Reconciling sustainability preferences and behavior — The case of mutual fund investments

This study analyzes the interaction between sustainability preferences and investment behavior, particularly in the context of mutual fund investments. Based on survey data from a representative sample of Swedish mutual fund investors, we observe that while a majority of respondents express a will-ingness to sacrifice returns for more sustainable investments, only a minority claim to have actively invested in sustainable funds. This highlights a discrepancy between preferences and behavior, which we show can be understood by (in)attentiveness in the financial decision-making process. We reveal that sustainability-motivated investors are less attentive than those motivated by returns, leading to potential misalignment with their preferences. This finding emphasizes the significance of banks taking (in)attentiveness into account when communicating with customers. Information is effective for return-focused investors, while nudges and boosts may better facilitate decisions for sustainability-focused investors.


Introduction
The financial sector has been identified as a key sector to leverage capital necessary to reach the temperature goals of the Paris Agreement.Still, the financial investment gap to decarbonize the global economy is significant.For instance, the Commission of the European Union has estimated that for the EU alone, the annual investment gap corresponds to approximately 180 billion EUR to reach 40 percent reductions in greenhouse gas emissions until 2030.1 A large part of the investment gap is expected to be filled by the private sector by redirecting private capital and financial flows towards green and sustainable investments.2 Retail investors, often active in mutual fund investments, play a critical role.Yet currently, only 3 percent of the global fund market comprises sustainable funds.(UN, 2021).Significant measures are therefore necessary to direct capital towards sustainable investments.
While disclosure and increased awareness of sustainability issues are clearly prerequisites for efficient sustainable capital allocation, regulators seem to make the assumption that if financial products correctly disclose their environmental and social impacts, then investors will make informed decisions. 3In this paper, we show that the relationship between preferences and investment behaviour is much more complex, and that disclosure of sustainability impacts and awareness of such issues may not be enough to steer capital towards green and sustainable investments.Specifically, zooming in on one important product group, mutual funds, we find that retail investors do not invest in sustainable funds to the same degree as they hold sustainability preferences.Based on a survey among a representative sample of Swedish mutual fund investors, we find that, while 71 percent state that they would be willing to give up some returns for making sustainable investments, only 43 percent have actively invested in sustainable funds.One explanation suggested by our findings is that those who are sustainability motivated are more likely to make their fund choice inattentively than those who are primarily motivated by profitability and high returns.As a consequence, sustainability motivated investors face a larger risk that their investments are not aligned with their preferences.This in turn implies inefficiencies and welfare losses-for individuals who fail to make preference-aligned investments, as well as for the global community when the transition towards sustainable and green investments becomes slower than if all investors acted rationally.
The overall objective of this study is to further our understanding of the interplay between sustainability preferences and investment behavior.More specifically, we study whether individuals make preference-aligned mutual fund investment decisions and what role the decision-making process as such has.While we are neither the first to study preferences for sustainable funds, nor to identify a gap between preferences and investment behavior, we are able to show that this seemingly paradoxically behaviour is consistent with a theory of decision making that accounts for contextual factors of the choice situation.Hence, we can explain what mechanisms are likely to imply the discrepancy between preferences and behavior.Our results also provide practical insights for banks and financial advisors.To facilitate preference-aligned investments, we highlight the importance of considering how different types of communication can influence investment decisions based on individual motivation and attention levels.
There is a growing literature on investor preferences and behavior concerning sustainable investments (e.g., Bollen, 2007; Nilsson, 2008; Krüger, 2015; Riedl and  Smeets, 2017; Bassen et al., 2019; Hartzmark and Sussman, 2019; Gutsche et al.,  2019, 2023, 2021; Barber et al., 2021; Bauer et al., 2021; Anderson and Robinson,  2022).Regarding preferences, we find, in accordance with many of the abovementioned studies, that women, younger individuals, and those who are more concerned with environmental issues tend to have stronger preferences for sustainable investments.Regarding behavior, previous literature is more inconclusive, to some extent depending on the setting.Experimental studies often find that sustainability preferences carry over to investment behavior (e.g., Riedl and Smeets, 2017; Gutsche  et al., 2023; Lagerkvist et al., 2020).However, behavior revealed in experimental settings is not necessarily the same as holding sustainable mutual funds in real life and studies based on data on actual investments are less conclusive.Anderson and  Robinson (2022) report that females hold stronger environmental preferences but are not more likely than men to actively invest in environmental, social, and governance (ESG) pension funds.Neither do Riedl and Smeets (2017) find any gender J o u r n a l P r e -p r o o f Journal Pre-proof ested in financial matters (referred to as financial disengagement) and subsequently hold financial assets to a lesser extent (Anderson and Robinson, 2022).Likewise, Bassen et al. (2019) find that those who place more weight on climate than on returns tend to make their financial decisions in a less reflective and attentive way.
The results from these studies highlight the role of (in)attentiveness in the financial-decision making process and the interdependence between preferences and attentiveness.Analysing our survey data in the light of the theory outlined in Löfgren and Nordblom (2020) offers further insights into the mechanisms that can explain the discrepancy between sustainable preferences and investments in sustainable funds.We indeed find that sustainability motivated investors are more likely than return motivated to make their investment decisions inattentively.This implies that banks should consider this disparity while communicating with their customers, facilitating appropriate mutual fund investment decisions.Not only would more targeted communication towards sustainability-motivated investors enable them to make preference-aligned investments; it would also increase overall investment in sustainable assets, which is necessary for the sustainable transition.
The structure of the paper is as follows.In Section 2, the focus is on providing insights into sustainability preferences and investment behavior based on survey data, in particular, whether sustainability is an important consideration for mutual fund investors.In addition, we investigate how trade-offs between return and sustainability are perceived and if there are specific sustainability aspects that seem to be more or less important when investing in mutual funds.In Section 3, we employ the theoretical model presented in Löfgren and Nordblom (2020) for an indepth analysis of decision-making processes and fund investment strategies.This aids in understanding the interdependence between sustainability preferences and investment behavior, and whether certain groups are less likely to make preferencealigned investment decisions.Drawing on analyses and findings from Sections 2 and 3, Section 4 provides guidelines and implications for banks and financial advisors concerning the communication of sustainability facets of mutual funds.Section 5 concludes the paper.To get a better understanding of mutual fund investors' preferences and behavior, we conducted an online survey in Sweden between August 28'th and September 13'th 2021.Data were collected from a telephone-recruited web panel of over 67,000 individuals, provided by Norstat, one of the largest data collectors for market research in Europe, which is frequently engaged by researchers for conducting high-quality surveys.Recruitment to Norstat's Swedish web panel is made randomly geographically across Sweden as well as over gender and age groups (individuals recruited are 15 years and older). 4ur initial respondent sample is representative of the Swedish population, but the final data only includes those affirmatively stating they owned mutual funds (68% of the original sample).This inherently limits our data to individuals aged 18 or older, the minimum age to invest in mutual funds independently in Sweden.A total of 4,011 individuals completed the full survey.
When comparing our final sample of mutual fund investors to the Swedish population in1 general, we see that Swedish mutual fund investors are on average younger and more educated than the population as a whole. 5Also, there is a small overrepresentation of men (52 percent compared with 50 percent within the population as a whole).To invest in mutual fund is common in Sweden.According to the Swedish Investment Fund Association 6 , 73 percent of the adult population had private investment in mutual funds in 2022, so our sample represents a vast majority of the general Swedish population.

The online survey
The main part of the survey (after the screening question) included questions eliciting respondents' financial decision-making process including questions on invest-4 Norstat filters out inactive panellists (where the definition of an of an active panellist is that the panellist should have opened at least one survey during the last 12 months).Panellists in the web panel are compensated in points that are paid out either as money or donated to charity.Hence, panellists are used to answering survey questions and generally do it thoroughly, which makes the data collected of high quality.
5 Notably, our sample closely aligns with the Swedish Investment Fund Association's (SIFA, 2022) survey in terms of mutual fund ownership.
6 https://www.fondbolagen.se/enJ o u r n a l P r e -p r o o f Journal Pre-proof ment strategies and preferences for various sustainability aspects of mutual fund choices.The last part of the survey contained socioeconomic questions.Based on earlier literature we are specifically interested in controlling for gender, educational level, and age group.After communication with financial advisors we also asked questions regarding investment amount, whether there are children in the household, and whether one lives in a metropolitan area.Due to potential sensitivity, we chose not to make the question regarding investment amount mandatory.Since not everyone chose to answer this question, our analysis contains 3,529 observations rather than 4,011.Summary statistics of the explanatory variables used in the analysis can be found in Table A1 in Appendix A.7 2.2 Survey results

Sustainability and importance of mutual fund characteristics
To get an understanding of the relative importance of sustainability for the fund choice, we asked how important various aspects are when investing in a new fund.Figure 1 shows the mean responses to the aspects "sustainability"and "historical returns."8Respondents answered on a 5-degree Likert scale, where 1 was defined as totally unimportant and 5 as very important.The mean value for sustainability is 3.1; hence, on average, it is neither important nor unimportant.We notice, though, that in line with earlier studies, women state that they consider sustainability to be significantly more important than men do (3.3vs 2.9, p<0.000).In comparison, the mean value for the importance of historical returns when making a fund choice is 3.5, and the discrepancy between genders is smaller than for sustainability (3.5 for women and 3.6 for men, p = 0.003).
Overall, 36.5 percent of the respondents stated that sustainability is important or very important (corresponding to scores of 4 or 5 on the survey's given scale) when choosing a fund.We divide our sample into groups depending on their sustainability preferences and label the respondents who state that sustainability is important or very important as "sustainability motivated."Regarding return, 56 percent of the respondents stated that historical returns are important or very important, and we label those "return motivated."As shown in Table 1, these groups are not mutually exclusive.Of our sample, 22 percent perceive both aspects to be important.

J o u r n a l P r e -p r o o f
Journal Pre-proof  Another indicator of sustainability motivation could be whether one is willing to give up returns in order to make one's portfolio more sustainable.Although this does not directly capture sustainability motivation, it hints at the comparative importance of sustainability versus expected returns.Interestingly, almost 72 percent of the full sample would be willing to give up some return to invest in a sustainable fund 9 , and 30 percent state that they would be willing to invest in a sustainable fund even if the expected return is more than 5 percent lower than for another fund.8.0 10.9 Gender difference significant at p < 0.001. 9Divided by gender, 78 percent of women and 65 percent of men. 10 35 percent of the women and 26 of the men.

Journal Pre-proof
Many are thus willing to give up returns to invest sustainably, but what are their expectations concerning the trade-off between sustainability and returns?Interestingly, most investors do not think there is a negative relationship between sustainability and return of funds.Of our sample, 30 percent think that sustainable funds are less profitable, 28 percent that they are more profitable, and 42 percent that there is no relationship between returns and sustainability.11This positions our Swedish sample as more optimistic about the profitability of sustainable investments relative to existing literature.Riedl and Smeets (2017) found that 14-17 percent and Gutsche et al. (2023) that 20 percent perceived sustainable funds to yield higher returns than conventional ones.12As expected, having a positive expectation regarding the returns of sustainable funds is more common among sustainability motivated investors.Of the sustainability motivated investors, 39 percent think that sustainable funds yield higher returns than other funds, compared with 22 percent of those who are not sustainability motivated.
Finally, among the 30 percent who think that sustainable funds in general yield less returns, as many as 70 percent are still willing to give up some return to invest in a sustainable fund.This observation is in line with those of, e.g., Riedl and Smeets  (2017), Bauer and Smeets (2015), Bauer et al. (2021), and Apostolakis et al. (2016,  2018) who all find that socially responsible investors expect to get lower returns but are still willing to invest in sustainable assets, sacrificing profitability.

Factors explaining attitudes towards sustainability and returns
Next, we run a bivariate ordered probit to find out what factors explain an individual's stated importance of sustainability and historical return for their fund choice. 13We control for gender, educational level, age group, investment amount, having children living in the household, and whether one lives in a metropolitan area.(see Table A1 in Appendix A for the definitions and descriptive statistics of the explanatory variables).Since income is highly correlated with investment, we do not include income in our econometric specifications.The regression results are presented in the two first columns in Table 3. 14 Our results show that men are somewhat more likely to care about returns than women, while women are much more concerned with sustainability than men, something already indicated in Figure 1. 15 Tables A3 and A4 show the marginal effects and reveal that men are about 2.9 ppt more likely than women to state that historical returns are very important, while women are 7.5 ppt more likely than men to mention sustainability as very important.The higher the level of education, the more important are returns for the fund choice. 16Interestingly, sustainability motivation is uncorrelated with education.Note though, that our sample of investors has a higher than average education level to start with.Young and Utz (2014), and Gutsche et al. (2023) who find women to be significantly more concerned than men with sustainability. 16Education is a categorical variable from 1 indicating less than high school to 4 indicating university.

J o u r n a l P r e -p r o o f
Journal Pre-proof investors are more likely than others to find both aspects important (marginal effects for sustainability and returns are of the same magnitude), while the oldest age group is relatively less concerned with returns.17Interestingly, sustainability motivation is unrelated to investment amounts, while those with the smallest investment amounts care less about returns than those who save larger amounts. 18ne may jump to the conclusion that young people are much more sustainability motivated than older age groups.However, in Column 3 in Table 3 we run an ordered probit with the willingness to give up returns for a sustainable investment, i.e., the importance of sustainability relative to returns. 19.Here we actually find the strongest effect on pensioners.However, the regression corroborates the indication from Table 2 that women are more willing than men to give up expected returns for sustainable investments.
Accounting for the multi-faceted nature of sustainability in mutual fund selection Given the broad nature of sustainability, we further explore detailed, well-defined sustainability aspects pertinent to mutual funds.Figure 2 shows descriptive statistics of how important the respondents find sustainability in general as well as five specific sustainability aspects often evaluated in ESG scores and rankings.Again, the respondents answered on a 5-degree scale, where 1 indicated completely unimportant and 5 very important.We find that the two most important aspects are that perceived unethical industries are excluded from the fund and that human rights are guaranteed.Furthermore, the oldest age group is most concerned with each of the more detailed sustainability aspects (the differences to younger age groups are significant, p < 0.000.)Running ordered probits of the sustainability aspects, gender and age disparities in concerns emerge, with women exhibiting greater levels of concern compared to men, and pensioners demonstrating higher levels of concern relative to younger age groups across all specific aspects.20

Behavior
As outlined above, preferences for sustainable funds seem rather prevalent among mutual fund investors, more so among certain groups.However, these preferences do not automatically translate into investment behavior.Although 71 percent of the Figure 2: Mean values of the importance of sustainability aspects, ranging from 1 to 5 where 1=completely unimportant and 5=very important respondents claim to be willing to give up profitability in return for sustainability, only 43 percent say that they have actively invested in a sustainable fund at some point.Still, as anticipated, a much larger share of the sustainability motivated than the non-sustainability motivated investors have done so (62 as compared to 33 %).But even among the sustainability motivated investors, a surprisingly large share state that they never made an active choice to invest in a sustainable fund.Among the sustainability motivated women, as many as 41 percent state that they have never actively invested in a sustainable fund.Overall, 23 percent answer that they hold no sustainable funds at the moment, while 21 percent think that more than half of their portfolio consists of sustainable funds.
Based on our survey, we thus have two potential measures of sustainable investment behavior: the first is the answer to the question whether one has made any active investment in sustainable funds and the second is the stated share of current holdings allocated to sustainable funds.While current holdings may provide the

J o u r n a l P r e -p r o o f
Journal Pre-proof most accurate reflection of preferences, survey-based estimations are likely subject to recall bias and ignorance regarding sustainability characteristics.21Therefore, active investment henceforth serves as our primary behavioral measure, while the estimated sustainable share is utilized in robustness checks with the acknowledgement of potential recall bias.
To examine systematic variations in investments in sustainable mutual funds, we employ probit regressions.The dependent variable takes a value of 1 if the individual claims to have actively invested in a sustainable mutual fund and 0 otherwise.
In Table 4, we present the results in terms of marginal effects.Column 1 utilizes the full sample and the same explanatory variables as the previous preference regressions.In column 2, we also control for whether one perceives sustainable funds as more profitable than other funds, while column 3 instead controls for sustainability motivation.Column 4 incorporates both variables.The last two columns restrict the sample exclusively to sustainability-motivated fund investors. 22hile women are more likely to be sustainability motivated and more willing to give up expected returns to invest in a sustainable fund than are men, they are not more likely to have invested in sustainable funds (controlling for being sustainability motivated, women are even less likely than men to invest).Although sustainability motivation is uncorrelated with education and investment amounts, stated investment in sustainable funds is more common the higher the education and the larger the investment amount; these effects remain when controlling for sustainability preferences. 23The oldest age group exhibits the lowest likelihood of stating to have made investments in sustainable funds, despite expressing the highest level of concern for each specific sustainability aspect.The age effects are substantial.In the full sample, a young person is 15 percentage points (ppts) more likely to actively have invested in sustainable funds than a middle-aged person, while a retired person is 10 ppts less likely.The subsample of sustainability motivated fund investors contains more young people, but there is still a substantial age effect in behavior where the youngest are 20 ppts more likely than the oldest to have actively invested in sustainable funds.Marginal effects; Standard errors in parentheses (d) for discrete change of dummy variable from 0 to 1 * p < 0.05, * * p < 0.01, * * * p < 0.001 As expected, perceiving sustainable funds as more profitable is positively correlated with the likelihood of active investment.However, when both variables are included in the regression as explanatory factors, the effect of being sustainability motivated is nearly five times larger than that of perceiving sustainable funds as more profitable (30 compared to 6 ppts, as observed in column 4).Hence, the most important indicator for sustainable investment is being sustainability motivated.Moreover, for those who are sustainability motivated, perceived profitability is of no significant importance for the investment choice (see column 6).This is consistent with the claims by, e.g., Hartzmark and Sussman (2019), Gutsche et al. (2023)  and Bauer et al. (2021), that those who invest in sustainable funds tend to value sustainability significantly more than returns.

Preference and behavior discrepancy
While sustainability-motivated investors, in general, are more inclined to have invested in sustainable funds, a notable proportion of respondents still indicate that they have not done so.For instance, among the sustainability motivated investors,

J o u r n a l P r e -p r o o f
Journal Pre-proof only 30 percent of retired women with low education state that they have actively invested in sustainable funds, while the corresponding share among young men with high education is 73 percent.The likelihood that one is sustainability motivated, but not have invested in sustainable funds is significantly higher for women, for those with low education and smaller mutual fund investments, and for those older than 64.A similar result is obtained when instead analyzing the sustainability motivated investors who state that they currently have no sustainable holdings. 25ven if some earlier studies also find that investors do not invest according to their preferences (e.g., Anderson and Robinson, 2022), the observation may seem surprising.In the next section, we employ the decision-making model proposed by Löfgren  and Nordblom (2020) to gain insights into what mechanisms hinder investors from making financial decisions that align with their preferences.
3 Financial decision-making and attentiveness

Attentive versus inattentive fund investment decisions
Löfgren and Nordblom (2020) develop a decision-making model.In its basic form, the model posits that individuals make decisions either attentively, i.e., they gather information to make an informed and optimal choice, or inattentively, i.e., without deliberation, risking a decision that may not align with their preferences.Three distinct properties of the decision are crucial for whether it is made attentively or inattentively: (i) how important the decision is, (ii) how difficult or demanding it is to make an attentive decision, and (iii) how confident one is that the outcome of an inattentive decision would be preference-aligned.
The basic mechanisms are firstly, that the more important a decision is, the more likely it is worth the effort of making it attentively.Secondly, the more challenging one perceives the effort required to make an attentive choice, the more inclined one is to instead opt for an inattentive choice.Lastly, if individuals have high confidence that the outcome will align with their preferences without them having to exert effort, they are more likely to make the decision inattentively, relying on gut feelings, habits, or other heuristics.While each of these components plays a role, it is the combination of the three that is decisive for whether the choice is made attentively or inattentively. 26n this section, we go through each of the three components-subjective impor-

J o u r n a l P r e -p r o o f
Journal Pre-proof tance, difficulty, and confidence-within our sample of mutual fund investors.By doing so we can identify whether certain groups can be expected to behave more or less inattentively and therefore be more or less likely to make an investment decision that is not preference-aligned.In addition, we create a composite measure called "inattentiveness," which quantifies the likelihood of making an inattentive investment decision.In subsection 3.2, we investigate the potential correlation between our inattentiveness measure and sustainability preferences, as well as its explanatory power in relation to investment behavior.As a reference for the discussions, we present coefficients from ordered probit regressions for each of the three components as well as the combined inattentiveness measure in Table 5. 27 Given the potential relationship between the three components and knowledge about the financial market, we introduce the dummy variable Knowledge to capture such knowledge.This variable takes a value of 1 if the respondent is at least somewhat familiar with the Morningstar Sustainability Rating.As described above, the more important the choice, the more likely that it will be made attentively.Respondents were surveyed on the importance of fund choice, rating their responses on a 5-point scale: 1 indicated Completely unimportant choice and 5 Very important choice. 29The mean is 3.5, indicating that respondents on average seem to perceive mutual fund investments as a somewhat important choice.
As observed in the first column in Table 5 men state the fund choice to be more important than women do, and the degree of stated importance is decreasing in age. 30It can be noted that the marginal effects of being younger than 30 (or older than 64) are much stronger than the gender effect. 31Furthermore, it is worth noting that the stated importance increases with the size of investments.Conversely, there is no observed correlation between education and the perceived importance of fund choice.
Regarding financial knowledge, respondents who are familiar with the Morningstar Globes find the fund choice more important; they are 9 ppts more likely to state it is very important than those who do not have knowledge about the rating. 32or given levels of confidence and difficulty, we would thus expect young men with large investments to be the most likely to make an attentive fund choice.Conversely, women aged 64 or older, with small investment amounts, would be more prone to making the choice inattentively.

Difficulty
Making an attentive choice requires a certain level of effort, as it involves acquiring and evaluating information.If an individual perceives the task as highly challenging, they are more likely to make the choice inattentively.We asked respondents how difficult or demanding they would find an attentive choice on a 5-degree scale. 33hoosing 1 indicated Not at all demanding and 5 Very demanding.The overall mean is 3.1.As observed in the second column in Table 5, men are less likely than women to think that fund choice is demanding 34 Notably, the marginal effects of 29 Direct translation: "Some choices are of great importance to people, while others seem unimportant.How important would you say that choice of funds is to you?" 30 Table A8 in Appendix A shows the marginal effects.
31 Men are 2 ppt more likely than women to state very important, while the youngest age-group is 6.5 ppts more likely and the oldest 5.9 ppts less likely than the middle-aged to state that fund choice is very important.
32 Excluding Knowledge leaves the other marginal effects largely unchanged.
33 Direct translation: "To make a well-thought-out choice of funds, one could need to gather and value certain information.How demanding would it be for you to make a well-thought-out choice of funds?" 34 Table A9 in Appendix A shows marginal effects.

J o u r n a l P r e -p r o o f
Journal Pre-proof gender are much larger than concerning importance. 35The youngest age group is the one perceiving the choice most difficult.Hence, these results are consistent with those of Lusardi (2019), who finds that women and young people have the lowest financial literacy (and should therefore find the investment decision more difficult).Interestingly, the higher the level of education, the more difficult one perceives the fund choice to be.This may seem counterintuitive.However, it is important to note that the measure of "difficulty" is subjective and may not necessarily reflect the objective capability of making the best choice.Individuals with higher levels of education may be more aware of the inherent complexity involved in making the choice and, as a result, recognize that it would require significant effort to make a well-thought-out decision.A similar interpretation can be applied to the non-significant knowledge coefficient.On the one hand, those with higher financial knowledge find it less difficult to understand any given set of information.On the other hand, they may also be aware that a lot of information would be needed in order to make the optimal attentive decision.
Nevertheless, the perceived difficulty of the choice also relies on its objective complexity, which, in turn, depends on individual's preferences.As proposed by Pedersen et al. (2021), individuals who solely prioritize profitability will have a decision task that is objectively less complex compared to those who consider additional factors such as sustainability, risk, and business orientation.We therefore create an index of complexity (ranging from 0 to 10).The index represents the number of aspects (outlined in Table A2) that are important for the fund choice.These aspects encompass various factors such as historical return, fund fees, and sustainability, among others.36Although this index is not a perfect measure of choice complexity, it serves as a good proxy for the objective complexity of an optimal investment decision for an individual.Including this index as an explanatory variable for how demanding the fund choice is perceived (results are available in Table A10 in Appendix A) we observe that (as expected) a more objectively complex choice is perceived as more difficult.Also, when accounting for complexity, the knowledge measure becomes statistically significant in the expected way: individuals with higher knowledge tend to perceive fund choice as less difficult.The gender effect remains unchanged, while marginal effects and significance levels associated with education and age are reduced when controlling for complexity. 37Also, those J o u r n a l P r e -p r o o f Journal Pre-proof with the largest investments find the choice less difficult, although they score higher on the complexity index than those with smaller investment.

Confidence
Making a fund choice attentively can indeed be challenging and demand considerable effort but would result in making a preference-aligned choice (Löfgren and  Nordblom, 2020).An inattentive fund choice on the other hand does not require any effort, but there is a risk that the outcome of the decision will not be in line with one's preferences.Some individuals are more confident in their ability to make the right choice without effort.This confidence may stem from factors such as familiarity with the decision task, overconfidence, or other factors.The higher this confidence, the more likely that one will make the choice inattentively (why expend effort if you think there is a sufficiently high probability of selecting the preferred option based solely on intuition or gut feeling).We asked the respondents how certain they were that they could pick the preferred option just by intuitive judgement (gut feeling).Here again, the answers were on a 5-degree scale, where 1 indicated Very uncertain that the choice would be preference-aligned and 5 Completely sure that the choice would be preference-aligned. 38.The mean response was 2.7, and only 2.4 percent of the respondents answered 5.
As observed in the third column in Table 5 men have higher confidence than women, and the marginal effects for gender are of the same magnitude as concerning difficulty. 39It should however be noted that we cannot tell whether it is genuine confidence or overconfidence as in, for example, Barber and Odean (2001), who find that men are much more overconfident than women in stock market trading.
Regarding age, the youngest age group has the highest level of confidence, and the oldest the lowest level, and the marginal effects are sizable. 40Further, there is a positive correlation between the size of one's fund investments and the level of confidence in making a fund choice based on gut feeling.One reason for this may be that as individuals accumulate larger fund investments, they gain experience and familiarity with the financial market and its dynamics.Consequently, they may develop greater confidence in their own instincts and intuition.
Concerning education we observe that the higher the education level, the lower 38 Direct translation: "Sometimes we don't think too much when we make choices but listen more to our gut feeling.If you would choose a fund completely by gut feeling, how certain are you that you would make the right choice (i.e., the choice you would have made had you thought it through carefully)?" 39 Table A11 shows the marginal effects. 40For example the oldest are more than 9 ppts more likely than the youngest to have the lowest confidence score.Table A11 shows the marginal effects.

J o u r n a l P r e -p r o o f
Journal Pre-proof the confidence, which is consistent with the interpretation in Section 3.1.2that those with higher education have a higher awareness of the complexity of the fund choice.Finally, knowledge is positively associated with high confidence and the marginal effects are of the same magnitude as for gender. 41verall, we can conclude that young men with a low level of education and large fund investments have the strongest confidence in their own inattentive choice.For given levels of perceived importance and difficulty, they would therefore be the most likely to make the choice of funds inattentively.

Attentive or inattentive?
There is no distinct group that the analysis of each of the three components can single out as the most or the least likely to make an inattentive choice.To gain further insights we therefore combine our three components into the new variable Inattentiveness, for which a higher value indicates a higher likelihood of making the choice inattentively.42 Figure 3 illustrates the interdependence between the three components in determining attentiveness.Combinations of importance (∆U ), confidence (θ), and difficulty (σ) to the upper left indicate attentive choices, while combinations to the lower right imply inattentive choices. 43n column 4 of Table 5 we observe that women and those with low investments score significantly higher on the inattentiveness measure than other groups.The youngest age-group has a higher probability of making an inattentive fund choice since they find the fund choice more important than older age-groups.On the other hand, they perceive the attentive choice to be more difficult and have a stronger confidence in making a preference-aligned choice inattentively, which goes in the opposite direction.Overall, younger have a somewhat higher degree of inattentiveness than older investors.We also find (as expected) that knowledge is negatively associated with overall inattentiveness.
To further validate our measure of inattentiveness we use some descriptives concerning stated behavior for fund choice in general.We asked the respondents about their strategies when they are about to invest in new funds.Respondents answered on a 5-degree Likert scale, where 1 indicates do not agree at all and 5 agree completely.Mean values for the full sample are found in the first column of

J o u r n a l P r e -p r o o f
Journal Pre-proof  6.To gather as much information as possible is a clear indication of making an attentive decision, while picking a random fund, not investing or listen to gut feeling are highly indicative (and correlated) with inattentiveness.Strategies that are significantly correlated with Inattentiveness with p < 0.001 are marked in bold in Table 6.When analyzing in more detail the relationship between respondent attributes and the strategy used when choosing a new fund (see A13 in Appendix A), we observe that men are more likely to gather information, while women seek advice from the bank and are overrepresented in just picking a fund or not investing at all.Young people tend to seek information but are also more likely than others to just pick a random fund or listen to their gut feeling.Those with the smallest invested amounts are more likely than others to just pick a random fund and not to save in new funds at all.They are also the least likely to gather information, while those with the largest invested amounts are the most likely to do so.Including our inattentiveness measure leaves the coefficients almost unaffected, but inattentiveness itself is significantly associated with all strategies except Name of the fund and Sustainability label (see Table A14 in Appendix A).This makes us confident that

J o u r n a l P r e -p r o o f
Journal Pre-proof our variable Inattentiveness captures inattentiveness in a meaningful way.

Sustainability and inattentiveness
As observed in column (4) in Table 5, women, those with small investment, and, to some extent, young people, score higher than others on our inattentiveness measure.Referring to Table 3, it is evident that being young and female is also more common among individuals who are sustainability motivated.Hence, certain attributes are more frequently observed among both inattentive and sustainability motivated investors.To further explore the link between sustainability and attentiveness, let us return to the three components of (in)attentiveness.
In columns ( 5)-( 7) in Table 5, we control for sustainability and return motivation.The overall results are robust to these inclusions.Naturally, both return and sustainability motivated investors find fund investments to be a more important decision than others.44However, most importantly, sustainability motivated investors perceive the fund choice to be more difficult than others do.The explanation to this does not seem to be a higher degree of complexity, as suggested by Pedersen  et al. (2021), as the results prevail also when controlling for complexity. 45We find that the complexity index is positively correlated with perceived difficulty for both sustainability and return motivated investors.
In columns ( 8)-(10), we instead control for the importance of sustainability relative to returns. 46Those results are in line with those in columns ( 5)-( 7), i.e.,

J o u r n a l P r e -p r o o f
Journal Pre-proof fund choice is perceived as less important and somewhat more difficult the more relative weight is put on sustainability.Although the confidence indicator does not turn out significant, these result all suggest that the more one is concerned with sustainability rather than with returns, the more likely one is to act inattentively.
Due to a high degree of multicollinearity and endogeneity, we have not shown any regressions with Inattentiveness as the dependent variable including the preference variables (sustainability and return motivation). 47However, the variable correlates positively with sustainability motivation (p < 0.01), negatively with return motivation (p < 0.05), and positively with the relative importance of sustainability (p < 0.01).Those who are solely sustainability motivated are the most inattentive according to our inattentiveness measure, and those who are only return motivated are the least inattentive. 48Respondents who find both or none of the two motives important are somewhere in between (c.f., the four groups in Table 3).This is in line with the findings by Bassen et al. (2019) that "intuitive" investors place more weight on climate performance than on financial performance, while those with the highest cognitive reflection rank financial performance much higher.
Our measure of inattentiveness is thus positively associated with sustainability preferences and if our theoretical reasoning is correct this could be an explanation to why sustainability motivated investors to a larger degree fail to make investments that are aligned with their preferences.Rerunning the regressions reported in Table 4 with inattentiveness as an explanatory variable, we unsurprisingly find that the likelihood of actively having invested in sustainable funds is decreasing in the degree of inattentiveness (the marginal effects for the other variables are stable).Results are presented in Table 7.The effects are sizable.Our variable Inattentiveness has a maximum value of 5, which means that those who are the most inattentive are 72-89 percentage points less likely than those who are most attentive (scoring just 0.04 on the inattentiveness scale) to have actively invested in sustainable funds. 49n this subsection, we have thus seen that inattentiveness is common among sustainability motivated investors than among those who are primarily return motivated, something that may have consequences for how well actual investments align with preferences.Moreover, our inattentiveness measure is a very strong preinvest in a sustainable fund, ranging from No to Yes, even if the expected return is more than 10 percent lower.
47 Indicative results, however, suggest that sustainability motivation is positively and return motivation negatively related to our Inattentiveness variable. 48The difference is significant, p < 0.001. 49As a robustness check, we run the same regressions, but with the dependent variable indicating whether or not one claims to currently hold any sustainable mutual funds.Those results (that are very similar to those in Table 7) are reported in Table A12 in Appendix A.

J o u r n a l P r e -p r o o f
Journal Pre-proof Marginal effects; Standard errors in parentheses (d) for discrete change of dummy variable from 0 to 1 * p < 0.05, * * p < 0.01, * * * p < 0.001 dictor of not having made an active sustainable investment and we find that those with observable attributes correlated with sustainability motivation make less attentive decisions than do those with attributes correlated with return motivation.This is important information for banks when setting up communication strategies.
The best way to communicate with sustainability motivated investors may not be the same as with those who are return motivated.

Discussion and implications for banks' communication strategies
An important task for banks is how to reach out to their customers to help them find (and invest in) financial assets that correspond to their preferences.The banks have various channels of communication at their disposal, including fact sheets, web pages, meeting with advisors, social media, and newsletters.However, investors will

J o u r n a l P r e -p r o o f
Journal Pre-proof process the content of the communication differently, depending on whether they make their decision attentively or inattentively.
In this setting it is useful to distinguish among three types of communication that work via different decision-making mechanisms (Löfgren and Nordblom, 2020): (i) information that is new to the customer and influences an attentive choice, (ii) nudges that are irrelevant to an attentive choice but may influence an inattentive choice, and (iii) boosts that simplify information with the purpose of reducing the effort required to make an attentive choice.
Regulations stipulate what kind of information banks and other financial institutions have to provide to their customers. 50However, reacting rationally to information requires an attentive decision.We have seen that return motivated investors are more attentive than sustainability motivated investors according to our inattentiveness measure.They are also more likely to gather information.The same holds for men and those with large investment amounts (attributes that are positively correlated with being return motivated).Hence, information may be an effective communication tool when reaching out to men with large investments who are motivated by high returns.However, it may be less effective in facilitating investment decisions by investors who make their decision inattentively.Then, a nudge or a boost would be more effective (Löfgren and Nordblom, 2020).
A nudge can be thought of as a change in the choice context that we would not expect to influence a (rational) choice that is made attentively with full information.Thaler and Sunstein define a nudge as "any aspect of the choice architecture that alters people's behavior in a predictable way without ... significantly changing their economics incentives" (2008, p. 6).A more formal definition is provided in Löfgren and Nordblom (2020): "A nudge is an alteration of an inattentive choice situation, which would not affect an attentive choice."Nudges of relevance specifically for banks' communication about mutual funds are, for example, changes in the order of options on a list, pre-selected options (defaults), different colors of text, changes in fund names, or different types of sustainability labels.

J o u r n a l P r e -p r o o f
Journal Pre-proof 2020).In this context, boosts could be easily available information or search tools that make the fund choice less complicated.
In Section 3, we found that sustainability motivated investors are more likely to make their choice inattentively.In Table 6, we report mean values for sustainability and return motivated investors, respectively, and we see that sustainability motivated investors are more likely to seek advice from the bank than are the return motivated (p < 0.001), which is consistent with the observation that they perceive the choice more difficult.51Picking a random fund, or simply not investing in new funds (i.e., behaviors associated with inattentiveness) are also significantly more frequent among sustainability motivated investors (p = 0.003).This implies that standard information may be less effective in guiding sustainability motivated investors towards preference-aligned investments.
The degree of (in)attentiveness thus varies with investment motive and investor characteristics.This means that the optimal way of communicating with investors also varies.Naturally, there are individual differences, but in general, sustainability motivated investors are more likely to make their investment choices inattentively than those who are return motivated.Hence, sustainability motivated customers are less likely to be influenced by the kind of information that benefits return motivated investors.Instead nudges or boosts, the order of fund options or clear labels may facilitate sustainability motivated customers' fund choices.
Note though, that the same communication content can be a nudge for one investor but serve as information or a boost for someone else.One example is sustainability labels.For an attentive investor who is actively looking for sustainable funds, a label could give the information that certain criteria are met.For someone who thinks that investigating sustainability aspects is not worth the effort, the label could serve as a boost that makes sustainability information more easily available.Finally, a green leaf next to the fund name could nudge inattentive investors to choose the fund, although they may not understand what the label actually implies.This would be in line with Gutsche and Zwergel (2020) who conclude that information costs are important obstacles to sustainable investments and that labels can reduce those costs.Bassen et al. (2019) find that climate labels are more effective among inattentive investors than among attentive ones, indicating that those labels work primarily as nudges or boosts rather than as information.

J o u r n a l P r e -p r o o f
Journal Pre-proof

Concluding remarks
To reach the climate goals, a greater share of global capital must be directed towards sustainable investments.Perhaps an overall increased awareness of sustainability issues is needed, but in this study we have found that also investors who consider sustainability to be an important investment factor to a large extent fail to make sustainable investments.
We find that in a large representative sample of Swedish mutual fund investors sustainability preferences are generally strong.71 percent of our sample would be willing to give up some expected return to get more sustainable investments and 36 percent claim that sustainability is an important or very important aspect when they invest in new funds.Still, as many as 38 percent of the sustainability motivated respondents state that they have never actively invested in a sustainable fund.Interestingly, we corroborate the findings by Anderson and Robinson (2022)  and find that women in general perceive sustainability to be more important than do men, but they do not hold sustainable funds to a larger extent.Moreover, older individuals are less likely than younger to invest in sustainable funds, although they perceive specific sustainability aspects to be more important than do younger investors. 52They also state to be more willing to sacrifice expected returns to make their investments more sustainable.It is, however, important to acknowledge that there may be individuals with pro-sustainable preferences who choose not to invest in sustainability funds due to a lack of options that align adequately with their sustainability criteria.Further research is needed to examine the implications of insufficient availability of sustainable funds that meet the desired levels of sustainability according to individual preferences.We leave for further research to explore the implications of supply of sustainable funds not meeting sufficient levels of sustainability.
The discrepancy between preferences and behavior in fund investment can be attributed to the decision-making process.An attentive choice, which is well-thoughtout and therefore by certainty aligns with preferences, requires effort.However, if the effort involved is perceived as too costly, individuals may opt for an inattentive choice, risking an outcome that deviates from their preferences.The likelihood of making an inattentive choice increases with the difficulty of the decision.Our findings reveal that sustainability-motivated investors encounter greater difficulty in making the investment decision and are significantly more prone to making inattentive choices, potentially resulting in outcomes misaligned with their preferences.

J o u r n a l P r e -p r o o f
Journal Pre-proof Conversely, those primarily concerned with returns find the choice easier and are more inclined to make attentive choices.Additionally, men are more likely than women to make attentive choices and exhibit slightly higher return motivation.
Our findings suggest that banks and financial institutions should employ tailored communication strategies to facilitate preference-aligned investments.Traditional information that caters to return-oriented investors may not be optimal for engaging with sustainability-motivated investors who are more likely to make inattentive fund choices.Instead, for this group, implementing nudges and boosts such as sustainability labels or easily accessible information can be more effective tools to guide investment decisions.Recognizing this is important not only for individual investors but also crucial for advancing sustainable development, given the significant role of the financial sector to steer capital towards sustainable investments.This also highlights the need for further research that applies behavioral insights to the financial domain.A-8 J o u r n a l P r e -p r o o f Journal Pre-proof A-9 J o u r n a l P r e -p r o o f Journal Pre-proof

A-10
J o u r n a l P r e -p r o o f Journal Pre-proof A-14 J o u r n a l P r e -p r o o f Journal Pre-proof

Figure 1 :
Figure 1: Stated importance of sustainability and historical returns (mean values, where 1 indicates completely unimportant and 5 very important) 24 28 J o u r n a l P r e -p r o o fJournal Pre-proof

A
Dependent variable ranges from 1=Completely unimportant to 5Very important.BDependent variable ranges from 1=Not at all demanding to 5=Very demanding.CDependent variable ranges from 1=Very uncertain that the choice would be preference-aligned to 5=Completely sure that the choice would be preference-aligned.D Dependent variable = Difficulty/(Importance * (6-Confidence) t statistics in parentheses * p < 0.05, J o u r n a l P r e -p r o o f Journal Pre-proof 3.1.1Importance

Figure
Figure 3: Attentive and inattentive choices

Table 2 :
Willingness to give up returns for sustainable investment (percent)

Table 3 :
Importance of investment aspects (sustainability and return) and willingness to give up returns for sustainable investments

Table 4 :
Propensity to actively invest in sustainable funds: Probits, marginal effects

Table 7 :
Propensity to actively invest in sustainable funds: Probits, marginal effects

Table A2 :
Stated importance of various aspects of the fund choice (mean values where 1 indicates completely unimportant at all and 5 very important)

Table A5 :
Importance of sustainability aspects: Ordered probit p < 0.05, * * p < 0.01, * * * p < 0.001 Note: The higher the value, the higher the importance.Marginal effects are available on request. *

Table A6 :
Propensity to hold sustainable funds: Probits, marginal effects The dependent variable equals one if the respondent is sustainability motivated but currently p < 0.05, * * p < 0.01, * * * p < 0.001 Note: 1 implies not at all demanding and 5 very demanding.
t statistics in parentheses * p < 0.05, * * p < 0.01, * * * p < 0.001 Note: 1 means Very uncertain that the choice would be preference-aligned and 5 Completely sure that the choice would be preference-aligned. *

Table A12 :
Propensity to hold sustainable funds: Probits, marginal effects Dependent variable is whether one currently holds any sustainable funds.Marginal effects; Standard errors in parentheses (d) for discrete change of dummy variable from 0 to 1 * p < 0.05, * * p < 0.01, * * * p < 0.001