues and Justification

Between 1980 and 2001, personal savings as a percent of disposable personal income fell from 10.2 to 2.3 (Office of Management and Budget, 2004). This relatively rapid decline is cause for serious concern. In recent years, doubt has arisen regarding the ability of the Social Security trust fund to pay promised benefits to the soon-to-retire Baby Boom generation. Employers and government have shifted greater responsibility for funding such things as health care and retirement to individual households. Increasingly, households that lack adequate savings can find it difficult, if not impossible, to achieve and maintain long-term financial stability. Without a financial cushion, households have little protection against the adverse effects of income loss due to unemployment, long-term illness, or the disability or death of a primary income earner (Schuchardt, 2002) and may have to rely on extended family or various forms of public assistance to survive. Insufficient savings can also have adverse consequences for the broader economic community. Home or business ownership, important elements in the economic vitality of local communities, are difficult to achieve without savings (Schaeffer, 2002). In times of economic downturn, loan default or bankruptcy become more likely among those who have not been savers, shifting the burden of economic loss to the community.


Given these concerns, motivating people to save if they are currently a non-saver or to increase their level of saving if they are a saver is an important objective. To meet this objective, it is essential to understand the specific impediments to saving and to ascertain the degree to which these impediments can be removed. Different barriers call for different intervention strategies. For example, a person unaware of the benefits of saving in an Individual Retirement Account (IRA) may be persuaded to open and use such an account after receiving financial education regarding IRAs. Alternatively, tax law changes might be needed to provide low-income individuals with a residual of income to save.


While several different social science disciplines have examined factors associated with both savings behavior and level of savings accumulated, existing studies are constrained by the specific assumptions and focus of the discipline within which they were conducted. Sociological studies of money management have centered on demographic characteristics such as differences in age, gender, family composition or social class (Lea, et al, 1987; Livingstone & Lunt, 1993), ignoring economic or psychological aspects of saving. Economists have focused on observed behavior, ignoring or addressing in only a cursory way internal motivations for savings. Further, economists generally classify savings behavior as a simple dichotomy: one either is or is not a saver, precluding consideration of a systematic, sequential decision process by which people may journey from a position of non-saver to saver. Psychologists consider attitudes and values, but typically ignore sociodemographic, economic, or public policy circumstances and constraints (Livingstone & Lunt, 1993). These disciplinary blinders make it difficult to gain complete understanding of factors affecting saving behavior. For example, economic theory suggests that individuals will not simultaneously be savers and dissavers (spending down financial reserves or borrowing money) and that saving is an activity undertaken by those in midlife, the prime income earning years, rather than the young or elderly and by those with high rather than low income. The fact that research findings contradict these expectations suggests that non-economic factors also motivate savings behavior. While psychological theory may point to some of these factors, it does not incorporate very real economic or legal constraints on or incentives for savings behavior. Consequently, there is a need for interdisciplinary research that recognizes that saving is an economic decision made within an existing social context, influenced by life cycle demands and the psychological characteristics of the potential saver.


This proposed research seeks to fill that gap. Unique contributions of this proposed cross-sectional study are characterizing saving behavior as a continuum and developing an index to measure that continuum and the use of both economic and psychological theory to examine the impact of both personality and financial resources as barriers to becoming a saver or to saving more. The impact of lifecycle stage on saving behavior and level of savings will be addressed through selection of the sample. The economic factors that will be investigated in this project include level of financial literacy, income, wealth, time preference, debt load, and bequest motive. The psychological factors that will be considered in relation to savings behavior includes self-control, self-efficacy, perfectionism, impulsivity, and materialism. Differences in demographic factors such as family size or in opportunity to save such as access to employer-sponsored retirement accounts will be controlled.

Related, Current and Previous Work

This study will not duplicate research and programs that currently exist. A search of CRIS indicated that no other research project specifically examines the psychological determinants of savings behavior. There are two projects that examine predictors of retirement savings: Retirement Planning: Macro-Environmental and Life Situation Factors, from the University of Missouri (PI: D. Sharpe) and Retirement Economic Well-Being of Women in Nebraska and Cross-Culturally, from the University of Nebraska (PI: S. Cramer). In contrast to these projects, this study would include micro-level psychological as well as economic factors that could influence saving behavior. This study would also encompass a broader array of savings motives than retirement and be broader in scope than a single state or region.


This research proposal was developed in consultation with numerous stakeholders, including university researchers, extension educators in family and consumer science, practioners, policy analysists, and advocates. The work will focus on a timely topic, using an interdisciplinary approach that will have implications for improving individual and community quality of life. Conducting the research as a multi-state project is beneficial. The involvement of researchers and extension educators in many states fosters development of nationwide partnerships promoting savings behavior. Further, this study relates directly to Community Vitality, a national research priority of the Agricultural Experiment Station Committee on Organization and Policy. Families that are financially secure are better able to contribute to their local economy by owning homes, starting businesses, and avoiding bankruptcy. In this respect, family financial security is a cornerstone of community economic vitality.


Review of Relevant Literature


Defining savings


In 1936, Keynes defined savings as simply the excess of income over expenditure (p. 62). Katona (1975) later classified this type of saving as residual saving and proposed expanding the concept of savings to include contractual savings  acquiring assets such as a home or durable goods via making contractual payments (e.g. mortgage payments or installment payments) out of current income  and discretionary savings  purposeful additions to deposit accounts or purchase of investment assets such as mutual funds, bonds, or stock. Wdrneryd (1999) differentiated saving, saving behavior, and savings. The act or process of setting funds aside is saving. Saving behavior focuses on the behavioral aspects of that act or process. Savings (a noun - a fund or an acquired asset) is the result of saving (a verb - an action taken).


Characteristics of savers and non-savers


A lack of psychological and attitudinal variables in the data sets used for analysis have typically constrained focus on economic and demographic variables. In general, savers are likely to have higher levels of education and income as compared with non-savers. Savers are also more likely to be white, married, homeowners and have relatively fewer children (DeVaney & Chien, 2001; Hogarth, 1991; Hogarth & Anguelov, 2001).
Following Wdrneryd (1999), DeVaney and Chien (2001) included past saving behavior in a planned behavior model. They examined attitude (risk tolerance), subjective norms (age, ethnicity, and marital status), perceived control (education, professional status, income, household size, and self-employment), and past savings (home ownership, amount of assets) as predictors of the amount saved in two types of retirement accounts: defined contribution accounts and Individual Retirement Accounts. They found that being more risk tolerant, having more education, being married, and being a homeowner was positively related to each of the retirement savings accounts.
Hogarth and Anguelov (2001) use respondent answers to survey questions to define current savers as those who spent less than income during the past year and usual savers as those who said that they either saved regularly, saved what is left over, or saved from other income. They found that 60% of households at or below poverty and 70% of households between 101 and 150% of the poverty level indicated that they saved using the definitions of current or usual savers. Their results indicated that as income and net worth increased, the likelihood of saving increased. Households who were rejected for credit or those who did not get as much credit as they applied for were less likely to be savers. Research regarding Individual Development Accounts, savings accounts for low-income individuals where dollars saved were matched according to certain guidelines, also indicates that low levels of income, per se, does not preclude saving as a substantial portion of the low-income group studied were successful in setting at least some amount of current income aside (Beverly, 1997).
Savers seem to have different attitudes and actions than non-savers. Lunt and Livingstone (1991) found that savers generally had an internal locus of control, valued hard work and getting respect. Non-savers, in contrast, tended to see themselves as victims of circumstance. Savers bought durable goods, used secondary markets and spent relatively less than non-savers on goods that were immediately consumed.



Factors affecting savings behavior  considerations from economic theory and research


The life cycle hypothesis of savings (LCH) (Ando & Modigliani, 1963) is a widely cited economic theory of saving that proposes that household savings and consumption behavior reflects the life cycle stage of the household. Following the permanent income hypothesis, the LCH assumes that household consumption is a linear function of the cash available now and the discounted value of future income (Bernheim, Shleifer & Summers, 1986). Decreasing marginal utility of income implies a utility maximizing strategy of consumption smoothing, where individuals estimate their future income and consume about the same each year during their lives. Given an income path where income increases during working years and drops at retirement and consumption smoothing, households will tend to borrow while young, save heavily during middle age, and spend down savings during retirement. Empirical results that conflict with this theory have been found, however, such as saving activity by younger households and little evidence of spending down by older households (Wdrneryd, 1999).
The LCH suggests that resource level is an important determinant of savings choice. Resources can be broadly considered as including such things as lifetime income/human wealth or annual income, net worth, liquidity of wealth. Much of the literature relating to savings behaviors has focused on allocation decisions (Zhong & Xiao, 1995; Gutter, Fox & Montalto, 1999) or on retirement (Yuh & DeVaney, 1996; Grable & Lytton, 1997). Studies of this type have generally found a positive relationship between the amount of income and wealth and stock ownership or investments in pensions. Considering the relationship between debt levels and participation in and level of discretionary retirement accounts, Cavanagh and Sharpe (2002) found evidence to suggest that those with credit obligations make relatively lower contributions to this particular type of savings.


Other studies have examined more general measures of savings beyond specific asset types or funds designated for retirement. Avery and Kennickell (1991) used panel data from the 1983 and 1986 Survey of Consumer Finances (SCF) to examine changes in household total non-pension wealth accounting for household changes occurring between survey waves. They show that the most influential predictor of savings was household income, with weaker predictability of savings through demographics.


Chang (1994) used data from the 1983 and 1986 SCF to examine patterns of savings behavior defined as increasing net non-housing assets. Chang found that households with higher income were likely to exhibit greater savings than households with less income. In addition, Chang also adds that households receiving windfalls in 1983 saved a significant portion of these amounts.
Also using SCF data, Rha, Montalto, and Hanna (2003) defined savings as spending less than income and incorporated savings goals as a means to control for self-regulation. They found that households with greater annual income and financial assets and homeowners were much more likely to be savers while households with consumer debt are less likely to be savers. This finding is consistent with the notion that households accumulating consumer debt are more concerned about current consumption and therefore less likely to save.


Factors affecting savings behavior  considerations from psychological theory and research


Katona (1975) proposed that savings was the result of the individuals ability to save (having income sufficient to save) and the individuals willingness to save (the ability to defer gratification for future benefit). Beverly (1997) reviewed previous research and theories that link economic factors, low-income households, and the possibility of saving. Like Katona (1975), she noted that both the ability to save and the willingness to save are important determinants of saving. Her review of literature suggested that low-income individuals may be willing to save, but lack ability to do so. A minimum level of consumption is necessary for survival. Households with incomes below this minimum cannot afford to save because survival needs cannot be deferred. Sherraden (1991) has also pointed out that low-income households have less access than middle and high income households to savings subsidies such as a tax deduction for a home mortgage or an employer match on a tax-deferred retirement account. Low-income households are also less likely to have access to payroll deductions, a strategy that many find useful for reducing the temptation to spend money earned. Clearly, economic resources and expenditures can be potential barriers to saving. Still, individuals with similar levels of economic resources and expenses may vary greatly in their savings behavior. Psychological variables can play a critical role in explaining these individual differences.


To date, little research has directly linked psychological concepts and savings behavior. Researchers in psychology and consumer behavior, however, have explored relationships between psychological concepts and various other economic behaviors. Psychological factors such as self-control, self-efficacy, perfectionism, impulsivity, and materialism have been found to influence other consumer actions and may also be important in savings behaviors.
Psychologists often use the terms self-control and self-regulation interchangeably; both refer to the individuals capacity to alter his or her own states and responses. Recent psychological studies have described self-regulation as a limited resource available to individuals (Vohs & Heartherton, 2000). For consumer behavior, self-control represents the capacity to resist temptations, especially those relevant to expenditures that are likely to be regretted later on (Faber and Vohs, in press; Baumeister, 2002). Impulse buying and compulsive buying have been linked to different types of failures in self-regulation (Faber & Vohs, in press). In a series of studies, Vohs and Faber (2002) found a negative relationship between self-control and impulse buying. That is, participants with depleted self-regulatory resources were willing to pay more for goods and to actually spend more money and buy a larger number of goods impulsively as compared with others.


Economists Shefrin and Thaler (1988) modified the Life Cycle Hypothesis of savings to include the non-economic variable of self-control. They describe savings and consumption as a function of the individuals ability to exert self-control to limit present consumption. Failures in self-regulation due to conflicting goals, poor monitoring or resource depletion may explain why some people are unable to save while others with similar financial resources are successful. In one of the few studies directly relating psychological variables to savings, Romal and Kaplan (1995) found a significant association between scores on a self-control scale and personal savings.


Impulsivity is the personality trait of acting uncontrolled versus controlled. People at the control end are cautious and sensible; they postpone immediate gratification. Impulsive people on the other hand are spontaneous, reckless and careless. They tend to be less future oriented and therefore tend more toward instant gratification. Those who are more impulsive are likely to spend more and have a harder time saving. Prior research has shown that more impulsive people engage more in impulse buying behavior (Youn & Faber, 2002).


We would, therefore, expect that impulsivity will be related to lesser savings.
Self-efficacy refers to the belief that an individual can accomplish goals and succeed at the behaviors they attempt (Bandura, 1977; Sherer, et al., 1982). It is seen as a pivotal construct that mediates the application of knowledge and skills in the pursuit of behavioral goals (Maibach & Murphy, 1995). People avoid tasks that they feel are beyond their abilities (Bandura, 1977). Self-efficacy has been found to be an important variable in explaining the adoption of illness prevention or health promotion behaviors (Bandura 1997). Indeed, this construct has been shown to be important in both the initiation and maintenance of behavioral change (Bandura 1992). Self-efficacy is likely to also be important in explaining peoples willingness to engage in savings behaviors. Individuals may want to save, but fail to do so because they believe they will not be successful in this effort.


Self-efficacy is often seen to be similar to locus of control. Locus of control refers to the generalized expectation of where control over a persons life resides. It can be seen as either internal to the individual or external. Self-efficacy, on the other hand, refers to the extent to which people feel capable of undertaking specific behaviors that influence their lives. Research in health behaviors have shown that self-efficacy is much more predictive of health behaviors than is locus of control (Barrios, 1985; Maibach & Murphy, 1995).


Perfectionism refers to the desire to set and achieve high standards for oneself. There are several measures of perfectionism that have been used to measure personal levels of achievement, and to measure general tendencies to set high goals for oneself and to try to live up to the high goals of others. Failure to meet these high goals often leads to problem behaviors. For example, perfectionism has been linked to bulimic symptoms and binge eating patterns (Vohs et al., 1999; 2001). Perfectionism has also been found to be a characteristic of compulsive buyers (Faber, 2000; OGuinn & Faber, 1989). Perfectionism may be related to a failure to save. People may set too high a goal and when they are unable to meet it, they give up the attempt or spend money in an effort at self-soothing.


Materialism refers to the beliefs held by individuals about the importance of material goods in their lives. Richins and Dawson (1992) have attempted to measure the extent to which individuals believe that acquiring possessions is central to their lives, an important route to happiness, and as necessary to live a successful life. Family structure is an element that appears to influence material values. For example, Rindfleisch, Burroughs, and Denton (1997) found that young adults who had grown up in disrupted households were more likely to be materialistic than those from intact families. Since materialism can affect savings behavior, this variable may help to explain differences in savings by household or family composition.


Linking economic and psychological theory


Variation in individual rates of temporal discounting, or the opportunity cost of trading present for future utility, may represent a bridge between the economic concept of utility maximization and psychological constructs measuring impulsiveness or impatience. The concept of time preference has existed within the field of economics for over a century; however there have been a number of recent studies that empirically estimate time preference and its impact upon decision making involving intertemporal consequences (Frederick, Loewenstein & Odonaghue, 2002). Many of these studies have both estimated individual rates of time preference and confirmed the ability of time preference to explain smoking, preference for durable goods that enhance future welfare, choice of a healthy diet, educational attainment and other behaviors involving a conscious reduction in present utility to increase future well being. This reduction in present well being to improve future well being necessarily precedes the decision to reduce consumption in the present period to increase consumption in future periods  the essence of saving.


The discounted utility model (Friedman 1957) posits a utility maximization framework that includes both current and future time periods. Utility occurring in future time periods is discounted due to uncertainty, preference for current consumption, and a finite life span. It is possible to estimate individual time preference either by asking individuals to compare varying dollar amounts in the present and future (Fisher, 1930), estimating the tradeoff between fatality risk and wage rates (Viscusi & Moore, 1989), or by the willingness to spend on improving future health (Chapman, 1996). According to Barsky, Juster, Kimball, and Shapiro (1997), rates of time preference will have a large impact on predicted accumulation of assets over time. Those who have lower rates of time preference, consistent with indicating a greater value placed on future utility, will follow an increasing consumption path with heavy accumulation of assets during working years. Those with a higher rate of time preference, indicating heavier discounting of future utility, will follow a decreasing or constant consumption path consistent with reduced asset accumulation. As asset accumulation allows increased consumption during periods of reduced income, such as retirement, maintaining increasing consumption requires increased savings. Since rates of time preference vary among individuals, and since individual utility maximizing behavior requires consideration of time preference when making choices about asset allocation over time, time preference may play an important role in explaining why some save while others do not.


Research by Finke, Huston and Weaver (2003) confirms the relative importance of time preference as a predictor of household net worth. Time preference, estimated through a factor composed of behaviors related to intertemporal decision making, such as smoking, exercising, having ones cholesterol checked, and having a flu shot, was found a strong and significant predictor of net worth among non-retired middle aged households. The time preference factor explained less variation in net worth than household income, however time preference was stronger than all demographic characteristics, such as race, as a predictor of wealth.


Specifics of Proposed Research


The proposed research is normative rather than positive. The ultimate purpose of this research is to find ways to bolster the rate and level of personal savings in the United States. Toward this end, a broader view of the full scope of savings behavior is needed along with a clearer understanding of the factors - personal, situational, institutional - that can inhibit or constrain the decision to save and the amount saved. With that understanding, appropriate intervention strategies can be developed to help remove or reduce these impediments. It is expected that some interventions would involve a change in the environment. Making payroll deduction plans more widely available would be an example. Results of this research can identify these external barriers. Other interventions will likely focus on behavior change. The transtheoretical model of behavior change proposed by Prochaska, DiClemente, & Norcross (1992) can provide a template for considering the steps involved in changing behavior related to saving. This model proposes that specific mental changes must occur before behavior can change and that change be maintained. Previous research using this model indicates that counseling, education and experience are important in fostering these changes. This research can help identify the types of financial counseling, education and experience most likely to be effective in fostering behavior change. Findings of this research can also help policy makers recognize the groups or the situations that are not amenable to change. This information is important, as it is likely that it is these groups or situations that would most likely place a demand on public resources.

Objectives

1. 1. To develop an index of savings behavior that reflects a progressive journey from non-saver to saver versus a simple dichotomy of save/do not save.
   
2. 2. To identify the specific factors that inhibit or motivate progress from non-saver to saver.
   
3. 3. To evaluate the impact of both economic and psychological factors on both the index of savings behavior and the level of saving accumulated, controlling for differences in sociodemographic characteristics and access to tax-advantaged savings vehicles.
   
4. 4. To ascertain whether or not the relationships between the economic and psychological factors and savings behavior and level vary significantly by race and gender.
   
5. 
   

Methods

Brief description of methods 1. Previous attempts to measure saving behavior have focused on only limited aspects of that behavior. From the literature reviewed for this proposal, an inventory of these measures has been compiled. As a first step in this research, gaps in those measures identified, and measures developed to fill those gaps. The result of this work will be incorporated in the survey instrument that will be administered. 2. In the psychological literature, various indexes have been developed to assess the psychological constructs of interest to this study. The contribution of these existing indexes to this research will be evaluated and, if need be, adjusted to better fit the requirements of this particular research endeavor. 3. A survey instrument will be developed and pilot tested with a small sample of current participants in an Outreach and Extension based program such as America Saves. The survey instrument will be revised as needed, based on results of the pilot test. 4. A stratified national sample will be drawn from the 48 contiguous United States. Following the classic principles outlined in Dillman (1978, 1991), a mail survey will be used to obtain data. 5. Descriptive and multivariate analysis, factor analysis, and structural equation modeling analysis will be conducted to meet the objectives previously outlined. 6. Findings of the study will be used to guide development of educational programs and public policy recommendations related to savings Characteristics of population and sample This study will focus on households with a head between ages 35 and 50. According to the life cycle hypothesis, it is this midlife group that is disposed to saving, while young or elderly households are expected to spend disproportionately more than their income. Also, as a practical matter, the ability to save and incentives for saving for these two age groups can differ greatly from those of the midlife group. Thus, excluding the young and elderly from the sample will provide a more homogeneous sample for analysis. Low- and moderate-income households would be over-sampled since savings for emergencies and major expenses may not be as crucial among higher income families. The sample would consist of at least 2,000 households in a stratified random sample. With such a sample size, the results are expected to lie within two standard deviations of the mean response. Participants would be selected by classifying all United States counties into metropolitan (those with more than 50,000 population), urban (between 50,000 and 2500 population), and rural (under 2500 population) census designations. Then, using random numbers, counties would be selected from each population designation. From a list of all cities and towns in each county selected, specific cities and towns would be selected again using random numbers. Then, from randomly selected block/s in the towns and cities, an updated alphabetical property tax and/or tenant listing of all residents will be made. Using random numbers, those selected to participate in the study will come from this list. If they do not meet the criteria for this study, another name on the list will be drawn randomly to participate. This procedure will be followed until a cooperating participant who meets the criteria is found. If no household qualifies within the selected block, another random numbered block will be selected and the same procedure will be followed as aforementioned to select the participant. Reasons why the identified person does not participate will be noted to determine possible bias in the sample. Method of Data Collection The Technical Committee will identify and contract with an independent data collection agency (e.g., National Opinion Research Center of the University of Chicago) to sample, collect, and clean the data for this project. An outside data collection agency is preferred to ensure that the instrument is developed and administered uniformly and to increase efficiency in administration and data collection. A delivered mail survey would be used because it has less bias than telephone surveys (with declining response rates and inappropriate for longer surveys) or interview surveys (very costly and difficult to control interviewer bias). Respondents can answer questions at times that they find convenient; have time to think about their answers privately and refer to their personal financial records. It is anticipated that the data collection agency will be contracted to undertake the following activities: a. Sample identification. The contractor will be asked to collect the multi-state sample to participate in the study, dividing all U. S. counties by population counts -- metro, urban, and rural. A random selection of these counties will be made; towns and cities in each of these counties will be randomly selected and a random sample selected from the blocks in each town or city where low- to moderate-income dwellers selected. The respondent must meet the selection criteria or another household would be selected (using random numbers) to participate from the same block. b. Survey delivery. The survey, cover letter, and directions for completion and return of the survey will be delivered to each respondent person-to-person. The interviewer would make up to three attempts at different times of the day within a two-week period to meet with the household head. In this face-to-face meeting, the interviewer will highlight the importance of the study and the contribution the respondent can make. They would be reminded of the importance of accuracy and completeness in answering the survey questions. The respondent would be asked to complete the survey and return in the mail by the designated date to receive a post-payment incentive. To send the payment to the respondent, their name, address and phone number would be requested and coded to correspond with the numbered survey. If no contact is made with the head of household, they do not qualify as a respondent, or he/she refuses to cooperate in the study another random selection will be made. One week later, a survey thank you and reminder postcard would be sent to each contacted respondent. After an additional week passed, a second copy of the survey, cover letter and instructions for survey completion and mailing would be sent by mail. A week later, telephone calls will be made by interviewers to encourage each person selected who has not yet responded to complete the survey and remind them of the incentive to return it as soon as possible. A third and final copy of the survey with other inserts will be sent (via certified mail) to those who have not yet completed and returned the survey. The monetary incentive would be sent with a thank you and notification of receipt of the completed survey; for incomplete surveys a smaller amount of money would be provided. Person-to-person contact by the interviewer with the potential respondent has been shown to be important in establishing a relationship under the theory of social exchange (Melevin, Dillman, Baxter, and Lamiman, 1998). Under this theory, the act of responding to a survey incurs certain costs for the respondent (time and effort in reading and answering questions) and that response can be encouraged by rewards to offset the costs (thanking respondent, complementing them for their contribution to improve the well-being of society, monetary incentives). The theory also assumes that the respondent must be able to trust that the incentives offered will be provided (Dillman, 1978). Research has consistently shown that, though costly, the extent of the follow-up (number of contacts) has a powerful positive effect on response rates (Scott, 1961; Linsky, 1975; Dillman, 1978; Heberlein & Baumgartner, 1978). Financial incentives have also been shown to be a strong inducement for response (Kanuk and Berenson, 1975; Linsky, 1995; Heberlein & Baumgartner, 1978; Dillman, 1991). Therefore, with more and different types contacts and incentives, the return rate will be higher. Melevin, et.al. (1998) experienced a 75% return rate using the proposed method of data collection; mail surveys with few mail-only contacts generally draw about a 50% to 60% return rate. The higher the return rate, the less chance there is for bias to enter into the results. If necessary due to funding constraints, the number and type of contacts and incentives used may have to be lowered. This decision will depend on project funding. Although data collection methods can be altered in several ways to reduce cost of data collection, a lower return rate for completed surveys can be expected with each form of contact that is eliminated. With a lower return rate, there is risk the sample will become biased and less representative of the total population that is being described. If the contractor cannot supply and train the interviewers throughout the U.S. to deliver the surveys face-to-face, perhaps they could train and/or direct state or county Extension faculty with a strong vested interest in the USDACREES national initiative of Retirement Security in Later Life to volunteer their services to participate in a small sample collection. An initial letter might be sent to request participation by the sample selected. Upon confirmation, the survey (with enclosures) could be sent in the mail and, at weekly intervals: a follow-up thank you/reminder postcard sent, another copy of the survey and other information sent a second time (by certified or non-certified mail) without a follow-up phone call reminder and without using incentives. This basic procedure may generate about a 55 % to 60% completed survey return rate  a barely acceptable response rate. c. Instrument refinement. The data collection agency will offer suggestions for further refinement of the instrument before administering. Several reliable tests used in previous studies will be included in the survey. Trained interviewers will conduct a second pilot study asking respondents to voice their thinking as they respond to survey items to determine if the meaning of the each survey question is clear. Other guidelines for a good survey that will be followed include putting questions in a logical sequence, using the proper format for answering, having an easily read layout with clear and concise statements, and not going beyond what is reasonable to expect for people to respond. d. Administration. The data collection agency will deliver the survey to the identified sample following the techniques as suggested above. Uniform directions for completion and a cover letter will be included. The agency will receive all completed surveys, make appropriate contacts with the respondents, and record/track response rate. e. Data cleaning. The data collection agency will do initial data cleaning/coding. Care will be taken to chronicle what is done. f. External funding will be sought to pay for the data collection. If insufficient external funding is available, the number and types of contacts may be altered as previously explained. Participating research institutions will be responsible for their prorated share of the cost to help cover the costs associated with the study. Data Analysis The data collection will result in a national data set available to the research committee. The committee as a whole will be involved in the data analysis. Descriptive analysis will be used to ascertain the socio-demographic, economic, and psychological characteristics of survey participants. It is anticipated that researchers will utilize multivariate analysis to accomplish study objectives. Factor analysis will be used to develop an index of saving behavior. Regression analysis and structural equation modeling will be used to explore the relationship between economic and psychological factors and to assess the impact that both economic and psychological factors have on both savings behavior (as measured by the developed index) and level of saving. Components of Survey Instrument Development of the survey instrument will be collaborative effort of participating researchers. Based on previous literature, three broad elements are proposed: 1. Economic factors such as temporal preference, income, bequest motive, wealth, debt load, level of financial literacy, access to financial institutions with convenient saving mechanisms, access to saving-related mechanisms such as employer matching contributions and tax benefits, access to services such as payroll deduction and direct deposit, and the possibility of asset-restriction to qualify for means-tested transfer programs; 2. Psychological factors such as motivation to save, expectations regarding the success of saving, and scales to assess the presence of psychological factors such as self-control, self-efficacy, perfectionism, impulsivity, and materialism; and 3. Demographic factors such as marital status and age of household head, age and number of children, race (ethnicity), gender, education, neighborhood quality, parental behavior regarding money, employment status, and occupation. Hypotheses to be tested Regarding demographic and economic variables, it is expected that, as compared with respondents who do not save or who save relatively less, respondents who save or who save relatively more: 1. have higher levels of financial literacy 2. have higher levels of income and wealth 3. have a lower rate of time preference 4. have lower levels of debt 5. have a bequest motive for saving 6. are in the later years of mid-life 7. are male 8. are white 9. have relatively smaller family size Regarding psychological variables, it is expected that 1. higher self-control will be related to greater savings 2. self-control will be an important determinant of moving from consideration and intent stages of saving to actual behavior 3. higher perfectionism will be related to greater savings 4. perfectionism may help to distinguish between people who have attempted to save, but have been unable to maintain this behavior and those who have been successful in saving; in this case, contrary to expectations previously outlined in point 3, high perfectionism will relate to the inability to maintain savings behavior 5. low impulsivity will be related to greater savings 6. low impulsivity will be an important determinant of moving from consideration and intent stages of saving to actual behavior 7. high self-efficacy will be related to greater savings 8. self-efficacy will be an important determinant of moving from consideration and intent states of saving to actual behavior 9. low materialism will be related to greater savings 10. the relationship between psychological variables and saving behavior will be mediated differently by race and gender

Measurement of Progress and Results

Outputs

• A model that identifies progressive saving stages based on economic, psychological, and demographic factors
• Identify psychological and economic strategies to influence consumers such that they are able to actualize to a higher stage of saving behavior

Outcomes or Projected Impacts

• Identify economic and psychological factors that contribute to increased consumer saving behavior
• Enable educators to develop educational interventions with consumers to increase their savings

Milestones

(2004): 1. The committee will develop and refine the survey instrument. The committee will seek external funding for the national data collection. 2. Bids will be solicited from data collection agencies.

(2005): 1. The preliminary pilot test of the survey instrument will be conducted using participants who are already involved with a savings program such as America Saves 2. The results of the preliminary pilot test will be used to refine the survey instrument. 3. The stratified national sample will be developed. 4. A second pilot test will be conducted with the stratified national sample and results used to modify the survey instrument, if needed. 5. Preliminary results will be examined, and if changes in the survey instrument are needed, the survey instrument will be modified.

(2006): 1. Data will be collected from the stratified national sample during the first six months of the year. 2. Data entry, cleaning, and analysis will be conducted by the data collection agency. 3. Data analysis by the research committee members will focus on the relationship of the economic and psychological factors to the stages of the savings model. 4. Drafts of papers will be prepared and papers will be submitted to professional and practitioner organizations.

(2007): 1. Data analysis relating the demographic factors to the stages of the saving model will continue. 2. Drafts of papers will be prepared and papers will be submitted to professional and practitioner organizations. 3. The committee will continue to disseminate the results by presentations at conferences and the submission of journal articles to appropriate professional organizations. 4. The committee will develop outreach materials based on the results of the study.

(2008): 1. The committee will develop and submit a proposal to study the effect of the outreach materials and programming. 2. The committee will continue to disseminate the results and implications of the study.

(0):0

Projected Participation

View Appendix E: Participation

Outreach Plan

The results of the study including the assessment tool, recommendations for educational curriculum, and recommendations for further research will be made available through journal articles in appropriate journals, conference proceedings, and a special topics pre-conference initiated by the research committee. This ensures that results of the findings will be shared with a large, diverse audience that is interested in savings behavior.


Special emphasis will be placed on disseminating the study results and outcomes to organizations and entities with experience and relationships working with underserved communities. Through extension education and private entities, the technical committee will maximize the dissemination of this information.

Organization/Governance

The committee will use a standard form of governance. The research committee will be comprised of a chair, vice-chair, secretary, and members at large. Any investigator from an agricultural experiment station (AES) may be a member. Additionally non-AES affiliated researchers may be granted membership upon consent and approval of the technical committee. The technical committee retains all powers except those specifically delegated to a subcommittee.
1. A chair, vice-chair, and secretary will be elected to their offices by the membership. Each state or agricultural experiment station will receive only one vote. 2. The chair will be responsible for calling and coordinating the meetings, tracking progress of each station, and overseeing the dissemination of results. 3. The vice-chair will assist the chairman and be responsible for duties as assigned. 4. The secretary will be responsible for minutes of all official project meetings, maintaining official project correspondence and other duties as assigned. 5. Members, the investigators, will be responsible for updating the technical committee on progress of the project in their state/station.

Literature Cited

Ando, A. & Modigliani, F. (1963). The life cycle hypothesis of saving: Aggregate implications and tests. The American Economic Review, 53(1), 55-84.


Avery, R.B. & Kennickell (1991). Household saving in the U.S. Review of income and wealth 37(4), 409-433.


Bandura, A. (1992), Exercise Of Personal Agency Through The Self-Efficacy Mechanism, in R. Schwarzer (ed.), Self-Efficacy: Thought Conrol Of Action, Washington, DC: Hemisphere, 3-38.


Bandura, A. (1997), Self-Efficacy: The Exercise of Control. New York: Freeman.


Bandura, A. (1977), Social Learning Theory. Englewood Cliffs, NJ: Prentice-Hall.


Baumeister, R. F. (2002). Yielding to temptation: Self-control failure, impulsive purchasing, and consumer behavior. Journal of Consumer Research, 28, March, 670-676.


Barsky, R., Juster, F., Kimball, M., & Shapiro, M. (1997). Preference parameters and behavioral heterogeneity: An experimental approach in the Health and Retirement Study. The Quarterly Journal of Economics, 112(2), 729-758.


Barrios, F.X. (1985), A Comparison of Global and Specific Estimates of Self-Control, Cognitive Therapy and Research, 7, 543-554.


Bernheim, B. D., Shleifer, A. & Summers, L. H. (1986). The strategic bequest motive. Journal of Labor Economics, 4(3), S151-182.


Beverly, S. (1997). How can the poor save? Theory and evidence on saving in low-income households. St. Louis: Washington University. Retrieved May 30, 2001. http://gwbweb.wustl.edu/Users/csd/workingpapers/wp97-3.pdf.


Cavanagh, J. & Sharpe, D. L. (2002) The impact of debt levels on participation in and level of discretionary retirement savings. Financial Counseling and Planning, 13(1), 47-59.


Chapman, G.B. (1996). Temporal discounting and utility for health and money. Journal of Experimental Psychology, Learning, Memory, and Cognition, 22(3), 771-791.


Chang, R.Y. (1994). Savings behavior of U.S. household in the 1980s: Results from the 1983 and 1986 Survey of Consumer Finance. Financial Counseling and Planning, 5, 45-64.


Cramer, S. (1996-2001). Retirement economic well-being of women in Nebraska and cross-culturally. University of Nebraska.


De Leeuw, E. D., Mellenbergh, G. J., and Hox, J. J. (1996). The influence of data collection method on structural models: A comparison of a mail, a telephone, and a face-to-face survey. Sociological Methods and Research, 24:443-372.


DeVaney, S.A. & Chien, Y (2001). A model of savings behavior and the amount saved in retirement accounts. Journal of Financial Service Professionals, LV, 72-79.


Dillman, Don. A. (1978). Mail and Telephone Surveys: The Total Design Method. New York, NY: John Wiley and Sons.


Dillman, Don A. (1991). The design and administration of mail surveys. Annual Review of Sociology, 17:225-249.


Faber, R.J. (2000), A Programmatic Investigation into Compulsive Buying, in A.L. Benson (ed.), I Shop Therefore I Am: Compulsive Buying and the Search for Self, Northvale, NJ: Aronson Press, 2000, 27-53.


Faber, R.J. & Vohs, K. (in press), To Buy or Not to Buy?: Self-Control and Self-Regulatory Failure in Purchase Behavior, in R.F. Baumeister & K.D. Vohs (eds), Handbook of Self Regulation.


Finke, M., Huston, S.J. & Weaver, D. (2003) "Time Preference and Intertemporal Decision Making: Evidence From the Health and Retirement Study," In Jinkook
Lee (ed.). Consumer Interests Annual, 49.


Fisher, I. (1930). The theory of interest. New York: Macmillan.


Frederick, S., Loewenstein, G. & ODonoghue, T. (2002). Time discounting and time preference: A critical review. Journal of Economic Literature, 40, 351-401.


Friedman, M. (1957). A theory of the consumption function. New York: Princeton University Press.


Grable, J., & Lytton, R. (1997). Determinants of retirement savings plan participation: A discriminant analysis. Personal Finances and Worker Productivity 1(1), 184-189.


Gutter, M.S., Fox, J. J. & Montalto, C. P. (1999). Racial differences in investor decision making. Financial Services Review, 8(3), 149-162.


Heberlein, T.A., and Baumgartner,R. (1978). Factors affecting response rates to mailed questionnaires: A quantitative analysis of the published literature. American Sociological Review, 43:447-462.


Hogarth, J.M. (1991) Saving and dissaving in retirement. Family Economics Review, 2 (2), 13-21.


Hogarth, J. M. & Anguelov, C. E. (2001). Can the poor save? Proceedings of Association for Financial Counseling and Planning, 42-50


Kanuk, L., and Berenson, C. (1975). Mail surveys and response rates: A literature review. Journal of Marketing Research, 12,440-453.


Katona, G. (1975). Psychological economics. New York: Elsevier.


Keynes, J. M. (1936). The general theory of employment, interest, and money. Reprinted in 1997. New York: Prometheus Books.


Lea, S. E. G., Tarpy, R. M., & Webley, P. (1987). The individual in the economy: A textbook of economic psychology. New York: Cambridge University Press.


Linsky, A. S. (1975). Stimulating response to mailed questionnaires: A review. Public Opinion Quarterly, 39:82-101.


Livingstone, S., & Lunt, P. (1993). Savers and borrowers: Strategies of personal financial management. Human Relations, 46(8), 963-999.


Lunt, P. K., & Livingstone, S. M. (1991) Psychological, social, and economic determinants of saving: Comparing recurrent and total savings. Journal of Economic Psychology, 12(4), 621-641.


Maibach, E. and D.A. Murphy (1995), Self-Efficacy In Health Promotion Research And Practice: Conceptualization And Measurement, Health Education Research, 10 (1), 37-50.


Melevin, Paul T., Dillman, Don A., Baxter, Rodney, and Lamiman, C. Ellen. (1998). Personal delivery of mail questionnaires for household surveys: A test of four retrieval methods. Sociology and Social Research.


Office of Management and Budget (2004). Table B-32. Gross Saving and investment, 1959-2002. Accessed May 16, 2003 at http://w3.access.gpo.gov/usbudget/fy2004/sheets/b32.xls


OGuinn, T.C. and R.J. Faber (1989), Compulsive Buying: A Phenomenological Exploration, Journal of Consumer Research, 16, 147-157.


Prochaska, J. O., DiClemente, C. C. & Norcross, J. C. (1992). In search of how people change: applications to addictive behaviors. American Psychologist, 47, 1102-1114.


Rha, J.Y.; Montalto, C.P.; & Hanna, S. (2003). The Effect of Self-Control Mechanisms on Household Saving Behavior. Working Paper The Ohio State University.


Richins, M. L. & Dawson, S. (1992). A consumer values orientation for materialism and its measurement: Scale development and validation, Journal of Consumer Research, 19, 303-316.


Rindfleisch, A., Burroughs, J. E., & Denton, F. (1997). Family structure, materialism, and compulsive consumption, Journal of Consumer Research, 23(4), 312-25.


Romal, J. B. & Kaplan, B. J. (1995). Differences in self-control among spenders and savers. Psychology: A Quarterly Journal of Human Behavior, 32 (2), 8-17.


Schaeffer, P. (2002). Connecting community and household assets. Southern Perspectives.


Schuchardt, J. (2002). Building household assets. Southern Perspectives, Summer, 8-9.


Scott, C. (1961). Research on mail surveys. Journal of Royal Statistical Society, Series A, 124:143-205.


Sharpe, D. L. Retirement planning: Macro-environmental and life situation factors. University of Missouri-Columbia.


Shefrin, H. M. & Thaler, R. H. (1988). The behavioral life-cycle hypothesis. Economic Inquiry, 26, 609-643.


Sherer, M. Maddus, J. E., Mercadente, B., Prentice-Dunn, S. Jacobs, B. & Rogers, R. W. (1982). The self-efficacy scale: Construction and validation. Psychological Reports, 51, 663-671.


Sherraden, M. W. (1991). Assets and the poor: A new American welfare policy. Armonk, NY: M. E. Sharpe


Viscusi, W. K., & Moore, M. J. (1989). Rates of time preference and valuations of the duration of life. Journal of Public Economics, 38(3), 297-317.


Vohs, K.D., A.M. Bardone, T.E. Joiner, L.Y. Abramson & T.F. Heatherton (1999), Perfectionism, Perceived Weight Status, and Self-Esteem Interact to Predict Bulimic Symptoms: A Model of Bulimic Symptom Development, Journal of Abnormal Psychology, 108, 695-700.


Vohs, K.D., T.F. Heatherton and M. (2001), Disordered Eating And The Transition To College: A Prospective Study, International Journal of Eating Disorders, 29, 280-288.


Vohs, K.D., & Heatherton, T.F. (2000). Self-regulatory failure: A resource-depletion approach. Psychological Science, 11(3), 249-254.


Vohs, K.D., & Faber, R. (2002). Self-regulation and impulsive spending patterns. In P.A. Keller & D. W. Rook (Eds.), Advances in Consumer Research.


Wdrneryd, K. E. (1999). The Psychology of Saving: A study on economic psychology. Cheltenham, UK: Edward Elgar Publishing.


Youn, S. and R.J. Faber (2002), Impulse Buying: Its Relation to Personality Traits and Cues, in S. J. Hoch and R. J. Meyer (eds.), Advances in Consumer Research, Vol. XXVII, Provo, UT: Association for Consumer Research, 2000, 179-185.


Yuh, Y. & DeVaney, S. (1996). Determinants of couples defined contribution retirement funds. Financial Counseling and Planning 7, 31-38.


Zhong, L., & Xiao, J. (1995). Determinants of family bond and stock holdings. Financial Counseling and Planning,6, 107-114.

Attachments

Land Grant Participating States/Institutions

GA, KY, MN, MO, ND, OH, PA, SC, SD, TN, VA

Non Land Grant Participating States/Institutions

Emeritus Collaborator