Short research note
Evidence of lower risk tolerance among public
sector employees in their personal financial
Michael J. Roszkowski1* and John E. Grable2
1La Salle University, Philadelphia, Pennsylvania, USA
2School of Family Studies and Human Services, College of Human Ecology,
Kansas State University, Manhattan, Kansas, USA
Critics claim that excessive risk avoidance is characteristic of public sector workers.
To test this contention, the financial risk tolerance scores of public sector and private
sector employees who had used financial planning services were compared on a
financial risk tolerance scale. Public sector employees scored lower on financial risk
tolerance relative to private sector employees. Differences remained even after
controlling for other variables linked to risk tolerance.
Critics of the public sector often contend that civil servants are excessively risk averse,
which supposedly renders them ineffectual, especially in managerial positions
(Bozeman & Kingsley, 1998). This charge – which has been levelled against government
workers in the US (Gingrich, 2005), New Zealand (Wright & de Joux, 2003), and Great
Britain (Prowle, 2000) – is at the heart of the ongoing debate about the need to ‘reform’
and ‘privatize’ the public sector (e.g. Fitzgerald, 1988; Gingrich, 2005; Gore, 1993).
However, as Bozeman and Kingsley (1998, p.109) aptlyobserved, the notion of low risk
tolerance in the public sector is ‘widely embraced but rarely tested’.
Nature of the evidence
Some of the evidence is very tenuous, based on the conjecture that since risk averse
individuals fear unemployment (Feinberg, 1977) and government jobs are more secure
than private sector jobs (Utgoff, 1983), then people working in the public sector must
be more risk averse. Frequently cited as direct evidence of the risk aversion of public
servants is a study by Bellante and Link (1981) in which risk tolerance was measured on
the basis of seat belt usage, insurance ownership, smoking, and drinking. Public sector
*Correspondence should be addressed to Dr Michael Roszkowski, La Salle University, Philadelphia, PA 19141-1199, USA
Journal of Occupational and Organizational Psychology (2009), 82, 453–463
q 2009 The British Psychological Society
employees scored lower on this measure. Fitzgerald (1988) implied that this research
offered ample proof of the poor managerial skills of the public sector employee.
Starr (1989), however, expressed dismay at this suggestion: ‘In other words, we are
asked to believe that people who do not smoke but wear seats belts and carry insurance
are exhibiting a general aversion to risk that makes them poor managers of
Contexts – specific nature of risk tolerance
Starr’s skepticism about the notion of ‘general’ risk aversion is supported by research
conducted nearly a decade earlier by Jackson, Hourany, and Vidmar (1972), which
identified four contexts for risk – physical, monetary (i.e. financial), social, and
ethical/legal. For the most part, Bellante and Link (1981) were assessing physical risk
taking, which on the face of it, has rather limited relevance to competence in most
positions in government. Although there is some debate as to the exact number of
contexts, it is well established that consistency of risk taking between contexts is rather
low (Blais & Weber, 2006; Jackson et al., 1972) and that it is hard to identify a ‘general’
risk taking factor (Slovic, 1972). According to Blais and Weber (2006), the within-person
variation in risk taking across different domains of risk taking is about seven times larger
than the one between-persons.
Differences between sectors on ‘general’ risk taking
A recent study in Germany (Dohmen et al., 2005) nonetheless asked public and private
sector employees for self-ratings on ‘general’ risk taking. Public sector employees rated
themselves lower than private sector employees. The meaning of this difference is
ambiguous, however, because it is unclear which type of risk taking was being rated in
the abstract. Inadvertently, this study too might have picked up differences in physical
risk taking given research showing that when asked for a self-rating on risk taking in the
abstract (i.e. without a context), respondents tend to base their assessment mainly on
their physical risk taking proclivities (Rohrmann, 2004).
Low agreement between different approaches to assessment
Different means of assessment, even within the same context – such as self-rating versus
choice on a lottery – can produce contradictory results. Consider the research
conducted by Dohmen et al. (2005), which examined differences by sector in self-rated
willingness to take risks in driving, financial matters, sports and leisure, career, and
health. In addition to the self-ratings, a question involving a hypothetical lottery (a test of
financial risk taking) was administered. Public sector workers rated themselves lower
than private sector employees on career risk taking, but not on financial risk taking.
Surprisingly, in terms of the lottery, public sector employees were actually more risk
taking than private sector employees. Yet a similar study conducted in The Netherlands
(Hartog, Ferrer-i-Carbonell, & Jonker, 2002), which also measured risk tolerance on the
basis of a hypothetical lottery, found greater risk aversion among persons employed in
the public sector, but only in two of their three samples.
Limitations in studies focusing on financial risk tolerance
Financial risk tolerance is perhaps the most relevant context to consider when
addressing the issue of differences between public and private sector workers because
454 Michael J. Roszkowski and John E. Grable
nearly all key decisions have fiscal implications on an agency’s budgets and capital
investments (Walls, 2005). The few studies that have considered this aspect of risk
between sectors have as their major limitation either a reliance on global self-ratings or
the measurement of risk preferences with just a single lottery question. Single items of
any type are notoriously unreliable. Furthermore, global self-ratings of risk tolerance are
susceptible to self-stereotyping (Roszkowski & Grable, 2007). Risk preferences on
lottery questions, in turn, can easily change across pay-off domains and response modes
(Schoemaker, 1990). For instance, framing the same alternatives as either losses or gains
exerts a marked influence on choice (Highhouse & Yuce, 1996; Roszkowski &
Whether a person is classified as a risk avoider or a risk taker on a single question
thus may depend on the nature of the question. Assessment of any construct, including
risk tolerance, is best accomplished with a multi-item test (Roszkowski, Davey, &
Grable, 2005). Moreover, since the overlap between different approaches to measuring
financial risk tolerance is low (Yook & Everett, 2003), in the absence of information
about which procedure works best with a particular individual, it is prudent to diversify
assessment techniques and average the results.
Contribution of the current study
The aim of our research was to compare the public and private sectors on a multi-item
measure of financial risk tolerance that incorporates multiple approaches to measuring
this construct. In this analysis, the private sector was split into two parts – employed by
another and self-employed – in order to account for the possibility that self-employed
persons might be more risk tolerant (Ekelund, Johansson, Jarvelin, & Lichtermann,
2005). The analysis also controlled for variables linked to financial risk tolerance in prior
research (see Hallahan, Faff, & McKenzie, 2004) to determine if perhaps the relationship
is spurious. Since the questionnaire dealt with matters of financial risk tolerance
unrelated to the job, it allows one to see if differences exist outside the workplace.
Financial risk tolerance was assessed using the surveyof financial risk tolerance (SOFRT)
published by The American College (Roszkowski, 1992). The scale was created to help
financial advisors make investment recommendations that are consistent with a client’s
risk tolerance. It consists of 51 items that are averaged to produce a total score which
can theoretically range from 0 (extreme risk aversion) to 100 (extreme risk loving).
The SOFRT employs a comprehensive approach to assessing this construct, including:
(1) minimum return required to prefer a risky venture over a sure one; (2) minimal
probability of success required to take a risky option instead of a guaranteed one;
(3) preferences for different investment vehicles; (4) reactions to sample portfolios;
(5) investment objectives; (6) emotional reactions to risky situations; (7) life-style
characteristics; and (8) self-classification. Both loss and gain frames are used. Sample
items are shown in the Appendix.
The internal consistency reliability (Cronbach’s a) of the scale was .91 in the
developmental sample, with alphas ranging from .81 and .86 in samples of actual users.
A 45-day interval test–retest reliability equalled .83 (Roszkowski, Delaney, & Cordell,
2004). Scores on the SOFRT correlate with an adviser’s impressions of the risk tolerance
Financial risk tolerance455
of the client (Roszkowski & Grable, 2005), and they are predictive of actual investing
behaviour (Roszkowski, 1992).
The final section of the SOFRT requests basic demographic information. Age is
recordedas an exact value (inyears). Sex isa twooption question, and level ofeducation
is assessed with seven options: less than high school; high school; some college;
bachelor’s degree; master’s degree; law degree; and doctorate. Employment sector, has
four choices: (1) an employee of a private company or business, or an individual
working for wages, salary, or commissions; (2) a government employee (federal, state,
county, or local); (3) self-employed in your own business, professional practice, or farm;
and (4) retired. Information about personal income, household income, and net wealth
are collected in brackets. The brackets for income are: under $50,000; $50,000–
$99,000; $100,000–$149,000; $150,000–$199,000; $200,000–$249,000; $250,000–
$500,000; and over $500,000. Wealth levels are: under $250,000; $250,000–$499,000;
$500,000–$999,000; $1,000,000–$2,499,000; $2,500,000–$5,000,000; and over
$5,000,000. These ranges were coded 1 through 7 for income and 1 through 6 for
The sample consisted of clients of financial planners who had used the SOFRT in the
process of client advising during years 1992 through 1998. The database contained 946
cases, but clients who were retired were excluded from this analysis. The 745 cases with
employment sector information (excluding retired) were distributed as follows across
the sectors: 399 private sector; 260 self-employed; and 86 public sector. The average age
was 45.39 (SD ¼ 10:01). The average personal income, expressed on the seven-point
code (brackets), equalled 2.33 (SD ¼ 1:57). On this samemetric, household income was
2.86 (SD ¼ 1:54). Net wealth equalled 2.22 (SD ¼ 1:26). The educational level of the
sample was distributed as follows: 1.5% less than high school; 6.6% high school
graduate; 21.8% some college; 35.3% bachelor’s degree; 21.7% masters degree; 2.8% law
degree; and 10.4% doctorate. For the purpose of the present analysis, the last two
categories were collapsed, and the levels coded 1 through 6.
The mean risk tolerance score for the norm group for the SOFRT is 43 (Roszkowski,
1992). In the current sample (N ¼ 745), the mean was 41.43 (SD ¼ 11:30;
minimum ¼ 6, maximum ¼ 88; kurtosis ¼ :16; skewness ¼ :20). For the 654 cases
used in the multiple regression analysis, the mean SOFRT score was 41.54 (SD ¼ 41:54;
minimum ¼ 6, maximum ¼ 88; kurtosis ¼ :18, skewness ¼ :21).
Evidence suggesting the sample is representative
Since the sample was one of convenience, tests were conducted to determine how
closely the sample matched the profile of the public sector employee reported in prior
research. Because it can be argued that some of the variables are ordinal rather than
interval in nature or that some data fails to meet normal distribution assumptions, both
parametric and non-parametric statistical tests were used to judge the representative-
ness under such circumstances.
Overall, the patterns are generally consistent with what is known about the public
sector. Therefore, the results can be generalized to the population of public sector
456 Michael J. Roszkowski and John E. Grable
workers rather than being limited to clients of financial planners. The details are as
(1) Females were overrepresented in the public sector [x2ð2Þ ¼ 7:85, p ¼ :020].
Women constituted 31.08% of the self-employed, 39.84% of the private sector, and
45.35% of the public sector.
Personal income differences between men and women were smaller in the public
sector. The ANOVA sex by sector interaction was significant [Fð2; 711Þ ¼ 6:08,
p ¼ :002] with a smaller sex difference in income occurring in the public sector
(Cohen’s d: self–employed ¼ :83; private ¼ :79; public ¼ :59).1Income differ-
ences between males and females within each sector were also statistically
significant in terms of the non-parametric Mann-Whitney test (self-employed:
U ¼ 3;347:50, p ¼ :000; private sector: U ¼ 8;881:50, p ¼ :000; public sector:
U ¼ 586:00, p ¼ :002). The average income ranks for males and females,
respectively, within each sector were as follows: 143.87 versus 82.63 for the self-
employed; 228.88 versus 134.93 in the private sector; and 49.76 versus 35.03 in
the public sector. The ratios of the ranks indicate a smaller differential in income
in the public sector relative to the other two sectors (74.11% for the self-employed,
69.63% for the private sector, and 42.05% for the public sector).
The distribution of personal income was tighter in the public sector. The standard
deviations on coded income were 1.23 for the private sector and 0.76 for the
public sector. Levene’s test showed that the variance in income in the public
sector was different from the private sector (F ¼ 77:73, p ¼ :001). In terms of the
Siegel-Tukey test, the probability of the difference being due to chance was also
fairly low, but it failed to reach conventional levels of statistical significance
(Z ¼ 21:54, p ¼ :125).
Public sector employees were slightly older [tð139Þ ¼ 22:13, p ¼ :035]. The
average age of public sector employees was 46.65 (SD ¼ 1:24), compared to 44.34
(SD ¼ 3:95) for private sector employees.
On the six-point scale used to represent level of education, the mean for the
public sector was 4.55 (SD ¼ 1:24), whereas for the private sector it was 3.95
(SD ¼ 1:08), a statistically significant difference [tð114Þ ¼ 24:15, p ¼ :001].
The average ranks were 300.91 and 230.52 for the public sector and private
sector, respectively, and the difference between private and public sector
employees was likewise statistically significant in terms of the Mann-Whitney test
(U ¼ 12;176:50, p ¼ :000).
Risk tolerance by sector and sex
Differences in risk tolerance as a function of employment sector are reported in Table 1.
Sex is included in the breakdown because there were clear disproportions in the
number of males and females employed in each sector, and it is well established that
sex is related to risk tolerance (see meta-analysis by Byrnes, Miller, & Schaefer,
1999). Risk tolerance differed significantly by both sector [Fð2; 737Þ ¼ 7:59, p ¼ :001]
and sex [Fð1; 737Þ ¼ 33:69, p ¼ :001], but the interaction was not significant
[Fð2; 737Þ ¼ 0:03, p ¼ :973]. The lack of an interaction is supported by the values of
1Generally, effect sizes can be interpreted as follows: .4 and below as small; .5–.7 as medium; and .8 and above as large.
Financial risk tolerance457
Cohen’s d on male–female risk tolerance differences, which are almost identical in the
three sectors (range of .51–.53). Public sector employees of both sexes exhibited lower
financial risk tolerance relative to their private sector counterparts. The difference was
on the order of about one-third of the common standard deviation (Cohen’s d was .32
for males, .31 for females, and .33 combined).
Controlling for possibly confounding variables
The data were also analysed in terms of several ordinary least squares regression models.
The aim was to determine the nature of the relationship between employment sector and
risk tolerance on an ‘as is’ basis and then again when control was introduced for the
effects of six variables known to be linked to risk tolerance (i.e. sex, age, education,
personal income, household income, and net wealth). The sample was reduced to 654
for net wealth, and 66 for household income). Employment sector and sex were
In the first model, employment sector was the sole predictor. The model was
significant [Fð2; 651Þ ¼ 11:35, p ¼ :001] and R2equalled .034 (.031 adjusted). The
standardized Beta (b) for public sector was 20.11 (t ¼ 22:68, p ¼ :008) and 0.12
(t ¼ 3:09, p ¼ :002) for self-employed, indicating that relative to the holdout category
(private sector), public sector employment was characterized by lower risk tolerance
whereas self-employment was associated with higher risk tolerance.
correlation (R2¼ :106 unadjusted/.097 adjusted) that was statistically significant
[Fð6; 647Þ ¼ 12:73,p ¼ :000].Withtheexception ofeducation(b ¼ 20:01,t ¼ 20:13,
p ¼ :895), allother regressionweightswerestatisticallysignificant or nearlyso:personal
income (b ¼ 0:25, t ¼ 3:18, p ¼ :002); sex (b ¼ 20:18, t ¼ 24:41, p ¼ :000); age
(b ¼ 20:17, t ¼ 24:18, p ¼ :000); household income (b ¼ 20:15, t ¼ 21:92,
p ¼ :056); and net wealth (b ¼ 0:10, t ¼ 1:95, p ¼ :052).
Table 1. Financial risk tolerance as a function of sector and sex
458 Michael J. Roszkowski and John E. Grable
The independent variables in the third model consisted of the six control variables
andthefocalvariable.Themodelresultedinastatisticallysignificant[Fð8; 645Þ ¼ 10:91,
p ¼ :000] multiple correlation (R2¼ :119 unadjusted/.108 adjusted). The DR2was
significant in relation to both the first model [Fð2; 645Þ ¼ 10:67, p ¼ :000] and the
second model [Fð2; 645Þ ¼ 4:98, p ¼ :007]. Of the six control variables, three had
significant Betas: personal income (b ¼ 0:21, t ¼ 2:69, p ¼ :007); female sex
(b ¼ 20:19, t ¼ 24:50, p ¼ :000); and age (b ¼ 20:16, t ¼ 24:02, p ¼ :000). The
standardized regression weights of two control variables, household income
(b ¼ 20:15, t ¼ 21:94, p ¼ :053) and net wealth (b ¼ 0:08, t ¼ 1:59, p ¼ :113),
although not reaching conventional levels of statistical significance, nonetheless had
fairly low probabilities of occurring by chance. The probability level of the standardized
regression weight for education (b ¼ 0:01) remained highly insignificant (t ¼ 0:36,
p ¼ :718). With all the control variables in the equation, the b for public sector dropped
to 20.08 (t ¼ 22:13, p ¼ :033) and the b for self-employed fell to 0.07 (t ¼ 1:81,
p ¼ :072). Holding the control variables constant, the association between public sector
employment and risk tolerance was abated, but not eliminated.
Relative to their private sector counterparts, public sector employees scored lower on a
test of financial risk tolerance which was administered as part of a financial planning
process. The lower risk tolerance of people employed in the public sector was evident
even after controlling for demographic variables related to risk tolerance. Considered
along with the other literature reported to date, it is reasonable to conclude that there is
some basis in reality to the claim of lower risk tolerance among public servants. It is not
merely a stereotype.
Table 2. Summary of hierarchical regression analysis predicting risk tolerance
Model 1 Model 2Model 3
Only focal variable Only control variables Control and focal variables
Public sector (b)
Personal income (b)
Sex (female) (b)
Household income (b)
Net wealth (b)
DR2Model 3 vs. Model 1
DR2Models 3 vs. Model 2
Note.*p # :05;**p # :01;***p # :001.
Financial risk tolerance 459
It has been argued that public sector employees may be more risk averse than their
private sector counterparts in work-related behaviours because the public sector is
charged with maintaining public welfare rather than maximizing yield on investments
(Bozeman & Kingsley, 1998). The items in the risk tolerance test dealt with personal
monetary issues and economic decisions, so this risk aversion extends beyond job-related
organizational demands and culture. Becker and Connor (2005) concluded that most
differences in values between the sectors are due to acculturation and not self-selection,
but risk tolerance was not a characteristic they studied. Interpreted according to
Schneider’s Attraction, Selection, and Attrition (ASA) model (Schneider, Goldstein, &
Smith, 1995), the cycle begins with the attraction of inherently less risk tolerant people
into government positions due to features such as greater job security and guaranteed
more likelyto select applicantswhoare similar to them in terms of risk tolerance. Even if
hired, an individual characterized by high risk tolerance is less likely to fit into the
organization’s culture and is therefore more likely to terminate. Thus, over time, the
organization becomes more homogenous with respect to risk tolerance.
Unfortunately, risk tolerance is value-laden. According to Lying (2005), today there is
an ‘: : : increased willingness to embrace risk taking in workplace settings where, in
earlier times, risk avoidance was the rule’ (p.25). It is questionable whether this is a
healthy trend given Reyna and Farley’s (2006, p.1) conclusion that ‘greater risk aversion
is generally adaptive, and that decision processes that support this aversion are more
advanced than those that support risk taking’. Preference for risk seeking over risk
avoidance cannot be the norm for all positions. Rather, the risk tolerance requirements
of the position need to be matched with the person’s inherent willingness to take risks.
Contrary to some opinions (e.g. Fitzgerald, 1988; Gingrich, 2005), both high risk
tolerance and low risk tolerance can constitute either competence or incompetence in
public sector employees, depending on the demands of the position.
One potentially relevant variable that remained uncontrolled in this study is
employment level. There was, however, an indirect control for it because personal
income and education were included in the regression analysis. Managerial positions
generally require higher levels of education and have greater compensation.[Holding
other relevant factors constant, Halek and Eisenhauer (2001) could not detect difference
in risk tolerance between managers and non-managerial employees.]
Suggestion for future research
Our results suggest that public sector employees, on the average, exhibit lower levels
of financial risk tolerance than private sector employees, especially compared
to individuals who are self-employed. It remains to be determined under what
circumstances risk aversion is a virtue and when it is a drawback. Sorely needed is
research assessing the relationship between risk tolerance and work performance on
specific jobs. Such information would take the discussion beyond conjecture when
value judgments are made about risk tolerance as an indicator of either competence or
460 Michael J. Roszkowski and John E. Grable
Becker, B. W., & Connor, P. E. (2005). Self-selection or socialization of public- and private-sector
managers? A cross-cultural values analysis. Journal of Business Research, 58, 111–113.
Bellante, D., & Link, A. N. (1981). Are public sector workers more risk averse than private sector
workers? Industrial and Labor Relations Review, 34, 408–412.
Blais, A.-R., & Weber, E. U. (2006). A domain-specific risk-taking (DOSPERT) scale for adult
populations. Judgment and Decision Making, 1, 33–47.
Bozeman, B., & Kingsley, G. (1998). Risk culture in public and private organizations. Public
Administration Review, 58, 109–118.
Byrnes, J. P., Miller, D. C., & Schaefer, W. D. (1999). Gender differences in risk taking: A meta-
analysis. Psychological Bulletin, 125, 367–383.
Dohmen, T., Falk, A., Huffman, D., Sunde, U., Schupp, J., & Wagner, G. G. (2005) Individual risk
attitudes: New evidence from a large, representative, experimentally-validated survey
(Discussion Paper No. 1730). Bonn, Germany: Forschungsinstitut zur Zukunft der
Arbeit /Institute for the Study of Labor.
Ekelund, J., Johansson, E., Jarvelin, M.-R., & Lichtermann, D. (2005). Self-employment and risk
aversion – evidence from psychological test data. Labour Economics, 12, 649–659.
Feinberg, R. M. (1977). Risk aversion, risk, and the duration of unemployment. Review of
Economics and Statistics, 59, 264–271.
Fitzgerald, R. (1988). When government goes private: Successful alternatives to public services.
New York: Universe Books.
Gingrich, N. (2005, April 10). Entrepreneurial public management as a replacement for
bureaucratic public administration: Getting government to move at the speed and
effectiveness of the information age. Washington, DC: American Enterprise Institute, Working
Paper #109. Retrieved November 23, 2007 from http://www.aei.org/publication22267/
Gore, A. (1993). Creating a government that works better and costs less: The report of the
national performance review. New York: Plume.
Halek, M., & Eisenhauer, J. G. (2001). Demography of risk aversion. Journal of Risk and
Insurance, 68, 1–24.
Hallahan, T. A., Faff, R. W., & McKenzie, M. D. (2004). An empirical investigation of personal
financial risk tolerance. Financial Services Review, 13, 57–78.
Hartog, J., Ferrer-i-Carbonell, A., & Jonker, N. (2002). Linking measured risk aversion to individual
characteristics. Kyklos, 55, 3–26.
Highhouse, S., & Yuce, P. (1996). Perspectives, perceptions, and risk-taking behavior.
Organizational Behavior and Human Decision Processes, 65, 159–167.
Jackson, D. N., Hourany, L., & Vidmar, N. J. (1972). A four-dimensional interpretation of risk taking.
Journal of Personality, 40, 483–501.
Lying, S. (2005). Sociology on the edge: Social theory and voluntary risk taking. In S. Lying (Ed.),
Edgework: The sociology of risk-taking (pp.17–50). New York: Routledge.
Prowle, M. J. (2000). Changing public sector: A practical management guide. Aldershot,
Hampshire: Gower Publishing Limited.
Reyna, V. F., & Farley, F. (2006). Risk and rationality in adolescent decision making: Implications for
theory, practice, and public policy. Psychological Science in the Public Interest, 7, 1–44.
Rohrmann, B. (2004, August). Risk attitude scales: Concepts and questionnaires, Project Report:
University of Melbourne, Australia. Retrieved May 6, 2008 from www.psych.unimelb.
Roszkowski, M. J. (1992). How to assess a client’s financial risk tolerance: The basics. Bryn Mawr,
PA: The American College.
Roszkowski, M. J., Davey, G., & Grable, J. E. (2005). Insights from psychology and psychometrics
on measuring risk tolerance. Journal of Financial Planning, 18(4), 66–75.
Roszkowski, M. J., Delaney, M. M., & Cordell, D. M. (2004). The comparability of husbands and
wives on financial risk tolerance. Journal of Personal Finance, 3(3), 129–144.
Financial risk tolerance461
Roszkowski, M. J., & Grable, J. E. (2005). Estimating risk tolerance: The degree of accuracy and the
paramorphic representations of the estimate. Financial Counseling and Planning, 16(2),
Roszkowski, M. J., & Grable, J. E. (2007). Self-assessments of risk tolerance by women and men.
Psychological Reports, 100, 795–802.
Roszkowski, M. J., & Snelbecker, G. E. (1990). Effects of ‘framing’ on measures of risk tolerance:
Financial planners are not immune. Journal of Behavioral Economics, 19, 237–246.
Schneider, B., Goldstein, H. W., & Smith, D. B. (1995). The ASA framework: An update. Personnel
Psychology, 48, 747–773.
Schoemaker, P. J. H. (1990). Are risk-attitudes related across payoff domains and response modes?
Management Science, 36, 1451–1463.
Slovic, P. (1972). Information processing, situation specificity and the generality of risk-taking
behavior. Journal of Personality and Social Psychology, 22, 128–134.
Starr, P. (1989). Review of the book when government goes private: Successful alternatives to
public services. Political Science Quarterly, 104, 357–358.
Utgoff, K. C. (1983). Compensation levels and quit rates in the public sector. Journal of Human
Resources, 18, 394–406.
Walls, M. R. (2005). Measuring and utilizing corporate risk tolerance to improve investment
decision making. Engineering Economist, 50, 361–376.
Wright, A., & de Joux, V. (2003, May). Case studies of innovation in the public service (Amherst
Group Ltd. Working Paper). Retrieved January 20, 2008 from http://treasury.govt.nz/
Yook, K. C., & Everett, R. (2003). Assessing risk tolerance: Questioning the questionnaire method.
Journal of Financial Planning, 16(8), 48–55.
Received 7 February 2008; revised version received 30 June 2008
Sample test items by category
Guaranteed versus probable outcome
Assume you’re an executive. Your company offers you two ways of collecting your
bonus: either cash equivalent to 6 months’ salary or a stock option with a 50–50
chance of either doubling in value or becoming worthless in the next year. Which
would you take? Possible answers: definitely the cash; probably the cash; not sure;
probably the stock option; definitely the stock option.
Minimal probability of success
Assume you are a contestant on a TV game show. After winning a prize that’s
equivalent to 1 year’s salary, you are offered the option of walking away with this
prize money or taking a chance on either doubling it or losing it all. What are the
odds of success that you would require before agreeing to accept this gamble.
Possible answers: would not take the bet no matter what the odds; 9 in 10; 8 in 10;
7 in 10; 6 in 10; 5 in 10; 4 in 10; 3 in 10; 2 in 10; 1 in 10.
You are offered an investment in which you stand an even chance of either losing
half your personal net worth or making a certain amount of money. What’s the
lowest return you would need in order to make such an investment? Possible
462 Michael J. Roszkowski and John E. Grable
answers: I would not make the investment no matter what the rate of return; Download full-text
quadruple my net worth; triple my net worth; double my net worth; less than
double my net worth.
When you think of the word risk in an investment context, which of the following
Is it more important to be protected from inflation or to be assured of the safety of
your principal? Possible answers: much more important to be assured of the safety
of my principal; somewhat more important to be assured of the safety of my
principal; somewhat more important to be protected from inflation; much more
important to be protected from inflation.
Preferences for investment vehicles
Diversification is typically the soundest investment strategy. However, suppose an
all the money in only one of the following investments. Which one would you select?
Possible answers: savings account; mutual fund (moderate growth); blue-chip
common stock; limited partnership; naked option/commodities futures contract.
Reaction to sample portfolio
Which of the following investment portfolios do you find most appealing?
Possible answers: (a) 60% in low-risk, 30% in medium-risk, and 10% in high-risk
investments; (b) 30% in low-risk, 40% in medium risk, and 30% in high
risk investments; and (c) 10% in low-risk, 50% in medium-risk, and 40% in high-
Life style characteristics
Have you ever borrowed money in order to make an investment (other than a home-
mortgage loan)? Possible answers: no, yes
How would you rate your willingness to take investment risks in comparison to the
general population? Possible Answers: extremely low risk taker; very low risk taker;
low risk taker; average risk taker; high risk taker; very high risk taker; extremely
high risk taker.
Financial risk tolerance463