Pension Participants’ Attitudes, Beliefs, and Emotional Responses to the New Dutch Pension System

As of 2022, the Dutch pension system will be overhauled. The success of this as well as other pension reforms also depends on how participants react to and accept such changes. We therefore studied participants’ attitudes, beliefs, and emotions toward the new pension system. We composed a text to inform them about the new system and qualitatively analyzed their responses. We investigated which beliefs and attitudes prevail among different age groups. The results show that many participants base their comments on previous experiences, misconceptions and (sometimes false) interpretations of the information in the text. Moreover, we find that young people are more optimistic, whilst older participants tend to feel victimized. Since the new Dutch pension rules have yet to be introduced, the results of our study contain valuable information for policymakers and pension funds who should acknowledge and address the oftentimes intense emotions, beliefs, and attitudes that influence the way that intentionally neutral information is perceived and accepted. A diversified communication strategy, mindful of different beliefs, emotions and attitudes among participants should help to empower citizens to get insight in their financial situation after retirement and to make informed choices.


3 Introduction
In recent years, Italy, France, Germany, and the United Kingdom have all made changes to their pension system (OECD, 2015). The Netherlands is the next European country to overhaul its pension system in the second pillar, the pension build-up via the employer. The transition will take place in the period between 2023 and 2027. Key changes include the abolition of defined benefit agreements and the uniform premium system and fluctuations in (expected) pensions corresponding to changes in the financial markets. Participants can also opt for pay out of 10% of accrued capital as a lump sum when they retire.
The new system must be communicated to pension plan participants so that they have basic insight in how and what they accrue and can make informed choices. Yet, research shows that many participants lack financial literacy and have incomplete or incorrect knowledge about how pension systems work (e.g., van Rooij et al., 2011). This often results in beliefs (one that is commonly voiced is that "there will be no money left when I retire") and negative emotions such as anger or anxiety (e.g. Eberhardt et al., 2021a, Eberhardt et al., 2021b, which can influence the understanding and interpretation of information, as well as the acceptance of the reform. While research in behavioral economics and psychology confirms that people often rely on heuristics and emotions in their decision making (e.g. Kahneman 2003; Thaler 2016), there is a lack of knowledge about which emotions, beliefs and attitudes play a role in the context of pension reforms and whether there is heterogeneity among participants. Knowledge about the (diversity of) emotions, attitudes and beliefs towards a pension reform is important since they form the backbone against which participants process and interpret information about the reform, and influence their behaviors and acceptance of the reform.
In this study, we investigated pension participants' attitudes, beliefs and emotional responses to the new pension rules in the Netherlands. Specifically, we studied whether a description of the new system evokes positive or negative reactions and whether these reactions are based on the correct comprehension of the system or are instead founded on misconceptions. A manual content-coding of open-ended questions from a survey of 854 participants identified the most critical and frequent beliefs, attitudes, and emotions.
This study contributes to the existing literature in two ways. First, this study identifies the diversity of pension participants' beliefs, emotions, and attitudes after reading information about the new Dutch pension system. It reveals that many participants cannot grasp fundamental ideas about pension due to lack of knowledge and misconceptions, and that factual and intentionally neutral information stimulates a variety of emotions, beliefs and attitudes. The results are a wake-up call for a different type of (pension) communication. Even though knowledge about the relevance of behavioral biases and emotions in decision making is wide-spread among policy makers, pension communication is still predominantly 'rational'. This study reveals that to communicate effectively, the pension sector cannot rely simply on the provision of facts, but needs to acknowledge and address the emotions, beliefs, and attitudes that influence the way that intentionally neutral information is perceived and accepted. We argue that a different type of communication, reasoning, and scope is needed, 1 3 one that works to dispel false beliefs and related negative responses to the pension system, thereby improving not just future pension communication but engagement by pension participants as well.
Second, this study generates knowledge about the heterogeneity of pension plan participants by distinguishing different beliefs, emotions, and attitudes towards the new pension arrangement. Specifically, we identified four different groupings of beliefs, emotions, and attitudes: the optimistic/receptive, the victimized, the "us versus the system" thinkers, and the confused/uncertain. To date, studies on retirement planning have predominantly relied on age and gender to study individual differences, ignoring emotions, beliefs and attitudes that can provide richer insights for targeted pension communication. Policy-makers and pension funds can apply these insights to develop communication that acknowledges and better caters to the different emotions, beliefs, and attitudes.
The remainder of this paper is structured as follows: First, we explain the global contours of the new pension system. Next, we define beliefs, emotions, and attitudes and explain their hierarchy and relationships. We then explain our research design and methodology, after which we present the results of our study. Lastly, we discuss the findings and provide some conclusions.

The New Dutch Pension System
In response to a rapidly changing society, the Dutch government sought to overhaul the pension system, reaching a pension agreement in June 2020. The new legislation in which the new pension system is laid down is supposed to go into effect on January 1, 2023. In short, the strong points of the old system-solidarity and collectivitywill be retained. Yet, the new pension system will abolish the uniform contribution system. This means that the deposit of younger participants will yield more expected retirement income than the same amount paid by older participants. A second critical component of the new pension system is the "collective solidarity reserve." This reserve is intended to act as a buffer to compensate for steep decreases in pension capital (Koolmees, 2020). Another new element is that participants can opt for pay out of 10% of their accumulated capital at retirement as lump sum.
It is expected that the implementation of the new Dutch pension system will increase pensions in such a way that purchasing power is ensured to greater extent than it currently is. Yet, this will depend on the financial markets-i.e., positive market developments will increase, negative developments will decrease the accrual and payout of retirement funds. The developments will lead to larger fluctuations in expected retirement income-especially among young people. Yet, young people have a longer investment horizon than older workers and retirees. Thus, their accumulated pension capital is much more capable to counter setbacks on the financial markets-and, if necessary, the collective solidarity reserve can compensate for financial setbacks. Other articles in this special issue provide a more detailed explanation and assessment of the new Dutch pension system.

3 3 Beliefs, Attitudes, and Emotions
In this study, we investigated the beliefs and emotions and resulting attitudes of pension participants toward the imminent new pension rules in the Netherlands. We focused on beliefs, emotions and attitudes since these are eminent factors in psychology when identifying people's perceptions toward an attribute-in this case, the (new) Dutch pension system. Insights into the beliefs, emotions and attitudes towards information about the new pension contract are valuable and can be used to improve communication about the new pension system so that participants are able and willing to get insight in their financial situation after retirement in the new system and the choices they have. Ajzen and Fishbein (2000, p. 3) defined an attitude as "an evaluation of an object, concept, or behavior along a dimension of favor or disfavor, good or bad, like or dislike." A relevant example of a response reflecting an attitude is the approval or disapproval of a policy. An attitude can be held consciously or subconsciously and is a critical component in most models that predict behavior (Ajzen, 2014;Forgas et al., 2011). People's evaluations of, or attitudes toward, an object are determined by their accessible beliefs about that particular object as well as emotions (Ajzen & Fishbein, 2000). Fishbein and Ajzen (1975) defined a belief as the subjective probability that an object has a particular attribute. More specifically, individuals hold beliefs about the positive or negative association of characteristics, attributes, values, and goals with any given object or concept. Examples of beliefs in the pension domain are 'Investing is (too) risky' or 'Informing myself about my pension is difficult' (Eberhardt et al. 2021b). Beliefs are sometimes based on invalid or selective information, can be selfserving, and/or can even fail to correspond to reality (Allport, 1954;Eagly & Chaiken, 1993;Fishbein & Ajzen, 1975). For example, commonly voiced beliefs about pensions that actually do not correspond to the reality of the (Dutch) pension system are 'There will be no pension left for me when I retire', 'I have to continue working until I get a state pension', or 'I get back less than I personally pay for my pension'.
To date, there is no consensus on the definition of what constitutes an emotion. Kleinginna Jr. and Kleinginna (1981, p. 355) reviewed over one hundred different definitions of emotions and generated a definition that integrates the many possible aspects of emotions: "Emotion is a complex set of interactions among subjective and objective factors, mediated by neural-hormonal systems, which can (a) give rise to affective experiences such as feelings of arousal, pleasure/displeasure; (b) generate cognitive processes such as emotionally relevant perceptual effects, appraisals, labeling processes; (c) activate widespread physiological adjustments to the arousing conditions; and (d) lead to behavior that is often, but not always, expressive, goal-directed, and adaptive". Emotions strongly influence our response to and behavior toward events (Lazarus, 1991). For example, research in the UK showed that a negative experience when being in contact with the pension provider resulted in disappointment and consequently barred further contact (Whitfield et al., 2011). And Schlösser et al. (2013 show that not only anticipated emotions, but also immediate emotions significantly predict decisions under risk, making it highly relevant to examine emotional reactions to the new pension agreement more closely.

3
In line with research in psychology (e.g. Ajzen, 2014;Ajzen and Fishbein, 2000;Fishbein and Ajzen 1975), we argue that a person's belief as well as their emotions influence a person's attitude. This attitude (object) is then expressed toward an attribute-in this case, (aspects of) the new Dutch pension system. We were thus interested in obtaining an impression of participants' beliefs about and emotional reaction to (aspects of) a new pension system and the attitudes cultivated by these beliefs and emotions.

Research Design
The purpose of this research was to investigate pension participants' reaction to an explanation of the new Dutch pension system with reference to their beliefs, emotional reactions, and attitudes. To this end, Brüggen and van Hekken et al. (2020) drafted a text explaining the essence of the new system. The text was written from the participants' perspectives, which means that non-technical terms and explanations were used. For example, the text did not explain the uniform contribution rate system. Instead, it was stated that, in the current system, youngsters annually accrue a lower pension than what they pay for. The text does not contain information about the option of 10% payout at once since it was not yet known when writing the text. The text was non-persuasive, as short and neutral as possible, and as simple as possible in terms of wording. The content was based on the layout of the new pension system at that particular moment and covered the following aspects (see Appendix I for the full version): • What remains the same? -Pensions are still arranged together-you pay money together with your employer in order to receive a pension when you are retired. -You receive a pension as long as you live.
-At a certain age, you will receive AOW, a benefit from the government.
• What is going to change? -Pensions become fairer. Younger and older people receive the pension they pay for.  Thompson et al. 1994). In addition, we measured age, pension involvement (1 -no involvement, 5 -very involved), the grade given to the text (1 -lowest, 10 -highest), educational level, and pension situation (active accrual via employer, receiving pension). We recruited a sample that reflects the Dutch population through the research agency ETIL. Members of the ETIL panel were invited to participate in a research study on pensions. The data were compiled from June 8 until June 28, 2020. Respondents received a URL that directed them to the online survey. We excluded those respondents who do not accrue or receive a pension via their (former) employer.
We received 2,009 responses. In our analysis, we left out comments on the structure or writing style of the text as such. We only included responses that contained information about attitudes, beliefs, and emotional reactions to the new pension rules. This left us with 854 respondents. The low response to the non-mandatory openended questions is in line with prior research, which has established hat open-ended questions produce more missing data than close-ended questions (e.g. Reja et al. 2003, Scholz andZuell 2012).
The average age of the 854 respondents was 51.5 years (SD = 16.35), significantly higher than the age of the entire sample (M = 48.13 years; SD = 14.484, p < .001). The median level of education was intermediate vocational education for both the respondents (Mdn = 5) and the full sample (Mdn = 5). Of the respondents, 26% were receiving a pension (and hence, 74% were accruing a pension), which is significantly more than for the full sample (20.5% receiving a pension and 79.5% accruing a pension, p < .01). The median first impression was 2 (= neutral) for respondents and the full sample (respondents: 42.1% are positive, 47% are neutral, 10.9% are negative; full sample: 42.1% are positive, 49.1% are neutral, 8.9% are negative). Yet, respondents to open-ended questions are somewhat more involved with pensions (M = 3.04, SD = 1.24) than the full sample (M = 2.92, SD = 1.228; p < .01). The grade that respondents give to the open-ended questions (M = 6.90, SD = 1.417) and the grade that the entire sample (M = 6.78, SD = 1.493) gives to the text is significantly different at p  Table 1 summarizes the characteristics of the full sample and respondents to open-ended questions. In conclusion, respondents who responded to the openended question were somewhat older, were more often retired, were more involved with pensions, and graded the text slightly higher than the complete sample. Thus, responses to open-ended questions were not completely representative for all respondents to the survey. Yet, the purpose of the study is not to analyze the extent to which different beliefs, emotions, and attitudes occur in the Dutch population, but to get a rich understanding of the diversity of beliefs, emotions and attitudes and whether there is heterogeneity among respondents.
Responses to the open-ended question were at times rather long, respondents used different expressions and wording for similar points, and a variety of sayings. Therefore, responses could not be processed automatically but we had to code them manually. To identify respondents' beliefs, emotions, and attitudes, we used a thematic analysis. Thematic analysis is a well-known procedure to analyses qualitative information, and we followed established guidelines to rigorously and methodologically analyses the information (Nowell et al. 2017). The goal of the analysis is to examine the open-ended questions in a precise, consistent, and exhaustive manner through coding, systematizing, and iterating. The next section will explain the exact steps that we have taken.
First, the researchers extensively trained a coder who was not involved in designing the research and collecting the data in interpreting and understanding the comments. This was done by providing the coder with relevant background information, instructions on how to analyze the comments, and a hands-on training whereby the researchers coded the first 100 responses together with the coder. The researchers and the coder started to closely read the comments and created codes for beliefs, emotions, and attitudes. For example, one comment was: [translated] "I personally think that people have been the victims of the whole system for a long time. I have already paid pension for more than 30 years, but the entire investment has been invested with the risk that nothing remains. I find it outrageous that the money I earn is treated like this. It's mine and not someone else's so stay away from it" The coder observed the belief 'Investing is (too) risky' and noted a code for this belief, the emotion "angry" received a code under emotions, and the attitude "Protective (stay away from my/our pension)" was coded under attitude. Then the researchers and the coder continued with the next comment. If the comment contained a belief, emotion, or attitude that was already noted in an earlier comment, they noted the code that was created for that specific belief, emotion or attitude. If the researchers and the coder identified a new belief, emotion, or attitude, they created a new code and added this code to the respective list with codes for beliefs, emotions or attitudes. In this way, we created three lists with codes for beliefs, emotions, and attitudes, and a database, where each row contained the basic characteristics of one respondent, his or her comment, and codes for the content of the comment. This inductive approach did not impose a predetermined structure, but ensured that all beliefs, emotions, and attitudes that emerged from comments were observed, coded and counted. The coder then proceeded coding the next 100 comments. Once the coder finished coding the second 100 comments, the researchers reviewed and verified the coding. They then revisited the first block of a 100 responses to verify the consistency in coding and that 1 3 we did not neglect any information in the first block that emerged more strongly in the second block. The remaining comments were independently coded by the coder and he continued in this iterative manner: he analyzed a block of 100 comments, and then revisited all previous coding to ensure consistency and completeness. After coding 400 comments, we once more discussed the structure of the codes and the level of detail. We followed this method since the number of comments was so large that it was impossible to create an encompassing and tight coding scheme in advance. Only by thoroughly reading and interpreting texts, we were able to identify the codes, discover which codes touched upon similar points, and which codes could be more refined.

Results
To provide a descriptive overview over the coded data, Table 2 displays the top three beliefs, emotions, and attitudes per age group. We divided the sample into four age groups: young workers (< 34 years), middle-aged workers (35-49 years), older workers (50-64 years) and retirees (> 65 years). The new pension system bears different implications for different age groups. Young workers just start to accumulate their pension, and will build up most of their pension under the new scheme. Middle-aged workers have already accumulated a substantial part of their pension under the old system, but also still have to accumulate a substantial part under the new system, and will be most affected by the transition. Older workers and retirees are in a different work situation, but both built up most if not all of their pension under the old system.
The most accessible beliefs and emotions can be considered the most prevailing determinants of a person's attitude (Ajzen & Fishbein, 2000). The most prevailing attitude is positive in all age groups. The most common belief is that the belief that the new pension rules ensure a fairer distribution of pensions. However, in all age 1 3 groups, except the young workers, people believe that the new rules will negatively impact their financial situation. This is interesting since this belief does not align with the information in the text they read. The most prominent emotions were skepticism, hope and worry. Interesting to note is that we identified fewer emotions in the openended comments, than beliefs and attitudes.
Young workers (< 34) were on the whole rather optimistic about the new pension rules: "It seems like a good plan to better organize a system of pension. As a young person, I am already regularly concerned about later. This offers a perspective that things can be done differently and that pensions remain essential." Young workers perceived the new rules as an improvement and as being much fairer: "This gives me the idea that I, as a young person, will be treated more fairly concerning pension contributions. After all, quite a bit of pension is paid. This makes me happy." Some younger respondents however, questioned the consequences of the new rules: "I don't know if this is going to be an improvement." Other younger workers (7.8%) seemed uninterested because retirement was still far away for them: "I have little interest in it (pensions) at the moment. It will certainly take more than 30 years before I can retire." Youngsters had diverse views when it came to the connection between the economy and pensions. While most understood that their contributions were being invested (and their estimated pension was, therefore, dependent on developments of financial markets), some indicated that they don't know what to expect: "I don't understand why pension fluctuates based on the economy. Especially in these uncertain times. This makes me feel insecure." Somewhat worrying is that a part of the young workers who felt positively about the reform, did not understand that the reform means that their (estimated) pension will fluctuate more in the future: "Pensions become more stable in value this way. This ensures fewer fluctuations in the future." The middle-aged workers (35-49 years), expressed more concerns. They were prominently worried, especially about the possibility that the new rules would affect their financial situation: "I am in the age group that will probably be affected most. I am afraid that this will cost me in the future." Furthermore, the percentage of middleaged respondents with a positive attitude towards the new pension rules was lower than that for younger respondents. They were less hopeful and more often expressed that they are worried: "The question is: is this really necessary? I prefer assurance, not fluctuations." Analyses of open-ended questions for older workers (50-64 years) revealed many similarities to middle-aged workers. Although members of this older group still believe the new rules to be much fairer than the current system, they are less positive than those in the younger age groups. Instead, they have a more tentative view, being more skeptical ("Everything sounds like it is well organized, but that is always a matter of waiting in this country"), worried ("I am concerned if there is enough money"), and less hopeful. Furthermore, when compared to the other groups, their attitude is the least positive or even decidedly negative. Although the text stated that the new rules will not significantly impact older workers, these findings clearly demonstrate that this group worries the most: "I belong to the group that is getting cut. I still have to work for several years. Experience shows that what sounds beautiful on paper, in reality, often turns out differently (read: negatively). So, I am afraid 1 3 that I have paid for 47 years (!), and after retiring, I am below the social assistance standard." Many older workers asked questions and wanted to discuss matters. Some even very plainly stated that the provided information is not true. And others voice doubts about the underlying motives for implementing a new pension system, or disbelief the expected effects. Van Solinge and Henkens (2017) concluded that older workers are more likely to worry about their ability to keep up in their jobs. In the context of this study, it may be possible that this effect was further strengthened by beliefs about the new pension rules. Another factor that could explain these more negative emotions is the increasing state pension age in the Netherlands: "While we have to work longer, pension contributions have doubled in recent years! We need to go back to when people were allowed to retire at 60!" This response is an excellent example of an adverse reaction to the text based on a person's beliefs. This respondent was convinced that the legal state retirement age should be decreased and expressed his/her displeasure.
The retirees (65 years and older) generally understood the principle of the new rules quite well. As in all age groups, the belief "this is fairer for the young workers" was the most common belief. Some retirees understood that the current system had to change ("This new system is much sincerer for everyone. Well thought through") and thought that young workers deserved a good pension and future ("Everyone, and especially young people, is entitled to receive a good pension"). Those who perceived the new rules to be fairer also addressed the fact that their pensions will be less susceptible to fluctuation compared to the (estimated) pension of the young workers: "This is a good solution in my eyes because older people are better protected against fluctuations. Younger people have much more time to cope with possible windfalls or setbacks." At the same time, retirees were more skeptical than any other age group. Compared to older workers, retirees expressed less negative attitudes. When present, negative emotions and attitudes often stemmed from the absence of a pension indexation in recent years: "I am 82 now, and my pension hasn't been indexed for ten years, so the current system is no good. But I wonder if the new system is better for me. And on top of that: by the time the new scheme comes through, I will probably not be here anymore. I was promised in the past that my pension would be stable in value, but it deteriorated thousands of euros." Their trust had been violated before due to an absence of pension indexations, which is a possible explanation for why this group was so skeptical. Furthermore, some retirees felt like all of their bygone hard work was being abolished: "I worked hard for my pension. But now I am retired and have a pension gap. I need to live on my savings. I want to receive what is mine!" Subsequently, some retirees perceived the regulation changes to be unfair. There was a sense of injustice among them, resulting in some fierce emotions and attitudes. The idea that their pension will be "stolen" (i.e., lowered) made many of them worried or angry. They perceived their pension as their money, and as such other external entities (i.e., government, pension funds, youngsters) had no right to interfere. "Ridiculous! Cutting pensions off people who have worked hard for it is absurd! I have worked 40 (!!!) years to build up a pension. And you want to cut it, what are you thinking? Do you even have any sense?" Remarkable was that nowhere in the text was stated that retirees' pensions will be cut. This reaction was based merely on an assumption, one 1 3 which translated into fierce emotional comments on the text. There were relatively many similar reactions, especially among older workers and retirees.
An important insight when analyzing the results is that the top 3 beliefs, emotions and attitudes were sometimes, but not always expressed in combination with each other. This explains the low percentages: the code lists with beliefs, emotions and attitudes were quite long, due to a wide variety in responses. However, we saw patterns in the combinations of beliefs, emotions and attitudes. Specific beliefs were often expressed in combination with specific attitudes and/or emotions. To examine the sets of emotions, beliefs, and attitudes, we closely examined our codings and identified, which emotions, beliefs, and attitudes are mentioned in combination with each other. This analysis resulted in the identification of four common sets of beliefs, emotions and attitudes that were mentioned in combination.
The first set of joint beliefs, emotions and attitudes that emerged from the data centered around a positive expected impact on the personal situation and a positive attitude towards the change itself. 39% of all comments reflect these beliefs, emotions, and attitudes. A dominant belief in this cluster was that the new pension rules were much fairer and improved, especially for youngsters. Comments state that the pension system was due for a change. Especially considering that changing jobs happens more frequently in modern society, the new rules were considered a great response to such a trend. Comments stated that the new rules created a feeling of solidarity, and that the connection between the economy and pensions is realistic. However, we discerned that a reasonable amount of this optimism was based on misconceptions. For example, one respondent liked the idea of pensions fluctuating with the economy but stated one sentence later that "It is not fair that when you retire, you will receive less money than you paid." A few comments indicated that some respondents tended to forget that if the economy were to perform poorly, their pension would lose value. In addition, several respondents did not understand that the new system also intended to contain fluctuations for retirees; consequently, younger peoples' pension estimations would fluctuate much more. This connection did not always get through to people. Although much positivity was expressed about the new rules, respondents sometimes only saw the positive side and neglected the risks. Below are some direct quotes by respondents.
"Pensions fluctuating based on the economy sounds realistic. Frequently changing jobs is more of this. It is not fair that when you retire, you will receive less money than you paid." (ID3742).
"It makes sense that pensions change; times have also changed." (ID3464). "It seems as if this way of accruing pension is transparent and cheaper, with taking many changes of jobs into account. This is better suited to the current labor market." (ID4781).
"Very good and fair that these new rules ensure that young people accrue more pension. I also like the connection of pensions and the economy." (ID626).
"This will make sure everyone's pensions will be higher." (ID5392). "Clear new rules, and rules that are fair for everyone." (ID2519). "… I think this is a neat arrangement for the collective." (ID2377). "I think this is much fairer. Much more attention is now being paid to the economy and young people. I do hope that everyone gets a decent pension to which they are entitled." (ID5841).
"This seems like a fair system without major fluctuations." (ID6500).
A second set of joint beliefs, emotions and attitudes focuses on a negative expected impact of the reform on the personal situation. 11% of all comments reflect this set of beliefs, emotions, and attitudes. Comments to the description of the new pension rule show that many respondents in this cluster viewed the new system in light of the lack of indexation in the past years. They expect the pensions to be (further) lowered when the new rules apply. Therefore, the new rules' principles did not seem new to them, they perceive the current situation as negative and the new situation will only worsen it. The lack of indexation is seen as negative and unfair, and overshadows the fact that participants built up or receive pension within a high-quality system that usually has a good ratio between the deposits and pay-out. Many comments reveal disappointment.
"This will again cost money." (ID4398). "I haven't been indexed in ten years. This will also be the case with the new rules." (ID3061).
"I am going to retire in 18 years. I am afraid that I am part of the group who will be negatively affected." (ID3622).
"My pension is deteriorating, and I have little faith in the supplement. It is also always said that the invested pension money would probably increase, but the chance that it will decrease is just as great, leading to people my age receiving even less pension, which is not right. The current people who have a pension can live very generously, but it will soon be difficult for my low-educated generation to make ends meet." (ID1766).
"This causes even more uncertain returns. I understand a large group is going to deteriorate, which I am probably a part of. I want more clarity about the supplementation." (ID7920).
"I have paid much more than I will receive once retired. When I turn 65, I will already receive a quarter less than before. And, I even have to work two years longer!" (ID2704).

3 "…Take a savings account. At least you know what you will receive." (ID8482). "This results in even more insecure returns" (ID7920). "Investing remains risky and is by no means always positive, not even in the long run. I think that the supplementary pension I will receive will not be enough. Thus, I will (again) lose money. (ID5104)
"It is weird anyway to invest money instead of saving it. What do we do if an investment falls short but the economy performs well? It's a shame we cannot save anymore. That gave much more certainty. Investing is a big risk…" (ID2249).
"… I don't get why pensions have to fluctuate with the economy. Especially in these uncertain times." (ID1374).
A third set of beliefs, emotions and attitudes mainly focus on the system (13% of all comments). These comments reflect opposition, a lack of understanding and agreement to the system behind the pensions, and negative emotions and attitudes. The comments in this cluster indicate that some respondents do not see the advantages of the new system and distrusted the government's interference in pensions. A common belief was that the new rules were being introduced for pension funds to earn money, based on the misbelief that pension funds are commercial institutions that do not act in participants' best interest. Many respondents in this group expressed the belief that pension funds pay out less than participants' deposit. Several comments also revealed the belief that the government pulls all the strings. Many of these comments indicate a deep misunderstanding of the pension landscape in the Netherlands, where pension funds are non-profit organizations and agreements on pensions in the second pillar are made by employers and employees. The negative beliefs and attitudes show that the old system was not considered to be any good, but in the eyes of respondents, this new system was not the solution either. Another reoccurring belief in this group was that pension regulations change very often. This might strengthen distrust and a negligent attitude.
"Pensions change regularly. So, all these new laws and regulations don't really interest me at the moment." (ID6612).
"You can't just cut retirees' pensions like that. We have to maintain the current rules." (ID5504).
"Too much confusion and rules. Pension funds have more than sufficient money, but our pensions are not indexed for years. What should I believe?" (ID1329).
"I think you will lose money anyway. I am 39 now and have been accruing pension for several years. Now the system is being overhauled, and maybe that will even happen twice before I retire." (ID5380).
"I can already tell where this is going-the total breakdown of our pension system. All because other hobbies to be paid for with it (EU). The ordinary people always will soon be fobbed off with a small pension. Everything is so transparent." (ID7857).
"I am so tired of these rules constantly changing." (ID5508). "I am not sure if this new system is an improvement." (ID6137). The fourth set of beliefs, emotions, and attitudes centers around the impact of the reform on the personal situation and the uncertainty related to it. 37% of all comments reflect this set of beliefs, emotions, and attitudes. There were two different flavors in the emotions and beliefs: Some respondents were confused because they did not understand the message we intended to tell. These respondents were doubtful and 1 3 skeptical with regard to the impact of the reform. Other comments in this set showed a neutral of even interested response. Young workers, for example, thought that the new rules are promising, but they indicated that they could not draw an informed conclusion. They emphasized that much can still change in their professional and personal lives. Both groups concluded "let's wait and see", but their perspective is quite different.
"I am wondering if this is going to succeed." (ID4184). "I'll believe it when I see it." (ID1755). "The intentions seem good, but I'll just wait and see. As a retiree, you are at the mercy of power." (ID7281).
"I find it too unclear and do not see what is different from the current system. Now employers and employees are already putting money aside for retirement. Nowadays, you are often told this money is invested, and that does not yield enough return… I do not understand this because investments are often positive and rising (according to the news), and this is never the case with pensions. Furthermore, given the current Corona crisis, many older people have died, so you would say fewer pensions need to be paid (bluntly). (ID8652) "… This change significantly impacts me because I am middle-aged myself. Therefore, I am wondering if the supplement is high enough. Although I support the fact that something must change to the pension system in order to keep it sustainable in the future…" (ID6675).
"I first need to see how this works. It sounds very nice, but I don't know if this works when put to practice." (ID1480).
"I am still young, so I cannot really judge this." (ID7399). "I first have to look at how this will work out in practice for me as an individual." (ID6756).
"I have little interest in this at the moment. It will certainly take more than 30 years before I can retire." (ID8370).
"My retirement is so far away that I wonder if I am ever even going receive a pension." (ID6055).
We concluded that expected impact (positive or negative) of the reform on the personal situation was an important factor, as well as the evaluation of the system and the change in general.

Conclusions
This study was the first to investigate on a large scale the beliefs, emotions, and attitudes of workers and retirees when informed about the new Dutch pension system. Participants to this study received a text explaining the essence of the new Dutch pension system. The text was factual and intentionally neutral, and avoided technical terms and complex explanations. 854 open-ended responses formed the input for our analysis. While such a qualitative approach is not very common in economics, research in other fields has shown that such an analysis can provide rich insights into how participants see the world (e.g., Belk et al. 2013, McCracken 1988, Thompson et al. 1994. Understanding beliefs, emotions and attitudes to the new pension system 1 3 is highly relevant, since they influence the way that communication about the new pension system is perceived and accepted. Our results verify that many participants base their comments on previous experiences, misconceptions and (sometimes false) interpretations of the information in the text. Results showed clearly that Dutch participants are not blank when they receive information about the reform. They are already filled with and colored by previous communication from pension providers or other information sources around them. The idea that factual, and even very simple information, is processed straightforwardly, is not sustainable. When this is not taken into account while creating information materials, there is a risk that the target group interprets the message incorrectly. Informing people, even if done as neutrally as possible, can evoke resistance due to existing beliefs, emotions and attitudes. This resistance can obstruct information processing and challenge the acceptance of a new pension system.
We also confirmed that participants' knowledge about the current Dutch pension system and the role of pension funds is limited. The open comments demonstrated that various principles of accumulating pensions/the Dutch retirement system were not understood correctly. For example, people lack knowledge about how pension funds deal with participants' contributions. At the same time, they are convinced that they know 'how it works' and have firm beliefs. Many responses revealed this contradiction between a lack of knowledge and strong beliefs.
In addition, we identified heterogeneity in emotions, beliefs, and attitudes. First, we divided the sample into four age groups since the new pension system bears different implications for each age group. Second, we studied patterns in responses, and identified which beliefs, emotions and attitudes are often expressed in combination with each other. With respect to age, we find quite some differences between age groups (see Table 2). For example, young workers were on the whole rather optimistic about the new pension rules. Middle and older aged workers already expressed more concerns and were more worried and skeptical. When studying the pattern of responses, we identified four sets of beliefs, emotions, and attitudes, which are often stated together. First, some respondents are optimistic and receptive, and understand that the system needs to change. They do see the advantages of the new system and belief that it is fairer and will positively affect their pension. Second, respondents perceived themselves as victims of the new system, are angry, afraid and taunted, and belief that their pensions will be (further) lowered. A third set of joint beliefs, emotions, and attitudes reveal distrust, suspicion and aversion, where respondents believe that there is a hidden agenda, the government meddles in pensions, and pension funds profit from the change. Fourth, respondents were confused and uncertain. They either lacked knowledge or do not oversee the impact (yet), and hence "wait and see".
This study contributes to the existing literature in several ways. First, this study identifies that factual information about the new pension system resulted in a variety of beliefs, emotions, and attitudes, that have a strong influence on how the new pension system is viewed and accepted. The methodology as well as focus of the study are novel in the field of economics, where the research interest generally centers on macro or microeconomic questions of the new pension system as such. Communication of the new pension system is often more of an afterthought whereby the approach is typically very rational: let's just tell them factually how the new system 1 3 works. Our results imply that simply providing the facts is not enough. Policy makers and pension providers need to acknowledge and address the oftentimes intense emotions, beliefs, and attitudes that influence the way that intentionally neutral information is perceived and accepted. Second, this study identified heterogeneity in beliefs, emotions, and attitudes among pension plan participants of different age groups. Moreover, we identified four sets of beliefs, emotions, and attitudes, that often occur together. These results emphasize the importance of pension communication that is relevant and tailored for groups of people that hold different emotions, beliefs, and attitudes. In the next section, we give some practical advices, based on the insights we obtained from this study.
Policy makers, social partners and pension providers have to communicate about the process towards the new pension system and the upcoming changes. Communication has to answer the question why a reform is necessary. Based on our results, we argue that legislation should not oblige pension providers to inform participants and retirees with detailed, technical information. Instead, we argue for a clear and short statement why the reform is necessary. Even the very basic text that we used in this was already perceived as too difficult by many. Moreover, our results show that communication about the new system can evoke uncertainty or the belief that pension will deteriorate. It is important to keep in mind that participants and retirees are not blank, but 'filled' with beliefs and emotions. As a result, they have a certain attitude, that strongly influences the response on communication. It is important to create materials that connects with these beliefs and attitudes, especially for respondents who feel victimized and distrust.
Next to answering the 'why', it is important to communicate in a manner that misunderstandings are eliminated and certainly not created. Clear and simple language, no technical jargon and appealing materials can help to convey a basic message. It is important to acknowledge the limited role of information. Instead, policy makers, socials partners and pension providers can use other techniques, like persuasive communication and specific call to actions with clear instructions on which steps to take. Information should be layered, whereby the most essential information is actively communicated through emails or letters. Participants who are interested to learn more, can find more details or background information in a next layer. This layering technique, with persuasive elements ('we made it easy for you, it only takes you 2 minutes to read this') can help to engage participants, especially those who are still a bit confused and uncertain, and do not yet understand the impact. Providing only the most relevant information is also especially important for young people and other participants who believe that a pension reform does not concern them. We expect that communication that empowers participants to check the effect of the new pension system on their expected retirement income will be most effective. As our results show, many participants are interested in or worried about the impact of the reform on their personal retirement income. Yet, to understand and interpret their new situation after the reform, participants have to perform active behaviors: they have to give meaning to the new amounts that will be communicated and follow the personal amounts over time. Although it is widely known that communication is not processed in a rational way, many pension providers still fail to acknowledge that the presence of beliefs, emotions and attitudes requires a different way of com-1 3 municating with participants. Pension information is still too often formulated from the perspective of the pension provider, complex, and technical. To alter an attitude, or to activate participants to get insight in their financial situation, other forms than communication are needed that informing participants. Another conclusion that we can draw from our findings that many participants and retirees want to know 'what's in it for me?' is that pension providers should carefully consider the timing of providing about the new pension system. As long as the impact of the new pension systems cannot be shown as € changes in the expected retirement income per participant, it will be challenging to engage people in a conversation about the new rules.
The heterogeneity in beliefs, emotions, and attitudes triggers the question whether it is useful and feasible to segment communication about the pension reform. Since we found difference between age groups, the impact of the reform and the call for action differs for different age groups, and segmentation based on age is feasible, we recommend to segment communication about the reform based on age. The four sets of beliefs, emotions and attitudes that often occur together are more difficult to use as a basis for more personalized communication, since pension providers or policy makers cannot identify who holds which set of beliefs, emotions and attitudes without running a survey first. However, pension providers or policy makers should keep the different perspectives in mind and craft messages that integrate these insights and thereby appeal to as many people as possible.
Future research could focus on a continuation study investigating whether participants in this study altered their pension behavior after participating in this research. A follow-up study could examine whether their pension involvement-or their beliefs-changed. Furthermore, a similar study should be conducted once the new rules are initiated. Comparing these results with those of the current study could be valuable for determining whether progress has been made regarding communication. An additional study would be especially important, when certain aspects of the reform are more clear. For example, if certain groups that would stand to lose from the transition are not compensated, beliefs, attitudes, and emotions would most likely be more negative than expressed in this study. Another study could research the effects of concrete calculation examples on people's beliefs, emotions, and attitudes. Making the effects of the reform more tangible might decrease the room for individual interpretation of pension principles.

Appendix I: Text Used to Inform Participants About the New Pension Rules
Waarom verandert ons pensioen?

3
Een deel van ons pensioen krijgen we van de overheid: de AOW. De AOW betalen we met z'n allen via de belastingen. Een ander deel van ons pensioen regelen we via onze werkgevers. We betalen samen met onze werkgevers voor dit pensioen. Ten slotte kan iedereen zelf nog sparen of beleggen voor zijn pensioen. Dat alles bij elkaar maakt dat we in Nederland allemaal een inkomen hebben om van te leven als we met pensioen gaan. De nieuwe regels gaan over het pensioen dat we regelen via onze werkgevers.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creativecommons.org/licenses/by/4.0/.