Socioeconomic Institute
Sozialökonomisches Institut
Working Paper No. 1007
When Are Preferences Consistent? The Effects of
Task Familiarity and Contextual Cues on Revealed
and Stated Preferences
Felix Schläpfer, Baruch Fischhoff
August 2010
Socioeconomic Institute
University of Zurich
Working Paper No. 1007
When Are Preferences Consistent? The Effects of Task Familiarity and
Contextual Cues on Revealed and Stated Preferences
August 2010
Author's address:
Felix Schläpfer
E-mail: felix.schlaepfer@soi.uzh.ch
Baruch Fischhoff
E.mail: baruch@cmu.edu
Publisher
Sozialökonomisches Institut
Bibliothek (Working Paper)
Rämistrasse 71
CH-8006 Zürich
Phone: +41-44-634 21 37
Fax: +41-44-634 49 82
URL: www.soi.uzh.ch
E-mail: soilib@soi.uzh.ch
When Are Preferences Consistent?
The Effects of Task Familiarity and Contextual Cues
on Revealed and Stated Preferences
Felix Schläpfer1* and Baruch Fischhoff2
1
Department of Economics, University of Zurich and
Institute for Environmental Decisions, Federal Institute of Technology ETH
Zurich; Switzerland
2
Department of Social and Decision Sciences
Department of Engineering and Public Policy
Carnegie Mellon University
Pittsburgh, PA 15213-3890; USA
Tel: +1-412-268-3246; email: baruch@cmu.edu
*Address for correspondence: Department of Economics, University of Zurich,
Hottingerstr. 10, 8092 Zurich; Switzerland, E-mail: felix.schlaepfer@soi.uzh.ch.
June 2010
Abstract
Traditionally, economists make a sharp distinction between stated and revealed
preferences, viewing the latter as more fully meeting the assumptions of economic
analysis. Here, we consider one form of empirical evidence regarding this belief: the
consistency of choices in stated and revealed preference tasks. We show that both kinds
of task can produce consistent choices, suggesting that both can measure underlying
preferences, if necessary conditions are met. We propose that a necessary condition is
that task be either familiar to those facing it or offer contextual cues that substitute for
familiarity, such as prices in competitive markets or recommendations from trusted,
knowledgeable sources. We show that how well decision makers achieve such
understanding is often confounded with the method that researchers use. Considering
task familiarity not only clarifies some of the conflicting evidence regarding revealed and
stated preference methods, but raises potentially productive questions regarding the roles
of social institutions in shaping preferences.
Keywords: Consistency, contingent valuation, framing, public goods, revealed
preferences, stated preferences, validity.
JEL codes: D01, D03, Q51
We are grateful to James Murphy and his co-authors for making the dataset of their metaanalysis available.
1
1. Introduction
Economic analysis traditionally relies on revealed choice, arguing that, for practical
purposes, individuals‟ preferences meet standard economic assumptions (e.g., Friedman
1953). Economists have maintained this position despite growing challenges (e.g.,
Simon 1956; Conlisk 1996; McFadden 2001; Ariely et al., 2003; Kahneman, 2003;
Camerer and Fehr, 2006), partly because they lack an account of when these violations of
its assumptions matter and partly because they distrust the alternative, stated preference
methods. On theoretical grounds, many economists are wary of stated preferences
because they know how, in theory, individuals could respond strategically to questions
about alternative levels of public good provision (Gibbard, 1973; Satterthwaite, 1975).
On empirical grounds, many economists note the non-rational behavior first identified in
stated preference studies (e.g., Kahneman and Tversky, 1979).
Despite the strength of this opposition, the evidence against stated preference
methods is fairly limited (Dominitz and Manski 1997). Many anomalies have been
replicated in tasks with monetary stakes, indicating analogous problems with at least
some revealed preferences (Camerer and Hogarth, 1999). Moreover, the largest body of
evidence suggesting problems with stated preferences comes from the somewhat
anomalous context of contingent valuation surveys (Arrow et al. 1993; Mitchell and
Carson, 1989), in which respondents state their willingness to pay for public goods that
are far from their personal experience and have not been publicly discussed. When those
tasks are described in simple terms, they force respondents to figure out what exactly
they are evaluating (Fischhoff and Furby, 1988). When the tasks are described in specific
terms, they pose cognitive demands on respondents. It is hard for individuals to respond
consistently when they do not know what question they are answering or cannot keep all
its details in mind.
A continuing debate among dedicated scientists suggests that there is no single,
general answer to the question of the relative validity of revealed and stated preference
methods. However, there may be answers to where or when each is more valid (Fischhoff
1991, p. 846; Slembeck and Tyran 2004, p. 683; Druckman 2004, p. 683; Shogren 2006,
p. 169). Here, we examine the cognitive tractability of revealed and stated preference
tasks selected to vary on three features of the choice task: (a) whether the good is
familiar, (b) whether informative contextual cues accompany the task, and (c) whether
the good is private or public.
Section 2 defines our criterion for consistent preferences. Section 3 presents the
three features of the choice task. Section 4 examines selected studies representing the
different possible combinations of these features, in terms of the consistency of the stated
and revealed preferences that they produce. Section 5 reports statistical analyses of the
roles of these features in predicting one form of consistency in a large set of studies.
Section 6 concludes with a proposal for viewing revealed and stated preferences as
complementary, rather than competing approaches, illuminating different aspects of
preference-forming processes.
2
2. Empirical consistency criteria
Economic theory and analysis are based on the axiomatic assumptions that preferences
are complete and transitive. Preferences that satisfied these assumptions should
demonstrate the kind of consistency that psychologists call construct validity. They
should be (1) insensitive to irrelevant information, and (2) appropriately sensitive to
relevant information.
A well-known violation of the first condition arises with anchoring effects, in
which responses are affected by arbitrary numbers (e.g., think of the last two digits of
your Social Security number, before answering). An example is the anchoring effect in
responses to hypothetical referendum questions in contingent valuation surveys with
“alternative bid designs”: The binary (yes/no) responses express lower values when the
question is, “Would you pay [$1/$2/3$, etc]?” than when it is, “Would you pay
[$10/$20/$30, etc.]?” (e.g. McFadden 2001). An anchor can also convey information,
such as what researchers believe the value to be. If so, then responses should be sensitive
to it.
It is harder to establish whether sensitivity to relevant information is appropriate.
Scope tests ask whether preferences increase with increases in the quantity of a good.
However, these are vague tests, without specifying how great the sensitivity should be.
Moreover, subadditivity, as in the finding that 57 wilderness areas were valued only
about fifty percent more than one wilderness area (McFadden, 1994), has been explained
in terms of “substitution effects and diminishing marginal rates of substitution”
(Hahneman‟s (1994, p. 35). Hahneman‟s argument implies that respondents believe that
the 57 wilderness areas are almost indistinguishable (hence are substitutable) and that
there is little extra value in having an additional 56 (diminishing marginal utility).
Thus, construct validity entails a theoretical argument, regarding which features
matter to people (hence should evoke sensitivity) and how methodologically sound
studies are. Below, we outline construct validity arguments for the eight classes of study
mentioned above (i.e., all combinations of familiar and unfamiliar, public and private
goods, with and without informative contextual cues).
We restrict the discussion to studies with between-subject designs, where different
respondents evaluate, say, different amounts of a good. Within-subject designs, where
each respondent evaluates multiple versions of a good, inevitably provide hints as to what
kinds of consistency researchers expect (Poulton, 1968, 1994; Frederick and Fischhoff
1998, Amir and Levav 2008).
3. Classification of choice domains
In the analysis below, we use the following definitions:
(1) familiarity: extent of experience with the good (e.g., the proposed change of
the specific wilderness areas).
(2) informativeness of contextual cues: how well they convey the issues relevant
to the choice, either directly or through surrogates, such as market prices, medical advice,
or public debates.
3
(3) private vs. public goods: “private goods” are for personal consumption;
“public goods” are public services or amenities, as might be evaluated with votes in
hypothetical referenda or on rating scales.
“Revealed preferences” (RP) refers to consequential choices. With private goods,
they involve actual transactions. With public goods, they require implementation rules,
such as a simple majority voting rule with a coercive payment mechanism.
“Stated preferences” (SP) refers to hypothetical choices without a clear, coercive
implementation rule.
Table 1 summarizes the 2x2x2x2 design of the analyses that follow.
[Table 1 about here]
4. Exploratory analysis of consistency across choice domains
4.1. Familiar private goods with informative contextual cues
Many marketing studies and observational studies of consumer behavior have these
properties. Taken together, they provide overwhelming evidence of price sensitivity, an
essential form of choice consistency, in the preferences revealed in actual markets for
gasoline, computers, air trips and many other goods. Experimental studies of choice
consistency, in such domains are inherently rare because natural choice sets and cues
(e.g., equilibrium prices, advertisements, consumer advice) cannot be manipulated
without creating unnatural cues that might elicit unrepresentative behavior. For example,
manipulating actual prices can send misleading (or just confusing) signals about quality,
producing inconsistent choices. Conversely, consistent choices can be overdetermined,
potentially reflecting habitual responses, contextual cues, or social pressure, as well as
tapping robust preferences (List 2004).
Revealed preferences. In one well-known experimental test, with a familiar,
private choice, and clear (contextual) price information, Dickie et al. (1987) offered
boxes of strawberries door-to-door at different prices. The resulting demand curve
suggests sensible sensitivity to price information, perhaps reflecting personal experience
with the good, perhaps reflecting contextual cues from prices in ordinary markets.
Stated preferences. Dickie et al. (1987) also asked other people the hypothetical
question of how many boxes they would buy at the prices offered in the revealed
preference portion of their study. The hypothetical choices produced a demand curve
similar to that for the actual ones.
Indirect evidence of choice consistency with familiar private goods and
informative contextual cues is seen in the weaker endowment effect found with
experienced traders in market settings, compared to that found with inexperienced
traders, unfamiliar with the good and market context (List 2004).
4.2. Private goods, directly experienced but without informative contextual cues
Strong experimental tests of construct validity, for such choices, are difficult because it is
hard to remove natural context cues (e.g., competitive market prices) without creating an
unnatural choice, potentially inducing unrepresentative behavior.
4
Revealed preferences. Removing familiar contexts is, however, a staple of
marketers, hoping to entice consumers into making choices inconsistent with their normal
preferences. To this end, casinos use chips (rather than money), have no clocks, and ply
gamblers with drinks. Gamblers‟ remorse is a special, sometimes tragic case of
inconsistent choices in revealed preferences.
Stated preferences. Experiments involving familiar private goods, but without
their normal informational context are also a staple of psychological and behavioural
economics researchers, seeking inconsistent choices that will demonstrate inappropriate
sensitivity to contextual cues. Loomis et al. (1997), however, showed cases in which
choices regarding a fairly familiar good (a wildlife art print) offered in unusual contexts
were sensitive to experimental variations in price. In a demonstration of the power of the
cues provided by within-subject designs, Frederick and Fischhoff (1998) found that stated
willingness to pay for toilet paper was 2.5 times more sensitive to quantity when it was
manipulated within rather than between subjects, indicating either that the former choices
were too sensitive or that the latter were insufficiently sensitive (Frederick and Fischhoff
1998). Levin and Gaeth (1988) found a case in which great familiarity with a product
(ground beef) overcame an experimental attempt to manipulate contextual cues – seen in
failure to find a familiar framing effect.
4.3. Private goods, without direct experience, but with informative contextual cues
Here, although individuals are unfamiliar with the goods themselves, they have access to
useful cues, for instance from market context or from a trustworthy source with enough
information to make consistent choices possible. That source might be a knowledgeable
relative, consumer organization, or government regulator. The source must know enough
about the decision makers to identify the facts (and uncertainties) most relevant to their
choices. The relevant knowledge might include predicting how those individuals will
experience novel goods. Taking advantage of such sources requires being able to assess
their knowledge and honesty. Many investors feel that they have misplaced their trust in
financial advisors, who have proven to lack one property or the other.
Revealed preferences. Slembeck and Tyran (2004) find that the „three-door
anomaly‟, a systematic violation of payoff-maximizing individual behavior in an
experimental game, is eliminated when individuals have access to contextual cues in the
form of communications from competing, better-performing participants in the game.
Stated preferences. Haward et al. (2008) presented subjects with the (fortunately)
unfamiliar hypothetical choice between resuscitation and palliative (or “comfort”) care
for extremely premature infants. However, even an impoverished description provided
enough information for individuals who describe themselves as “highly religious” to
resist a common framing manipulation. Namely, they responded consistently whether the
outcomes were described in terms of survival and health or death and disability. Thus,
some preferences reveal themselves, as soon as individuals are asked to state them.
4.4. Unfamiliar private goods, without informative contextual cues
Life sometimes surprises people with unfamiliar choices without any clarifying context.
Experimental researchers do that as a matter of course, as they create novel formulations
5
designed to inform theoretical questions. Indeed, there can be a “curse of cleverness,”
whereby the more innovative a task is, the less individuals can perform it consistently.
When studies seek to highlight processes that are not easily seen in everyday life, the
prevalence of (in)consistent preferences in them says little about their prevalence under
more normal circumstances.
Revealed preferences. Preferences in this domain have been shown to be easily
influenced by irrelevant information and not appropriately sensitive to relevant
information. Behavioral economists have amply identified cases in which market choices
are inconsistent with one another, not to mention consumers‟ best interests (Ariely, 2008;
Thaler and Sunstein, 2008). For example, Ariely et al. (2003) showed that willingness to
pay for goods like bottles of wine (by students unfamiliar with the specific kinds) was
affected by obviously arbitrary numbers (digits from their social security cards), to which
their attention had previously been drawn. Tilman and Slembeck (2004) found that
participants in an unfamiliar game did not appropriately update their preferences, given
new information, unless they had access to informative contextual cues from better
informed participants (cf. 4.3). One non-monetary example is the much greater rate of
organ donation when it is the default, compared to where one must opt into it (Johnson et
al., 2003). Thaler and Sunstein (2008) show the power of such defaults, with unfamiliar
private goods and impoverished contexts.
Stated preferences. Studies with lightly described private goods are a staple of
stated preference studies designed to demonstrate respondents‟ sensitivity to normatively
irrelevant cues. For example, Coupey et al. (1998) show sensitivity (preference reversals)
to task details that should be irrelevant (choice vs. matching format) in stated preferences
for unfamiliar products like “reverse osmosis filters.” Even when the good is ostensibly
familiar (e.g., a coffee mug in an endowment effect study; Kahneman et al., 1990), the
social setting of the specific experiment will not be. If the preference applies to an
interpersonal exchange, then the good cannot be isolated from it – meaning the choice in
an experiment is unfamiliar even when the good is common.
One operational test for whether cues are “informative” is whether preferences are
affected by adding them in a non-coercive way. For example, Huffman et al. (2007)
found that preferences for GM food stabilized when accompanied by information about
the positions of politically involved third parties, who could be assumed to have worked
through the issues from their perspective.
4.5. Familiar public goods, with informative contextual cues
At the end of idealized political processes, the public goods at issue are familiar and
citizens‟ choices are informed (e.g., Lupia and Matsusaka, 2004). That reality, of course,
is always something less. Indeed, political scientists often focus on “issue publics,”
individuals who care enough to familiarize themselves with a public good and its context
(e.g., how will its costs and benefits be distributed, what precedents will it set). The issue
public is held to interpret issues for others sharing their values. Whether even they can
master the specific issue and context is an empirical question (Fischhoff, 1991).
Revealed preferences. Preferences for public goods are most clearly revealed in
behaviours like voting, canvassing, and making political contributions. Experiments are
uncommon, given that public goods often become familiar through political processes
6
that are not amenable to the experimental manipulations needed for consistency tests. As
a result, such tests tend to be limited to choices about contributions to public goods.
Such studies sometimes find inconsistent revealed preferences. For example, even after a
withering national election campaign, voter turnout (a measure of caring about the result)
could still be increased by several percentage points by the simple intervention of asking
registered voters where they would be prior to voting (Nickerson and Rogers, 2010).
Envisioning that place and the path from it to the polls provided context that was
otherwise missing. A growing research area examines how alternative frames can affect
charitable contributions for seemingly familiar public goods with ample contextual cues
(e.g., people give more for identified beneficiaries than for the general causes that they
represent; Small et al., 2007).
Stated preferences. Attitude surveys about public goods tend to provide few
informative contextual cues, in deference to reducing respondents‟ cognitive load. As a
result, they will be discussed below (4.6). Chong and Druckman (2007) found that, even
with relatively familiar public policy issues, which had been widely covered in the news
media, stated preferences were still subject to framing effects, until respondents received
(forged) arguments from political information providers (see also Druckman, 2004;
Druckman and Nelson, 2003; Tomz and Sniderman, 2005).
Although rarely used in experimental economics or psychology, deliberative
processes are often used in applied settings to help people to articulate their preferences,
by hearing and reflecting on alternative perspectives, chosen to provide contextual
information. Such clarifying interactions are central to decision analysis (Clemen &
Reilly, 2002), which continues until clients‟ preferences pass consistency tests. Morgan
et al. (2001) describe the consistency achieved through a complex group deliberation
about risk priorities. Well-designed deliberative processes show the limits to
consistency, while providing only indirect evidence regarding its natural occurrence.
4.6. Familiar public goods, without informative contextual cues
Preferences here involve public goods that may or may not seem to require the additional
information provided by contextual cues.
Revealed preferences. Political advertising counts on the malleability of voting
behavior for issues that have been regularly, but only superficially aired. Its impacts are
limited by the power of party identification. People do not need to know very much
about specific issues, if they follow party leaders who take consistent positions. Political
scientists study how individuals respond to such naturally occurring circumstances.
Researchers are rarely in a position to create them.
Stated preferences. Public opinion polls normally fall into this category, letting
the issues speak for themselves. Polls released to the general public sometimes elicit the
strength of respondents‟ preferences, but rarely examine their internal consistency.
Proprietary polls done for political candidates often look for inconsistency, seeking the
context that evokes the most desirable preferences. Several of the experimental studies
on political behaviour mentioned in Section 4.5 found that, absent cues such as party
labels, preferences tended to be inconsistent with one another (Tomz and Sniderman
2005) or with expressed political attitudes (e.g. Druckman and Nelson 2003).
7
4.7. Unfamiliar public goods, with informative contextual cues
It is a triumph of political deliberations when they take public goods from obscurity to
fully informed status. As mentioned, the reality is often less satisfying. Even voters who
know the main arguments of both sides still may not be comfortable making their
decisions without considering contextual cues, such as party labels.
Revealed preferences. In actual voting, although experimentally manipulating
information is impossible, it is possible to assess the consistency of preferences for voters
with different amounts of contextual information. Lupia (1994) finds that otherwise
uninformed voters who knew the positions of interested parties voted as consistently as
well-informed voters with similar stakes riding on an issue.
Stated preferences. Researchers who present unfamiliar public good issues often
provide little context, preferring to evoke intuitive responses. Indeed, many social
scientists are averse to providing context, lest they unnaturally bias or deepen
respondents‟ preference (Fischhoff, 1991). Other researchers embrace reactive
measurement, seeking to change respondents is a disciplined way. For example,
deliberative polling (Fishkin, 1991; Luskin et al., 1997) tries to present balanced
perspectives on an unfamiliar issue, in the context of group discussions moderated to
facilitate airing and clarifying of alternative perspectives. Like decision analysis, these
processes proceed until they reach the most consistent preferences possible, hence
provide only indirect evidence regarding the natural prevalence of consistent preferences.
In a demonstration of the power of trusted authorities to provide useful contextual
cues for unfamiliar public goods, Druckman (2001) found that forged party positions
reduced framing effects with Tversky and Kahneman‟s (1981) Asian disease problem. In
experimental studies, providing actual party positions afforded hypothetical voters
enough contextual information to form consistent preferences, as reflected in their
sensitivity to price (Schläpfer et al., 2008). The predictive validity of many pre-election
polls suggests that they can evoke the context of the actual voting, although they seldom
assess how well voters understand the context (Vossler and Kerkvliet 2003).
4.8
Unfamiliar public goods, without informative contextual cues
Psychologists have long known that most survey respondents offer some opinion on
almost any question, no matter how unfamiliar the topic or how impoverished the context
cues. A classic example is that 70% of survey respondents expressed an opinion on a
fictitious Metallic Metals Act although no such act existed and, therefore, the respondents
could have no actual knowledge (Gill, 1947; cited in Plous, 1991, p. 55). Some such
respondents do not want to admit ignorance. Some assume that they can guess what the
issue is, based on the brief description.
In such cases, the good is not only unfamiliar, but undefined. Fischhoff and
Furby (1988) present one framework for specifying choices well enough that consistent
preferences are possible. If everything needs to be spelled out, then the cognitive load
could be unmanageable. However, informative contextual cues can reduce that load by
directing respondents to critical issues and proposing possible decision rules.
Revealed preferences. Despite theoretical arguments regarding the informational
value of advertising, those using it often seek to manipulate contextual cues in their favor.
8
Their opportunities are greatest when the good is unfamiliar or misunderstood. When
marketers are successful, the contextual cues determine which features are essential to the
good‟s definition. For instance, Great Britain‟s close national referendum on joining the
EU was reportedly tipped by framing it as “stay in Europe,” rather than “join Europe,” a
distinction discovered through research into what is now called the “status quo bias.”
Contingent valuation surveys are sometimes described as producing results that
will affect policies regarding public goods. If that claim is accepted, then responses
might reveal (rather than just state) preferences. It also heightens theoretical concern
about strategic responses, understating or overstating a good‟s value, based on how
respondents assess the opportunities for slanting results to their advantage (Gibbard,
1973; Satterthwaite, 1975; see also Green et al., 1998; Champ et al. 2002; Flores and
Strong, 2007; Schläpfer and Bräuer, 2007; Schläpfer and Schmitt, 2007).
Stated preferences. Manipulation checks ask respondents to recall and make
inferences about details of questions that they have just answered. Without them, there is
no direct way of telling how familiar the issue was. The power of framing, and other
context effects, arises from respondents‟ need to rely on contextual cues, in order to
articulate reasoned answers to incompletely specified questions. In many experimental
studies, those cues are deliberately uninformative, so that researchers can demonstrate an
“effect,” whereby people respond to irrelevant information. The psychological and
contingent valuation literatures have many such examples (Tversky and Kahneman,
1981; Fischhoff and Manski, 1999; Lichtenstein and Slovic, 2005; Mc Fadden 2001).
These demonstrations show how inconsistent preferences can be with goods that are both
incompletely specified goods and without informative contextual cues. Ironically, the
same lack of information and context can also encourage insensitivity to relevant
information, such as the quantity or „scope‟ of a good (Boyle et al., 1994; McFadden,
1994). One of the few stated preference studies with manipulation checks found that
preferences were more consistent when evaluated in terms of what respondents believed
about the goods, compared to what was explicitly stated (Fischhoff et al., 1993).
4.9. Discussion
Thus, both consistent and inconsistent preferences have been found with both revealed
and stated preferences, in most of these eight domains. As the discussion suggested,
some domains are much more heavily studied than others, reflecting research
opportunities (e.g., for experimental manipulations) and researcher predilections (e.g., for
minimal or hypothetical descriptions). Because these predilections differ by academic
discipline, researchers typically work in a subset of possible domains and see the patterns
of consistency that it typically produces. It would only be natural, if they came to see
such results as typical in the world.
The present analysis finds that preference consistency seems to increase as
choices become better understood, whether through familiarity with the good or receipt
of informative contextual cues. That is seen in contrasts across domains, created to vary
on these two features, and within domains, for choices varying on these two factors. The
analysis did not suggest any systematic impacts of the other two features: whether the
good was public or private and whether preferences were revealed or stated. In studies
(and in the world), public and private goods tend to vary in how well they are understood.
9
So do revealed and stated preference tasks. As a result, differences in preference
consistency reflect those covariates (familiarity, informative contextual cues), not the
public/private or revealed/stated distinction.
The following section examines whether these patterns emerge in the universe of
preferences represented in one set of published studies.
5. Explaining “hypothetical bias” in stated preferences: a meta-analysis regression
The preceding analysis suggested that preference consistency depends on how well
choice tasks are understood. That depends, in turn, on the familiarity of the good and the
informativeness of the accompanying contextual cues. Preference consistency does not,
however, depend on whether the good is private or public good or whether the
preferences are revealed or stated.
The studies used as examples were selected to illustrate these cases. This section
examines these patterns in a set of studies collected for other purposes. Specifically, it
asks how well three of these features (public/private, familiarity, cue informativeness)
predict a form of consistency defined by the fourth: whether stated and revealed
preferences agree. A standard measure of such consistency is hypothetical bias, the ratio
of willingness to pay expressed in stated preference tasks (with hypothetical choices) and
in revealed preference tasks (involving actual payments). These ratios vary widely
enough, across studies, to justify the search for covariates.
The set of studies is drawn from Murphy et al. (2005). They used OLS to predict
hypothetical bias from six factors: whether the good is public or private, whether the
preferences are from students or other (perhaps more strongly motivated) individuals,
whether preferences are elicited from individuals or (possibly more thoughtful) groups,
whether preferences were expressed or an open-ended scale or choices among fixed
options that might suggest expected answers (e.g., possible prices); whether the design
was between-subjects or within-subjects (which might have suggested expected answers);
and whether preferences were somehow “calibrated” in order to be more consistent (see
Murphy et al. 2005, p. 318). They considered 28 studies, with 83 observations, selected
because of their accessibility in the peer-reviewed literature (see Appendix, Table A1
Model 1a in Table 2 shows regression results for all 83 observations included in
the original analysis. The only variable that strongly predicted a smaller hypothetical bias
was using a within-subject design. Murphy et al. note that “attempts to identify other
factors that may be associated with hypothetical bias yielded mixed results.”
For these purposes, two of these six predictors seem problematic. As mentioned,
within-subject designs can signal the features that interest the researchers, thereby
inducing consistency (Fischhoff and Bar Hillel, 1984). Calibration of responses attempts
to eliminate the phenomenon of interest. As a result, we reran the regression after
1
The original dataset is available at: http://faculty.cbpp.uaa.alaska.edu/jmurphy/meta/meta.html#data
(accessed 8 June, 2010.
10
eliminating the 44 observations having within-subject designs or calibration.2 Model 2a
shows that hypothetical bias is somewhat smaller with private goods and choice tasks.
We now repeat these regressions, adding the two variables central to our analysis:
familiarity with the good and access to informative contextual cues.3 Table A in the
Appendix shows our coding of the 28 studies (with 83 observations) on these two
additional variables. Among them, 13 had unfamiliar goods without informative context,
7 had unfamiliar goods with informative context, 5 had familiar goods without
informative context, and 3 had familiar goods with informative context.
Because Model 1b includes with-subject designs and calibration, we focus on
Model 2b. In it, hypothetical bias is significantly smaller for familiar goods and for
informative contexts. The significant interaction indicates that familiarity and context are
somewhat redundant contributors to more consistent preferences: contextual cues reduce
hypothetical bias less with familiar goods than with unfamiliar ones. Conversely,
familiarity reduces bias less when people have access to contextual cues.
[Table 2 about here]
6. Conclusions
Economists have traditionally made a strong distinction between revealed and stated
preferences. The former are usually held to meet the axiomatic assumptions for
preferences, while the latter are often dismissed for their perceived inconsistency.
However, as seen in Section 4, both consistent and inconsistent choices are found with
both revealed and stated preferences. We propose that consistent preferences are more
likely when individuals understand tasks better, regardless of whether the choices involve
public or private goods and regardless of whether the preferences are revealed or stated.
We propose further that understanding can come two complementary routes. One
is familiarity with the good. The second is receipt of informative contextual cues. If
revealed preferences studies are more likely to demonstrate consistent preferences, it is
because those studies are more likely to involve well-understood tasks. The analysis in
Section 4 provides reasons why this might be the case. For example, stated preference
studies can (and often do) use choices invented for research purposes. Conversely,
revealed preference tasks do sometimes occur in life, with relatively fixed and familiar
properties and context.
We supplement this illustrative discussion with statistical analysis of a sample of
studies, collected by Murphy et al. (2005) for other purposes (thereby reducing the
chances that they were selected to confirm our hypotheses). In their full sample, the only
predictor of choice consistency was using a within-subject design, which can signal how
to be consistent. Removing those studies, along with ones that “calibrated” responses to
be more consistent, revealed two predictors of consistency: using private (rather than
public) goods and eliciting preferences with choice (rather than rating scale) responses.
However, those predictions vanished when familiarity and contextual cues were added to
2
The dropped observations include 17 within-subject observations from Sinden (1988), multiple
observations in Blumenschein et al. (2001) and Johannesson et al. (1998), and calibrated observations by
List (2001) and List (2003), cited in Murphy et al. 2005.
3
The correlation between the two variables is r=0.16.
11
the equations. Thus, the other two predictors seem to have mattered because they were
more likely in studies postulated as having better understood tasks.
If this result is robust, then concern over revealed versus stated preferences has
been misplaced, obscuring the more important question of when people understand tasks
well enough to make consistent responses. We propose that one class of determinants
has to do with how familiar individuals are with a good. Whether a good is public or
private may predict familiarity, for researchers who focus on familiar private goods.
Private goods may also have a more natural set of informative contextual cues, such as
prices on competitive markets or popularity among personally relevant individuals.
However, those correlations are not exclusive. Political campaigns and policy debates
can make public goods quite familiar, as well as generate party positions providing
important contextual cues. Similarly, researchers and marketers can create entirely novel
private goods for decision makers‟ consideration.
From this perspective, the critical research question becomes what are the
conditions that favor understanding choice tasks and, hence, consistent preferences. The
answers to that question can be used prescriptively, when the goal is to create those
conditions (Fischhoff et al. 1999; Keeney and Raiffa, 1993; Payne et al. 1999). The
answers can be used descriptively when the goal is to understand how well natural
circumstances have allowed individuals to master choice tasks (Fischhoff et al., 1980;
Lichtenstein and Slovic, 2005). In a mature research program, the two activities would
be integrated. Descriptive research would direct prescriptive efforts to make tasks clear,
then assess their success. Prescriptive work would identify barriers to understanding that
could benefit from and inform descriptive research. That integrated research program
could, for example, examine the role of consequentiality of choices on their consistency.
In other domains, it has proven theoretically productive to ask when increased incentives
enhance or degrade performance or have no effect at all, absent help in understanding
tasks (Camerer and Hogarth, 1999; Lerner and Tetlcok, 1999.; Milkman et al., 2009).
The integration of descriptive and prescriptive research has long been part of the
behavioural decision research, which grew out of mathematical, cognitive, and
engineering psychology in the 1950s (Edwards, 1954; Kahneman et al., 1982
vonWinterfeldt and Edwards, 1986). It seems to be a historical coincidence that such
integration has lagged as economics has strengthened its behavioural foundations. One
barrier has been the theoretically founded scepticism regarding stated preferences,
reinforced by the fact that so many stated preference studies have involved unfamiliar
goods presented with limited contextual cues (Manski 2000). A second barrier has been
the focus of revealed preference studies involving familiar goods where these issues were
less relevant. A third barrier has been the institutional pressures to defend many stated
preference estimates, for example, when contingent valuations are used to defend damage
settlements. Such settings encourage doing the best study possible, then denying any
residual problems.
All researchers must work within the constraints of their funding and their
discipline‟s norms. However, all those concerned about assessing and informing
preferences could benefit from better fundamental understanding of the conditions that
favor consistent preferences. That basic science can exploit and inform research in
economics, psychology, political science, survey research, and other fields. It raises
questions such as what experiences make a good familiar enough that people can
12
anticipate their future experiences, which sources are trusted for providing contextual
cues, which cues are most informative, and what insight do individuals have into how
well they have articulated their own preferences.
13
References
Amir, O. and Levav, J., 2008. Choice construction versus preference construction: The
instability of preferences learned in context. Journal of Marketing Research,
45(2), 145-158.
Ariely, D., 2008. Predictably irrational: the hidden forces that shape our decisions,
HarperCollins, New York.
Ariely, D., Loewenstein, G. and Prelec, D., 2003. "Coherent arbitraryness": stable
demand curves without stable preferences. Quarterly Journal of Economics,
118(1), 73-105.
Arrow, K., Solow, R.., Portney, P. R., Leamer, E. E., Radner, R. Schuman, H., Solow, R.,
1993. Report of the NOAA Panel on contingent valuation. Federal Register,
58(10), 4601-4644.
Boyle, K. J., Desvousges, W. H., Johnson, F. R., Dunford, R. W. and Hudson, S. P.,
1994. An investigation of part - whole biases in contingent-valuation studies.
Journal of Environmental Economics and Management, 27(1), 64-83.
Camerer, C. F. and Fehr, E., 2006. When does “economic man” dominate social
behavior? Science, 311(5757), 47-52.
Camerer, C.F., and Hogarth, R.M., 1999. The effects of financial incentives in
experiments. Journal of Risk and Uncertainty, 19, 7-42.
Champ, P. A., Flores, N. E., Brown, T. C. and Chivers, J., 2002. Contingent valuation
and incentives. Land Economics, 78(4), 591-604.
Chong, D. and Druckman, J. N., 2007. Framing public opinion in competitive
democracies. American Political Science Review, 101, 637-655.
Clemen, R.T. and Reilly, T. 2002. Making hard decisions. Boston: Duxbury..
Conlisk, J., 1996. Why bounded rationality. Journal of Economic Literature, 34, 669-700.
Coupey, E., Irwin, J. R. and Payne, J. W., 1998. Product category familiarity and
preference construction. Journal of Consumer Research, 24(4), 459-468.
Dickie, M., Fisher, A. and Gerking, S., 1987. Market transactions and hypothetical
demand data - a comparative-study. Journal of the American Statistical
Association, 82(397), 69-75.
Dominitz, J. and Manski, C., 1997. Using expectations data to study subjective income
expectations. Journal of the American Statistical Association, 92, 866-867.
Druckman, J. N., 2001. Using credible advice to overcome framing effects. Journal of
Law, Economics & Organization, 17, 62-82.
Druckman, J. N., 2004. Political preference formation: competition, deliberation and the
(ir)relevance of framing effects. American Political Science Review, 98(4), 671686.
Druckman, J. N. and Nelson, K. R., 2003. Framing and deliberation: How citizens'
conversations limit elite influence. American Journal of Political Science, 47(4),
729-745.
Edwards, W., 1954. The theory of decision making. Psychological Bulletin, 54, 380-417.
Fischhoff, B., 1991. Value elicitation - Is there anything in there? American Psychologist,
46(8), 835-847.
14
Fischhoff, B. and Bar-Hillel, M., 1984. Focusing techniques: A shortcut to improving
probability judgments? Organizational Behavior and Human Performance, 34,
175-191.
Fischhoff, B. and L., Furby, 1988. Measuring values: a conceptual framework for
interpreting transations with special reference to contingent valuation of visibility.
Journal of Risk and Uncertainty, 1, 147-184.
Fischhoff, B., Quadrel, M. J., Kamlet, M., Loewenstein, G., Dawes, R., Fischbeck, P.,
Klepper, S., Leland, J., Stroh, P., 1993. Embedding effects - stimulus
representation and response-mode. Journal of Risk and Uncertainty, 6(3), 211234.
Fischhoff, B. and Manski, C. (eds.), 1999. The elicitation of preferences. Journal of Risk
and Uncertainty, 19(1-3).
Fischhoff, B., Slovic, P. and Lichtenstein, S., 1980. Knowing what you want: Measuring
labile values. In T. Wallsten (Ed.), Cognitive processes in choice and decision
behavior (pp. 117-141). Erlbaum, Hillsdale, NJ.
Fischhoff, B., Welch, N. and Frederick, S., 1999. Construal processes in preference
assessment. Journal of Risk and Uncertainty, 19(1-3), 139-164.
Fishkin, J., 1991. Democracy and Deliberation. New Haven: Yale University Press.
Flores, N. E. and Strong, A., 2007. Cost credibility and the stated preference analysis of
public goods. Resource and Energy Economics, 29, 195-205.
Frederick, S. and Fischhoff, B., 1998. Scope insensitivity in elicited values. Risk
Decision and Policy, 3, 109-124.
Friedman, M., 1953, The Methodology of Positive Economics. Essays in Positive
Economics. University of Chicago Press, Chicago.
Gibbard, A., 1973. Manipulation of Voting Schemes - General Result. Econometrica,
41(4), 587-601.
Green, D., Jacowitz, K. E., Kahneman, D. and McFadden, D., 1998. Referendum
contingent valuation, anchoring, and willingness to pay for public goods.
Resource and Energy Economics, 20(2), 85-116.
Hanemann, W. M., 1994. Valuing the environment through contingent valuation. Journal
of Economic Perspectives, 8(4), 19-43.
Haward, M.F., Murphy, R.O. and Lorenz, J.M., 2008. Message framing and perinatal
decisions. Pediatrics, 122, 109-118.
Huffman, W. E., Rousu, M., Shogren, J. F. and Tegene, A., 2007. The effects of prior
beliefs and learning on consumers' acceptance of genetically modified foods.
Journal of Economic Behavior & Organization, 63, 93-2006.
Johnson, E.J. and Goldstein, D.G., 2003. Do defaults save lives? Science, 302, 13381339.
Kahneman, D., 2003. Maps of bounded rationality: psychology for behavioral economics.
American Economic Review, 93(5), 1449-1475.
Kahneman, D., Knesch, J.L. and Thaler, R.H. 1990. Experimental tests of the endowment
effect and the Coase theorem. Journal of Political Economy 98(6), 1325-1348.
Kahneman, D., Slovic, P., and Tversky, A. (eds.), 1982. Judgment under uncertainty:
Heuristics and biases. New York: Cambridge University Press.
Kahneman, D. and Tversky, A. 1979. Prospect theory: an analysis of decision under risk.
Econometrica, 47(2), 263–292.
15
Keeney, R. L., and Raiffa, H., 1993. Decisions with multiple objectives: Preferences and
tradeoffs. New York: Cambridge University Press.
Leduc, L., 2002. Opinion change and voting behaviour in referendums. European Journal
of Political Research, 41(6), 711-732.
Lerner, J. S. and Tetlock, P., 1999. Accounting for the effects of accountability.
Psychological Bulletin, 125, 255-275.
Levin, I. P. and Gaeth, G. J., 1988. How consumers are affected by the framing of
attribute information before and after consuming the product. Journal of
Consumer Research, 15(3), 374-378.
Lichtenstein, S. and Slovic, P. (eds), 2005. The construction of preference. Cambridge
University Press, New York.
List, J. A., 2004. Neoclassical theory versus prospect theory: Evidence from the
marketplace. Econometrica, 72(2), 615-625.
Luskin, R. C., Fishkin, J. S. and Jowell, R., 2002. Considered opinions: deliberative
polling in Britain. British Journal of Political Science 32, 455-487.
Loomis, J., T. Brown, B. Lucero and G. Peterson, 1997. Evaluating the validity of the
dichotomous choice question format in contingent valuation. Environmental and
Resource Economics, 10, 109-123.
Lupia, A., 1994. Shortcuts versus encyclopedias - information and voting behavior in
California insurance reform elections. American Political Science Review, 88(1),
63-76.
Lupia, A. and Matsusaka, J. G., 2004. Direct democracy: new approaches to old
questions. Annual Review of Political Science, 7, 463-482.
Manski, C. F., 2000. Economic analysis of social interactions. Journal of Economic
Perspectives, 14(3), 115-136.
McFadden, D., 1994. Contingent valuation and social choice. American Journal of
Agricultural Economics, 76(4), 689-708.
McFadden, D., 2001. Economic choices. American Economic Review, 91(3), 351-378.
Milkman, K. L., Chugh, D. and Bazerman, M. H., 2009. How can decision making be
improved? Perspectives on Psychological Science, 4, 379-383.
Mitchell, R. C. and Carson, R. T., 1989, Using surveys to value public goods: The
contingent valuation method. Resources for the Future, Washington, DC.
Morgan, M.G., Fischhoff, B., Bostrom, A., and Atman, C., 2001. Risk communication: A
mental models approach. Cambridge University Press, New York.
Murphy, J. J., Allen, P. G., Stevens, T. H. and Weatherhead, D., 2005. A meta-analysis of
hypothetical bias in stated preference valuation. Environmental and Resource
Economics, 30, 313-325.
Nickerson D. W. and Rogers, T., 2010. Do you have a voting plan?: implementation
intentions, voter turnout, and organic plan making. Psychological Science, 21 (2),
194-9
Payne, J. W., Bettman, J. R. and Schkade, D. A., 1999. Measuring constructed
preferences: towards a building code. Journal of Risk and Uncertainty, 19(1-3),
243-270.
Plous, S., 1991. Judgment and decision making. McGraw-Hill, New York.
Poulton, E. C., 1968. The new psychophysics: Six models for magnitude estimation.
Psychological Bulletin,; 69: 1-19.
16
Poulton, EC., 1994. Behavioral decision making. Lawrence Erlbaum, Hillsdale, NJ.
Satterthwaite, M. A., 1975. Strategy-proofness and Arrows conditions - existence and
correspondence theorems for voting procedures and social welfare functions.
Journal of Economic Theory, 10(2), 187-217.
Schläpfer, F. and Bräuer, I, 2007. Theoretical incentive properties of contingent valuation
questions: do they matter in the field. Ecological Economics, 62(3-4), 451-460.
Schläpfer, F. and Schmitt, M., 2007. Anchors, endorsements, and preferences: a field
experiment. Resource and Energy Economics, 29(3), 229-243.
Schläpfer, F., Schmitt, M. and Roschewitz, A., 2008. Competitive politics, simplified
heuristics, and preferences for public goods. Ecological Economics(65), 574-589.
Shogren, J. F., 2006. Valuation in the lab. Environmental & Resource Economics, 34(1),
163-172.
Simon, H. A., 1955. A behavioral model of rational choice. Quarterly Journal of
Economics, 69(1), 99-118.
Slembeck, T. and Tyran, J. R., 2004. Do institutions promote rationality? An
experimental study of the three-door anomaly. Journal of Economic Behavior &
Organization, 54(3), 337-350.
Small, D.A., Loewenstein, G. and Slovic, P., 2007. Sympathy and callousness: The
impact of deliberative thought on donations to identifiable and statistical victims.
Organizational Behavior and Human Decision Processes, 102, 142-153.
Richard H. Thaler, R. H. and Sunstein, C. R., 2008. Nudge: Improving decisions about
health, wealth, and happiness. Yale University Press, New Haven, CT.
Tomz, M. and Sniderman, P.M., 2005. Brand names and the organization of mass belief
systems. Paper presented at the Annual Meeting of the Midwest Political Science
Association, Chicago.
Tversky, A. and Kahneman, D., 1981. The framing of decisions and the psychology of
choice. Science, 211(4481), 453-458.
vonWinterfeldt, D. and Edwards, W., 1986. Decision analysis and behavioural research.
Cambridge University Press, New York.
Vossler, C. A. and Kerkvliet, J., 2003. A criterion validity test of the contingent valuation
method: comparing hypothetical and actual voting behavior for a public
referendum. Journal of Environmental Economics and Management, 45(3), 631649.
17
Table 1. Classification of choice domains.
Private vs. public good
Private
Familiarity (own
experience)
familiar
unfamiliar
Public
familiar
unfamiliar
18
Informative contextual
cues
present
absent
present
absent
present
absent
present
absent
Table 2. Ordinary least squares estimates of models explaining the ratio of hypothetical to actual willingness to pay.
Trimmed dataseta
Model 2a
Model 2b
Coeff.
t-ratio
Coeff.
t-ratio
5.628*** 3.617
4.727***
2.841
2.221
1.030
-1.167
-0.494
1.551
0.812
4.801*
1.897
-2.935*
-1.935
2.184
1.042
-3.134** -2.066
-1.499
-1.002
Full dataset
Model 1a
Coeff.
t-ratio
3.628***
3.978
0.907
0.724
0.955
0.771
-0.065
-0.082
-1.167
-1.390
-3.241***
-3.237
-1.838*
-1.846
Model 1b
Variable
Coeff.
t-ratio
Constant
2.733***
2.700
Student
-2.434
-1.635
Group
4.906***
2.834
Private
3.390***
3.193
Choice
0.165
0.192
Within
-3.162*** -3.480
Calibrate
-2.466*** -2.696
Cues (C)
-1.991*
-1.704
Familiar (F)
-8.030*** -4.341
CxF
5.028**
2.402
n
83
83
39
Adj. R2
0.08
0.26
0.13
Notes: Significance levels: *: p<0.1; **: p<0.05; ***: p<0.01.
a
Within-subject comparisons and observations with “calibrated” values excluded.
19
-3.665*
-1.827
-9.062***
-3.123
6.723**
2.082
39
0.30
Appendix Table A. Description of goods and choice contexts used in the study sample of the meta-analysis
Coding
Study
Good and choice context
Familiar
Cues
Blumenschein,et al. (1997)
lab sunglasses to economics students
0
0
Blumenschein et al. (2001)
asthma management program to patients in “scientific
0
0
study”, general population
Bohm (1972)
preview of TV program at broadcast station, general
1
0
population
Botelho and Costa Pinto (2002) contribution to NGO information campaign on river
0
0
otters, at university, students
Boyce et al. (1989)
common house plant in experiment, university staff at
1
0
university
Brown et al. (1996)
contribution to remove roads along Grand Canyon rim,
0
0
mail survey/collection by university
Brown and Taylor (2000)
contribution to NGO environmental program, at
0
0
university
Cameron, et al. (2002)
green electricity contribution, in established renewable
0
1
energy and tree planting program (fixed 6$ surcharge)
Carlsson and Martinsson (2001) contribution to 3 environmental projects “rain forest”,
0
0
“Mediterranean”, “Baltic sea”, students prior to lecture
Champ et al. (1997)
contribution to remove roads along Grand Canyon rim,
0
0
mail survey/collection by university
Champ and Bishop (2001)
green electricity donation, actual wind power program
0
1
of electric company, fixed surcharge
Duffield and Patterson (1992)
contribution to Montana river mgt., contingent valuation
0
0
survey context
Frykblom (1997)
Swedish national atlas, volume “the environment”, prior
1
0
to lecture
Frykblom (2000)
Swedish national atlas, volume “the environment”, prior
1
0
to lecture
Heberlein and Bishop (1986)
hunting permits to hunters, issued by Department of
1
1
Natural Resources
Johannesson et al. (1998)
box of chocolates, displayed in front of students
1
1
Kealy et al. (1988)
familiar candy bars (Cadbury)
1
1
List (2001)
sports cards in existing market (list price of $200-250)
0
1
List (2003)
sports cards in existing market (book value of $12)
0
1
List and Shogren (1998)
sports cards in existing market ($350)
0
1
Loomis et al. (1997)
wildlife art print displayed to subjects, clerical and
1
0
administrative staff on campus, auction context
MacMillan, et al. (1999)
contribution to nature reserve, mail survey experiment,
0
0
although mimicking regular collection
Murphy et al. (2002)
contribution to environmental NGO, students at
0
0
university
Neill et al. (1994)
painting (list price $75), historical map replica (list price
0
0
$20) shown to subjects, students at university
Sinden (1988)
contribution to soil conservation program, students in
0
0
class
Spencer et al. (1998)
contribution to pond water monitoring program, student
0
0
respondents
Vossler et al. (2003)
pre-election poll on collectively provided public good,
0
1
issue subject to public debate, general population
Vossler and Kerkvliet (2003)
pre-election poll on collectively provided public good,
0
1
issue subject to public debate, general population
20
Working Papers of the Socioeconomic Institute at the University of Zurich
The Working Papers of the Socioeconomic Institute can be downloaded from http://www.soi.uzh.ch/research/wp_en.html
1007
1006
1005
1004
1003
1002
1001
0920
0919
0918
0917
0916
0915
0914
0913
0912
0911
0910
0909
0908
0907
0906
When Are Preferences Consistent? The Effects of Task Familiarity and Contextual
Cues on Revealed and Stated Preferences, Felix Schläpfer, Baruch Fischhoff, August
2010, 20 p.
Golden Balls: A Prisoner’s Dilemma Experiment, Donja Darai, Silvia Grätz, July 2010,
47 p.
Probability Weighting as Evolutionary Second-best, Florian Herold, Nick Netzer, July
2010, 32 p.
Trade Openness, Gains from Variety and Government Spending, Sandra Hanslin,
April 2010, 46 p.
Is the Welfare State Sustainable? Experimental Evidence on Citizens’ Preferences for
Income Redistribution, Ilja Neustadt, Peter Zweifel, March 2010, 32 p.
Preferences for Health Insurance in Germany and the Netherlands – A Tale of Two
Countries, Peter Zweifel, Karolin Leukert, Stephanie Berner, March 2010, 22 p.
Convex Treatment Response and Treatment Selection, Stefan Boes, January 2010,
28 p.
Bounds on Counterfactual Distributions Under Semi-Monotonicity Constraints,
Stefan Boes, December 2009, 38 p.
Rotten Kids with Bad Intentions, Nick Netzer, Armin Schmutzler, December 2009,
38 p.
Partial Identification of Discrete Counterfactual Distributions with Sequential Update
of Information, Stefan Boes, December 2009, 37 p.
How much do journal titles tell us about the academic interest and relevance of
economic reserach? An empirical analysis. Felix Schläpfer, December 2009, 14 p.
Fine Tuning of Health Insurance Regulation: Unhealthy Consequences for an
Individual Insurer, Johannes Schoder, Michèle Sennhauser, Peter Zweifel, August
2009, 18 p.
Capping Risk Adjustment?, Patrick Eugster, Michèle Sennhauser, Peter Zweifel,
September 2009, 27 p.
A Pharmaceutical Innovation: Is it Worth the Money? Whose Money?, Michèle
Sennhauser, Peter Zweifel, September 2009, 22 p.
Copula-based bivariate binary response models, Rainer Winkelmann, August 2009,
26 p.
Simulating WTP Values from Random-Coefficient Models, Maurus Rischatsch, July
2009, 6 p.
Physician dispensing and the choice between generic and brand-name drugs – Do
margins affect choice?, Maurus Rischatsch, Maria Trottmann, July 2009, 15 p.
GPs' preferences: What price fee-for-service?, Peter Zweifel, Maurus Rischatsch,
Angelika Brändle, July 2009, 21 p.
Social Mobility and Preferences for Income Redistribution: Evidence from a Discrete
Choice Experiment, Ilja Neustadt, Peter Zweifel, July 2009, 31 p.
Robust estimation of zero-inflated count models, Kevin E. Staub, Rainer
Winkelmann June 2009, 22 p.
Competitive Screening in Insurance Markets with Endogenous Wealth
Heterogeneity, Nick Netzer, Florian Scheuer, April 2009, 28 p.
New Flight Regimes and Exposure to Aircraft Noise: Identifying Housing Price Effects
Using a Ratio-of-Ratios Approach, Stefan Boes, Stephan Nüesch, April 2009, 40 p.
0905
0904
0903
0902
0901
0816
0815
0814
0813
0812
0811
0810
0809
0808
0807
0806
0805
0804
0803
0802
0801
0719
0718
Patents versus Subsidies – A Laboratory Experiment, Donja Darai, Jens Großer,
Nadja Trhal, March 2009, 59 p.
Simple tests for exogeneity of a binary explanatory variable in count data regression
models, Kevin E. Staub, February 2009, 30 p.
Spurious correlation in estimation of the health production function: A note, Sule
Akkoyunlu, Frank R. Lichtenberg, Boriss Siliverstovs, Peter Zweifel, February 2009,
13 p.
Making Sense of Non-Binding Retail-Price Recommendations, Stefan Bühler, Dennis
L. Gärtner, February 2009, 30 p.
Flat-of-the-Curve Medicine – A New Perspective on the Production of Health,
Johnnes Schoder, Peter Zweifel, January 2009, 35 p.
Relative status and satisfaction, Stefan Boes, Kevin E. Staub, Rainer Winkelmann,
December 2008, 11 p.
Delay and Deservingness after Winning the Lottery, Andrew J. Oswald, Rainer
Winkelmann, December 2008, 29 p.
Competitive Markets without Commitment, Nick Netzer, Florian Scheuer, November
2008, 65 p.
Scope of Electricity Efficiency Improvement in Switzerland until 2035, Boris Krey,
October 2008, 25 p.
Efficient Electricity Portfolios for the United States and Switzerland: An Investor
View, Boris Krey, Peter Zweifel, October 2008, 26 p.
A welfare analysis of “junk” information and spam filters; Josef Falkinger, October
2008, 33 p.
Why does the amount of income redistribution differ between United States and
Europe? The Janus face of Switzerland; Sule Akkoyunlu, Ilja Neustadt, Peter Zweifel,
September 2008, 32 p.
Promoting Renewable Electricity Generation in Imperfect Markets: Price vs. Quantity
Policies; Reinhard Madlener, Weiyu Gao, Ilja Neustadt, Peter Zweifel, July 2008,
34p.
Is there a U-shaped Relation between Competition and Investment? Dario Sacco,
July 2008, 26p.
Competition and Innovation: An Experimental Investigation, May 2008, 20 p.
All-Pay Auctions with Negative Prize Externalities: Theory and Experimental
Evidence, May 2008, 31 p.
Between Agora and Shopping Mall, Josef Falkinger, May 2008, 31 p.
Provision of Public Goods in a Federalist Country: Tiebout Competition, Fiscal
Equalization, and Incentives for Efficiency in Switzerland, Philippe Widmer, Peter
Zweifel, April 2008, 22 p.
Stochastic Expected Utility and Prospect Theory in a Horse Race: A Finite Mixture
Approach, Adrian Bruhin, March 2008, 25 p.
The effect of trade openness on optimal government size under endogenous firm
entry, Sandra Hanslin, March 2008, 31 p.
Managed Care Konzepte und Lösungsansätze – Ein internationaler Vergleich aus
schweizerischer Sicht, Johannes Schoder, Peter Zweifel, February 2008, 23 p.
Why Bayes Rules: A Note on Bayesian vs. Classical Inference in Regime Switching
Models, Dennis Gärtner, December 2007, 8 p.
Monoplistic Screening under Learning by Doing, Dennis Gärtner, December 2007,
29 p.