Identifying Cognitive Resources Associated with High ...



Identifying Cognitive Resources Associated with Retirement Saving

April 16, 2010

Project Director: Miles Kimball

Co-Director: Tyler Shumway

Other Key Personnel: Annamaria Lusardi and Laurie Pounder DeMarco

Statement of Policy Relevance: There is a strong policy interest in encouraging people, where feasible, to accumulate additional savings for retirement to complement Social Security income during retirement. Identifying why in similar situations some people save little while others save a great deal will inform efforts to encourage retirement saving.

Project Abstract: In order to better understand which cognitive resources are most closely associated with retirement saving, this project will study the statistical relationship between the propensity to save and three broad types of cognitive resources: (1) internal locus of control—that is, being convinced that one’s actions can make a difference for outcomes in one’s own life, (2) planning, budgeting and debt-management skills and attitudes, and (3) financial knowledge. High-quality HRS data is available for each of these concepts. The analysis will identify a high propensity to save by a low value of consumption in relation to the present value of wealth and labor income. The HRS Psychosocial Leave-Behind provides a detailed measure of internal vs. external locus of control and related measures, while this and the 2008 and 2010 Financial Sophistication and Investment Decision Making modules provide excellent measures of the other two types of cognitive resources.

Research questions and policy relevance: Trying to ensure that Americans have adequate financial resources during their retirement years is a matter of great concern. Given the constraints on the amount of directly provided financial resources that can be financed by payroll taxes, there is a strong policy interest in encouraging people, where feasible, to accumulate additional savings for retirement to complement their Social Security income during retirement. In trying to know how to encourage additional retirement saving, it is crucial to understand why—in what look like similar situations—one person saves a great deal, while another person saves very little. Understanding why in similar circumstances some people save more towards their retirement years than others is bound to be enormously helpful in figuring out how to influence saving.

TIn this project focuses, we will focus on understanding the role cognitive resources play in helping people save. Scientific research on the role of cognitive resources for retirement saving has the potential to inform the extensive efforts by many public and private agencies—including the SSA itself—s—including the Social Security Administration itself—at enhancing financial literacy. Because we include more in our concept of cognitive resources for retirement saving than is included in the narrowest estr conceptions of financial literacy, we hope to be able to identify additional avenues for improving retirement saving by enhancing financial literacy, broadly conceived to include all relevant cognitive resources. In particular, we will use two special data modules in the Health and Retirement Study to better establish the relationship (or lack of relationship) between three broad types of cognitive resources and the level of an individual or household’s retirement saving:

1. Internal locus of control—which in the context of retirement saving means being convinced that one’s own actions can make a big difference in what one’s retirement years are like;

2. Planning, budgeting and debt-management skills; and

3. Knowledge about finances and financial markets.

Since the potential effect of internal as opposed to external locus of control on retirement saving is the most unusual idea, it deserves additional discussion. (We will discuss the other two below in the context of our preliminary research.) Although locus of control is sometimes thought of as a personality trait, an important set of experiments have demonstrated that targeted interventions can have an important effect on perceived locus of control in a particular domain: Blackwell, Trzesniewski and Dweck (2007) showed for junior-high-school students and Aronson, Fried and Good (2002) showed for college students that relatively modest interventions fostering the belief that—as recent psychological research documented detailed in Nisbett 2009 indicates—hard work can increase one’s intelligence leads to significant improvements in academic performance. Our hypothesis is that being convinced that it is within one’s power to make one’s own retirement years better by saving in advance leads to more retirement saving. We will be able to test for correlations predicted by this hypothesis both with questions about sense of personal control over life in general and sense of personal control over one’s financial situation more specifically.

Existing Evidence: In previous work (Kimball and Shumway, 2009) we designed and analyzed a special module for the June 2008 Survey of Consumers to look for attitudes and abilities associated with high saving. We measure saving with an index based on responses to 10 survey questions, and regressed this saving index on responses to 34 different survey questions and standard demographics. Because we generated our list of 34 questions with the help of two focus groups, the survey dealt with a number of topics that have not been the focus of previous research. This work helped us draft several of the questions on the 2010 HRS Financial Sophistication and Investment Decision Making module (FS&IDM).

We had also designed a module to measure financial literacy and financial sophistication for the 2008 HRS, whose findings are analyzed in Lusardi, Mitchell and Curto (2009).

As in previous findings in Ameriks, Caplin and Leahy (2003) and Lusardi (19992004), in Kimball and Shumway (2009) we find a strong relationship between retirement savings and responses to questions about planning and budgeting. Moreover, aAgreement with statements such as “I often wonder, ‘where did all my money go’” and “thinking about money stresses me out” is significantly related to our saving index. We also find that measures of institutional trust are related to saving. Respondents that agree with “If I try to save through financial institutions, someone is likely to figure out a way to cheat me out of the money” are generally poor savers.

Most surprising, we find that responses to questions about locus of control and fatalism are related to saving. People who agree that “Many of the things that keep me from saving more money are out of my control” and “No one can predict the future, so trying to save doesn’t do much good” save relatively little. TWe find this evidence is intriguing, but we are reluctant to draw strong conclusions from the limited data available on the Survey of Consumers. Using HRS data, including the the new 2010 FS&IDM module we designed with Olivia Mitchell and the rich data on the HRS Psychosocial Leave-Behind (PLB) will allow us to explore our hypotheses with much more power.

Proposed Research: Our basic research design is to correlate HRS measures of the propensity to save with measures of locus of control, planning, and financial sophistication using multiple regression analysis to control for demographics and other factors known to affect retirement saving. Our preferred savings measure is the ratio of consumption to fulltotal wealth following Pounder (2007).[1] FullTotal wealth is the sum of net worth[2] (from Rand) and human wealth—the present value of future after-tax income from four sources: earnings, social security, pensions, and other government benefits.

Pounder, 2007 shows that the consumption to full wealth ratio is a clean measure of the propensity to save focusing on attitudes and preferences as opposed to an individual’s situation. A low consumption to full wealth ratio indicates high savers, while a high consumption to full wealth ratio indicates low savers.

For non-retired households, earnings are estimated from current wages incremented each year, using a tenure and experience equation similar to that in Gustman & Steinmeier (2002), until the self-reported expected retirement age. Future Social Security benefits are calculated from actual Social Security earnings histories and projected earnings, adjusted using rules for early or late retirement. The present value of pensions are estimated using the HRS Pension Calculator with adjustments such as actual earnings histories from W-2 data, supplemented with self-reports for defined contribution plans. Government benefits such as veterans or disability are extended, and an income floor from SSI is modeled. Tax brackets are estimated annually and the present value calculation includes a discount rate and a year-, age-, and race-specific mortality hazard.

The numerator of our savings measure is consumption from the CAMS supplement to the HRS, whose 26 categories of spending cover well over 90% of total expenditures based on the more detailed Consumer Expenditure Survey (CEX). To get from expenditures to consumption, we replace housing expenditures with imputed rental equivalence for homeowners, and vehicle purchases with imputed vehicle consumption.

Using multiple regression analysis to allow for other statistical controls, we will study the correlations between these savings measures and a wide range of measures of locus of control, planning and budgeting, and financial sophistication available on the HRS. (1) The Psychosocial Leave-Behind Participant Lifestyle Questionnaire (PLB) (administered to half of the HRS sample in 2006 and the other half in 2008) has a ten-question sequence on locus of control (e.g. “other people determine most of what I can and cannot do,” “what happens in my life is often beyond my control” vs. “whether or not I am able to get what I want is in my own hands,” “what happens to me in the future mostly depends on me), plus other questions on closely related beliefs such as fatalism, and optimism/pessimism (e.g. “There is no use in really trying to get something I want because I probably won’t get it,” and “In uncertain times, I usually expect the best” vs. “If something can go wrong for me it will.”). In addition there is a 0 to 10 scale for “amount of control you have over your financial situation these days” which provides a financial locus of control measure, as well as a set of questions on trust (e.g., “Most people will use somewhat unfair means to gain profit or an advantage rather than lose it”) which is necessary to feel one is in financial control when entrusting one’s savings to another person or institution. (2) Both the PLB and 2010 FS&IDM module provide data on planning, budgeting and debt-management (e.g. PLB: “I enjoy making plans for the future and working to make them a reality;” FS&IDM: “Thinking about money stresses me out,” “I often wonder ‘Where did all my money go?’” “It is a big financial mistake to use a credit card without paying off the balance every month”). (3) Finally, both the 2010 FS&IDM module and a similar 2008 module provide measures of financial sophistication. Our sophistication measure will be based on responses to 18 quiz-like questions that vary substantially in difficulty and topic, as discussed in Lusardi, Mitchell and Curto (2009). Having a large number of questions to measure sophistication should give us a much better measure than the comparatively simple financial literacy measures that have been used previously. In Kimball and Shumway (2010) we found a similar measure of financial sophistication to be strongly related with many different dimensions of portfolio choice, but could not look at the relationship to overall retirement saving for lack of a measure of overall retirement saving on the survey we used.

Potential Outcome of the Proposed Research: We view the proposed statistical analysis of HRS data as the first step in a scientific research agenda that will inform effective interventions to help people make and carry through better retirement saving decisions. In particular, existing experimental results showing that enhancing student’s sense of control over their own intelligence can lead to better academic outcomes raise the tantalizing possibility that effective interventions can be designed to enhance sense of control over one’s future retirement years and thereby significantly improve retirement saving choices. Such interventions could readily be tested both in lab experiments and in field experiments. (For example, the annual pamphlets sent out to Social Security participants provide an excellent platform for field tests, since randomization by the last digit of someone’s Social Security number would allow surveys with confidential links to Social Security data such as the HRS to track the outcome of the interventions.) But it is important to first identify—using high-quality HRS data—which types of cognitive resources have especially strong associations with retirement saving and therefore are especially promising candidates for later experimental tests.

Project Timeline:

September 30, 2010 Award Period Begins

September-December, 2010 Initial Data Analysis

January 10, 2011 Submit Quarterly Progress Report

January-March, 2011 Prepare First Draft of Working Paper

April 10, 2011 Submit Quarterly Progress Report

April 8 & 9, 2011 MRRC Researcher Workshop (dates tentative)

April-June, 2011 Revisit Data Analysis in Light of First Draft

July 10, 2011 Submit Quarterly Progress Report

July-August, 2011 Prepare Final Draft of Working Paper

August 4 & 5, 2011 Retirement Research Consortium annual meeting

(dates tentative)

September, 2011 Final edits and submission of all deliverables

September 29, 2011 Deadlines to submit project deliverables: working

Paper, executive summary, results abstract & key findings

November 14, 2011 Final date for invoice submission

References

Ameriks, John, Andrew Caplin and John Leahy, 2003. “Wealth Accumulation and the Propensity to Plan,” Quarterly Journal of Economics 118, 1007-1047.

Aronson, Joshua, Carrie B. Fried and Catherine Good, 2002. “Reducing the Effects of Stereotype Threat on African American College Students by Shaping Theories of Intelligence,” Journal of Experimental Social Psychology, 38, 113-125.

Blackwell, Lisa S., Kali H. Trzesniewski and Carol Sorich Dweck, 2007. “Implicit Theories of Intelligence Predict Achievement Across an Adolescent Transition: A Longitudinal Study and an Intervention,” Child Development, 78 (January), 246-263.

Gustman, Alan and Thomas Steinmeier, 2002. “The Social Security Early Entitlement Age in a Structural Model of Retirement and Wealth” NBER Working Paper 9183, September.

Hurst, Erik, “Grasshoppers, Ants and Pre-Retirement Wealth: A Test of Permanent Income Consumers”, MRRC Working paper 088, September 2004.

Kimball, Miles S. and Tyler Shumway, 2010, “Investor Sophistication and the Home Bias, Diversification, and Employer Stock Puzzles,” Working paper, University of Michigan.

Kimball, Miles S. and Tyler Shumway, 2009, “Fatalism, Locus of Control, and Retirement Savings,” Working paper, University of Michigan.

Lusardi, Annamaria, 1999. "Information, Expectations, and Savings for Retirement," in Henry Aaron (ed.), Behavioral Dimensions of Retirement Economics, Washington, D.C.: Brookings Institution Press and Russell Sage Foundation, 1999, pp. 81-115.

Lusardi, Annamaria, Olivia Mitchell and Vilsa Curto, 2010. “Financial Literacy and Financial Sophistication Among Older Americans,” NBER Working Paper n. 15469, November 2009.

Nisbett, Richard E. 2009, Intelligence and How to Get It: Why Schools and Cultures Count, Norton, New York.

Pounder, Laurie, 2007. “Life-Cycle Consumption Examined,” Ph.D. dissertation, University of Michigan.

-----------------------

[1] We will also run our tests using an alternative measure of savings, the residual of a wealth equation using income, employment, and demographic characteristics such as in Hurst (2004).

[2] We will use the flat files provided by Rand.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download