HOW TO BEAT PROCRASTINATION
THROUGH LETTER
INTERVENTIONS?
An experiment to seduce agricultural
entrepreneurs to ll in their Combined
Submissions earlier in the submission period
Hannah Nijsingh
10003365
Master Economics: Behavioural Economics and Game Theory
University of Amsterdam
Ministry of Economics Aairs
UvA supervisor
Prof. Dr. T.J.S. Oerman
Internship supervisors
Dr. E. van de Veer
Dr. T. Dirkmaat
19-07-2016
Abstract
Procrastination is a harmful form of self-regulatory failure and occurs in many contexts in life. Procrastination is the voluntary and irrational postponement of an intended course of action despite the knowledge that this delay will come at a cost or have negative eects on the individual. The government suers from the procrastination of people.
Submission periods can be come increasingly crowded towards the end of a period. The Ri-jksdienst Voor Ondernemend Nederland (RVO) also suers from this problem. The problem is approached with letter interventions. Sending letters with an earlier target date than the deadline had an eect, as well as a calender that signalled quiet and peak days on the site and call center for submitting the requested information. This reduced the crowds at the end and spread the submissions more equally across the submission period. However the eect size is not that big and therefore the prevention of procrastination needs more investigation. There is a whole group of farmers that does not ll in the GO, because there are no bene ts for them and no repercus-sions. One half of the group got a standard reminder letter and the other half of the group got an intervention letter with a social norm and a deadline date included. The intervention letter had signi cant more response than the control letter.
Statement
This document is written by Student Hannah Nijsingh who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Acknowledgement
I would like to thank my internship supervisors Evelien van de Veer and Thomas Dirkmaat from the Behavioural Insights Team of the ministry of Economic Aairs. I would like to thank them for their motivation, enthusiasm and knowledge. Their guidance helped me a lot with writing this thesis. I could not imagine a better place to have an internship. I would also like to thank my thesis supervisor, Theo Oerman from the University of Amsterdam. Thanks to his exibility, stimulation and critical view it was possible for me to write the thesis at an internship.As last I would like to thank Hein Holtrop from RVO for providing the database.
Contents
1 Introduction 3
2 What is procrastination? 5
2.1 What is procrastination? . . . 5
2.2 Task properties of procrastination . . . 5
2.3 Personality traits of procrastination . . . 6
2.4 Cognitive mechanisms . . . 7 2.5 Conclusion . . . 9 3 Discouraging procrastination 10 3.1 Reminders . . . 10 3.2 Self-imposed deadline/commitment . . . 10 3.3 Reciprocity . . . 11 3.4 Social norms . . . 13 3.5 Framing . . . 15 4 Casus 17 4.1 Field investigation . . . 18 4.2 Letter interventions . . . 20
4.2.1 Making the core components clear . . . 20
4.2.2 Introducing persuasive messages . . . 21
5 Method 22 5.1 Division groups . . . 22
5.1.1 Division sectors . . . 22
5.2 Interventions . . . 22
5.2.1 Control group . . . 24
5.2.2 Target date 22 April with an appeal to the common interest . . . 24
5.2.3 Target date 29 April with an appeal to the common interest . . . 24
5.2.4 Calendar with quiet and peak days . . . 25
5.2.5 Flyer and information pamphlet parcel registration . . . 25
5.2.6 Unexecuted interventions . . . 26
5.3 Non-response group . . . 27
5.3.1 Control group . . . 27
5.3.2 Social norm group . . . 27
6 Hypothesis 28 6.1 Unforeseen circumstances . . . 28 6.2 Hypothesis 1: . . . 28 6.3 Hypothesis 2: . . . 28 6.4 Hypothesis 3: . . . 28 6.5 Hypothesis 4: . . . 29
6.6 Hypothesis 5: . . . 29 6.7 Hypothesis 6: . . . 29 6.8 Hypothesis 7: . . . 29 6.9 Hypothesis 8 . . . 30 6.10 Hypothesis 9: . . . 30 6.11 Hypothesis 10: . . . 30
6.12 Hypothesis non-response group . . . 30
7 Results 31 7.1 Overall results . . . 31
7.2 Testing hypothesis . . . 31
7.3 Conclusion . . . 37
7.4 Results non-response group . . . 37
8 Discussion 38 8.1 Addition to literature . . . 39 8.2 limitations . . . 39 8.3 Future research . . . 40 9 Bibliography 41 10 Appendix 44
1
|
Introduction
Procrastination is a widespread and harmful form of self-regulatory failure that occurs by humans in dierent settings. It is safe to say that everybody in his life has at least considered delaying something, while delaying was not the best choice to do. Procrastination is present in dierent settings. Think of situations related to health, studying, work etc. The extent to which people procrastinate depends on dierent features. It has to do with the task properties as well as with the character trades of people. Next to that procrastination can be more present at certain ages or certain situations.
People deal with procrastination in their day to day life, but bigger institutions can also suer from procrastination. For example, the government. The government has to deal with the behaviour of her citizens, including the be-haviour of procrastinators. Policy makers have to deal with procrastination in the execution of policy with the information or actions they ask from citizens. The government asks citizens to declare their taxes, they handle the requests for extending passports/driver's license and these processes often have a deadline date. The behaviour of procrastination can be a problem in the mentioned examples, because it can cause large crowds at the end of the period the citizens have to comply with the request from the government. It can occur that the government cannot meet the increased demand, when it gets too crowded. This can be a problem because the capacity at the end of the period cannot be extended in nite. The increase of capacity is associated with extra costs and at a certain point the quality cannot be preserved. The government wants to provide good service and due to peak periods and rush hours the service level decreases. This does not improve the relationship between citizens and the government and will not stimulate the participants to cooperate earlier on in the process.
The Rijkdienst voor Ondernemend Ned-erland (RVO) is experiencing problems with an overloaded call-centre and website due to procrastination every year with the submission of the Gecombineerde Opgave (GO). With the GO farmers have to deliver information about agricultural census, registration of emissions, manure legislation and subsidies. The informa-tion will be used, among other European and national statistics for policy development and at the outbreak of animal and plant diseases. Each year the farmers have a period of 6 weeks to ll in the GO. The GO opens at the 1st of April and closes at the 15th of May. If the farmers do not meet the deadline, they will get a discount on the subsidies they receive.
In the past years RVO experienced that the last two weeks of the period are by far the busiest. 60% of the submissions are done in the last two weeks of the period, of which 40% submitted the GO in the nal week. The ICT systems and the call centre suered from the peak period and therefore the service level went down. The RVO cannot change the 6-week submission period. This period is de ned by the European Union and every member state must adhere to these 6 weeks. Each member state has some space with the starting and ending of the 6 week period; the period has to be between February and May, but it cannot be longer than 6 weeks. The problem that RVO has is that they cannot guarantee a good service level in the last weeks and RVO really wants to improves this.
The farmers also suer from the overload of the site in the last 2 weeks of the period. They have to wait a long time before they can reach the call center and if they get an employee on the phone the chance is high that this person is not a specialist. This is because RVO has to employ temporary workers in order to cope
with the pressure on the call center in the last 2 weeks. The site does not work as good as in the weeks in the beginning of the period, because too many people log in on the site at the same time. By postponing the GO to the last week it will cost the farmers more time and more stress and they risk the chance of an un nished submission. If they wait to the last moment, the farmers might not get an answer on their question from the call center or the site is not available due to overload. Therefore, it is more rational for the farmers to submit the GO earlier in the submission period
That is why RVO asked the Behavioural Insight Team (BIT EZ) from the Ministry of Economic Aairs to come up with interventions to try to spread the submissions of the farmers more equally across the 6-week period. If the submissions are spread more equally among the 6-week period, the farmers can be helped more accurately through the call center and have a better working program. The service level will increase if the peak period in the last 2 weeks can be reduced and is spread more equally over the 6-week period. The goal of RVO is to increase their service level and this can be done by removing the crowds in the last 2 weeks through reducing the procrastination behaviour of the farmers.
To try to get a clearer image of the farmers a eld investigation is done to get information about the GO, the farmers and the struggles the farmers have with lling in the GO. Sev-eral farmers are interviewed, as well as some accountancy bureaus.
Based on the eld investigation and the lit-erature investigation there are 3 interventions formulated, next to a control letter, that could be placed in the invitation letter the farmers re-ceive each year at the start of the period.
2 of the interventions letters presented the farmers with a target date, respectively the 22nd of April and the 29th of April and the third in-tervention letter included a calender that sig-nalled the quiet and peak days on the site and at the call center for lling in the GO.
Each intervention letter did have an eect compared to the control letter. All the inter-vention are signi cantly dierent at a 1% level from the control letter and submitted their GO earlier on in the period. The farmers that lled
in their GO through an accountancy bureau are not in uenced by the intervention letters. Some farmers received certain information after the submission period started and without this information they could not start the GO. These 2 groups of farmers are dierent and therefore are not included in the results. The results are signi cant dierent from the control group, but still the overload of submissions is done in the last 2 weeks of the submission period and the eect size of the results is relatively small.
Each year there is also a group of farmers that does not submit the GO. They do not comply with the request of RVO to ll in the GO, because this group does not receive a subsidy or other bene ts from submitting the GO and there are no repercussions for not lling in the GO. To persuade them to still ll in the GO, even though the submission period was over, they received another letter after the deadline date of the 15th of June. We divided this group in half and one half of the group received a control letter and the other half received a letter with a social norm and a deadline included. The group that received the letter with the social norm had a higher submission rate than the control group 38,41% compared to 32.58%.
There are not that many studies that focus on letter interventions. This thesis uses a real experiment with the entire population entering in the experiment. Most of the ex-periments about this subject are done with students or taxpayers and in that way this thesis distinguishes itself from prior research. This situation is about entrepreneurs that have to ll in the GO for their business. This is another setting than with students and might contribute to the knowledge, because the rela-tionship is between government and business instead of students who want to get their degree. The letter that was send to the non-response group adds to the literature about social norms. Adding a social norm did have an eect in this experiment and this adds to the research that is already done concerning social norms.
2
|
What is procrastination?
In a perfect world people can prioritize the most important tasks and make the right deci-sions and nish the tasks at the right moment. Unfortunately, the reality is dierent. Some-times people delay tasks is such a way that they are not able to complete the task in an accept-able way. This behaviour is called procrastina-tion. What is this behaviour and why do people postpone their tasks?
2.1 What is procrastination?
Simpson and Pychyl (2009) de ne procrastina-tion as "the voluntary and irraprocrastina-tional postpone-ment of an intended course of action despite the knowledge that this delay will come at a cost or have negative eects on the individual". The delaying of the beginning or completion of a task or decision may be inappropriate and self-defeating. This is called dysfunctional procrasti-nation (Ferrari & Emmons, 1995). Various fac-tors may have an in uence on the behaviour of procrastination, namely aspects of the task to be performed, individual characteristics traits of the person to perform the task, situational factors and the way people make decisions in general. These factors will be described in the following sections.
2.2 Task properties of
pro-crastination
Certain tasks are more likely for people to post-pone than other tasks. Which properties of a task in uence the act of procrastination? Attractiveness of the task
An aspect that can in uence the chance of delaying is the attractiveness of the task. The more people like a certain task, the less the task is postponed (Steel, 2007). Several studies have
showed this. Solomon and Rothblum (1984) found in their investigation that disliking a task, leads to greater procrastination. Their subjects were students. Students who considered a task to be unpleasant, boring, or uninteresting are much more likely to delay starting a task and/or completing it; this is true even among students whose academic successes are not negatively impacted by their delaying behaviour. Like Solomon and Rothblum other studies also used the Procrastination Assessment Scale in their investigations. In these studies the students are asked, among other things, which of the 26 reasons for procrastination the best explains the postponing of the writing of a paper. These studies showed that the not liking of the task was indicated by 45% of the participants as a reason for postponing writing the paper (Kachgal, Hansen, & Nutter, 2001) (Peterson, 1987). Other studies support these results and showed that procrastination is caused by tasks being boring and uninteresting (Anderson, 2001)(Briody, 1980) (Froehlich, 1987).
These studies are mostly in the context of studying, but procrastination due to the des-attractiveness of the task can be found in other settings next to studying. Think of the period right before Christmas. The stores are over crowed in the last week before Christmas. This is due to the fact that a lot of people dislike shopping and postpone this task right to the last moment (Ferrari J. R., 1993).
The Dutch tax authorities (2016) have to deal with procrastination and they de ne the follow-ing task properties that stimulate procrastina-tion. These properties are:
1. Boring tasks: In this case there is low in-trinsic motivation to complete a task. 2. Irrelevant tasks: Tasks that do not feel
rel-evant feel meaningless. When people feel something is meaningless, the intrinsic mo-tivation to complete a task is really low.
3. Tasks without feedback: Positive feedback is a great stimulant for people to perform certain behaviour. Feedback gives people approval that they are doing the right thing and that other people are happy when the task is nished. When people know in ad-vance that they will not receive any feed-back, they are less motivated to nish the task. This is closely related to point 2, ir-relevant tasks.
4. Task with low autonomy: When people have few choice electives, the intrinsic moti-vation decreases. A task is less interesting when you cannot adapt the task to your own preferences.
5. Di cult tasks: Task with a high level of di culty can provide an enhanced experi-ence of fear. To avoid this negative feeling of fear, people postpone task which might evoke these emotions.
6. Unclear tasks: Unclear task have the same eect as di cult tasks, it raises feelings of fear and stress.
7. Unstructured tasks:Tasks which are chaotic and unstructured need a high level of intrin-sic motivation. When people do not have this high level of intrinsic motivation, the chance is high that the task will be delayed. The task properties that the Dutch tax au-thorities describes have a lot of overlap with the task properties of attractiveness to the task. When a task does not appeal to people they pro-crastinate more.
2.3 Personality traits of
pro-crastination
Next to task properties, procrastination has to do with certain personality traits. Some even go as far as saying procrastination can be con-sidered as a personality trait itself. Arvey, Ro-tundo, Johnson and McGue (2003) did an inves-tigation about this subject. They asked identi-cal and fraternal male twins reared in the same family to indicate the degree to which they were procrastinators. The interclass correlations of this item were .24 for identical twins and .13 for the fraternal twins, suggesting that approx-imately 22% of the variance of this item was associated with genetic factors.
The following personality traits that are dis-cussed are found to have a strong correlation with procrastination.
Neuroticism
Neuroticism has been explored as a source of procrastination. Neuroticism is much the same as negative aect, anxiety and worrying. Neu-roticism can make tasks increasingly unpleas-ant, because people with this personality trait will indicate a task sooner as less attractive and therefore procrastinate more (Steel, 2007). Impulsiveness
People who are impulsive are more concerned with the here and now than with the future. They are more engaged in temptations in the present and therefore are distracted from other tasks (Blatt & Quinn, 1967). People who pro-crastinate prefer to begin with the most pleasant task at rst. People who do not procrastinate tend to save the most pleasant task to the end (Konig & Kleinmann, 2004). Thinking ahead is not something that procrastinators do a lot. Most of the time they will go for instant grati -cation, whereby they ignore or neglect long-term responsibilities.
Low self-e cacy and low self-esteem Low self-e cacy and low self-esteem are both correlated with procrastination. Especially low self-e cacy, which is when someone has very low con dence in his or her own abilities. It is possible that this goes with the irrational be-lieves that people might have and thereby doubt whether they can accomplish an okay job. A logical consequence is that people with a low self-e cacy often do not even start with certain tasks. Next to that they can believe that they are inadequate as a person when they not com-plete a task in an okay way. Bandura (1997), Burka and Yuen (1983) and Judge and Bono (2001) prove the direct link between procrasti-nation and low self-e cacy and low self-esteem. Conscientiousness
Conscientiousness is one of the big 5 personal-ity traits and is related to distraction, degree of organization, performance orientation and self-control. A low conscientiousness is strongly re-lated to procrastination. A low
conscientious-ness can consist of a lack of self-control, being distracted, being organized and being less fo-cused on achievements.
• Distraction: people who are easily dis-tracted will procrastinate more. When somebody cannot focus on only relevant cues this can lead to procrastination. When somebody is able to dierentiate between distracting and relevant cues, procrastina-tion can be prevented.
• Degree of organization: this degree refers to the ordering, structuring and planning of somebody's life. It is an important char-acteristic in self-control that has a big in- uence on reducing procrastination. Or-ganized people are better in the setting of goals (Locke & Latham, 1990), the re ect-ing on a process (Oettre ect-ingen, 1999) or de-veloping habits that prevent bad decisions to happen (Bargh & Barndollar, 1996). • Performance orientation: people with a
higher performance orientation set more di cult goals for themselves and are of-ten more motivated by achievements them-selves (Costa, McCrea, & Dye, 1991). Per-formance orientation has an in uence on procrastination, because the task itself is seen as motivating and therefore less annoy-ing to the people that focus more on per-formance orientation. Procrastinators turn out to have lower performance orientation and for that reason are less intrinsically mo-tivate to nish the task.
• Low self-control: next to the above charac-teristics in turns out that people with a low self-control have more troubles following up on their plans or intentions. They suer from an intention behaviour gap, described in the section about cognitive mechanisms. Age
Looking at chronic procrastinators there is an eect that people procrastinate more when they are younger. When procrastinators get older they seem to develop a better coping mech-anism with handling procrastination. Appar-ently people can learn to avoid procrastination if they practice more with this kind of behaviour (O'Donoghue, 1999).
Perfectionism
Perfectionism is also correlated with procras-tination. Capan (2010) found that students, who exhibit perfectionist character traits, show more procrastination in academic issues. Capan concludes that procrastination and life satisfac-tion of students were predicted by self-oriented perfectionist personality traits. In contrast, it was concluded that other-oriented and socially-prescribed perfectionism traits did not predict academic procrastination and life satisfaction of teacher candidates.
Absorption
Procrastination might be associated with higher levels of absorption. High absorption individu-als have a mental mind-set that is more exper-imental than low absorption individuals, who tend to have an instrumental mind-set. Being absorbed in more pleasurable activities is a way to facilitate mood repair. By focusing on the more pleasurable activities, people escape from negative emotions that come with the negative tasks. Procrastination can work as a mood-regulating function. Sirois (2014) found that the association between procrastination and absorp-tion was partly due to higher levels of anxiety, supporting the proposition that for procrastina-tors becoming absorbed in pleasurable activities may be a way to escape negative moods.
2.4 Cognitive mechanisms
In the sections before, task properties and in-dividual characteristics are discussed. This sec-tion discusses which cognitive mechanisms stim-ulate procrastinating. These cognitive mecha-nisms apply for everybody, but for one person more than for another.
Limited memories
People can simply forget the task they have to perform. It is not very di cult in this era to set reminders for yourself in an agenda or in a telephone, but still people forget certain things. It has turned out that people have limited mem-ories (Marzilli Ericson, 2014).
The intention behaviour gap
People have the tendency to think that an in-tention will lead to the desired behaviour, how-ever Gollwitzer and Sheeran (2006) showed that there is a low correlation between intention and behaviour. This is due to the fact that people do not take into account the conditions in the future, which might hinder people to do some-thing. Forming an intention in one context and performing it in another context appears to be a di cult task for people. For example, people nd they weigh too much standing on the scale in the morning, but during dinner they take an extra portion of food. Or people plan to go run-ning every morrun-ning, but when it rains they do not go any more. These people have made the intention but fail in the execution.
Planning fallacy
People are not very accurate in estimating the time they need to accomplish certain tasks (Kahneman & Tversky, 1979). People are not very accurate, because they fail to take circum-stances into account that might hinder them in ful lling a task. One can think of distractions or changes in the situation that people did not fac-tored for. Next to that people have the tendency to agree faster on requests that are far in the fu-ture. They do not take into account possible cir-cumstances that might be experienced when the time has nally come to perform the task. Tasks in the future are experienced as more abstract. When a task is perceived as more abstract, peo-ple are more likely to focus on the positive out-come of the request. In the abstract future peo-ple focus more on the bigger goals and therefore tend to agree more with the request. When the request comes closer to the present, people are more likely to think about the practical sides of the request. For example how many hours and how much eort the task will cost and this will lead to less rapid commitment to something in the nearby future. This goes jointly with the fact that most people are too optimistic about their potential to nish task correct and on time. Your future self is perfect
The planning fallacy and the intention be-haviour gap are enhanced by the ideas that peo-ple have about their future selves. Peopeo-ple as-sume that the future them is dierent than the
person at this moment. For some reason peo-ple think that in the future they will have more self-control, make the right decisions and have the motivation to do task that the present them does not want to do. This failure to predict future behaviour has to do with aective fore-casting. That means that when people have to make an assessment on how they feel in the fu-ture, they have a wrong impression of the future. People believe that they can do a task faster or smarter later in time, compared to themselves at the moment. They expect they will change in too something better, which is hardly ever the case (Wilson & Gilbert, 2003).
Your present self loves instant grati ca-tion
The human brain is focused on short term sat-isfaction. Every chance your brain gets to post pone something less pleasurable, it will try to take. People tend to overvalue pleasure in the present and underestimate the cost to do some-thing later instead of now. This pervasive ten-dency is known as temporal discounting (Joshi & Fast, 2013).
Rationalization
When there is a deadline for making the de-cision but no opportunity to shift responsibil-ity, the decision maker may develop rationali-sations that argue against the available infor-mation. This rationalization spreads the al-ternatives and one alternative is perceived as a more favourable solution in comparison with other options. Rationalisation can blind the de-cision maker to their own actions. Rationalisa-tion can take many forms, people can blame ex-ternal pressures, such as family or a superior for forcing one to make a decision. Or people can minimize self-doubt by denying consequences of a decision as bad or unfavourable (Evans, 1990). Inertia
Inertia is a form of resistance whereby someone does not want to move and is passive. Someone with inertia favours the status quo and prefers that nothing needs to change or need to be done. Departing from the status quo takes eort and the person does not look forward to put energy in an action. They are not necessarily against the desired behaviour, but it is mainly due to
the lack of wanting to put eort into some-thing (Pittman, Tykocinski, Sandman-Keiman, & Matthews, 2008).
Timing
Timing also has in uence on procrastination. Loewenstein (1992) traced its roots from a pre-dominantly economic standpoint in terms of temporal discounting. The further away the re-ward or eect is from performing a certain task, the less value is attached to it. The closer the reward or eect is, the greater the value will be that is detached to it. Therefore it might be better to let people perform certain behaviour when the consequences are not too far in the future. O'Donoghue and Rabin (1999) have res-urrected this preference for the present by using the economic discounted model to describe var-ious forms of procrastination. O'Donoghue and Rabin give as an example the tendency of people to save inadequately for retirement.
2.5 Conclusion
There are a lot of factors that can stimulate procrastination. But a lot of these factors can also be resisted when people are made aware of their behaviour. The next section will investi-gate which processes can reduce procrastination.
3
|
Discouraging procrastination
As can be read in the previous chapter there are a lot of factors that can in uence procras-tination. This chapter will review literature which shows which interventions will help duce procrastination and which will not help re-duce procrastination. Proportionally most liter-ature about procrastination is about taxpayers and tax complying. That is why relatively most examples in this chapter have to do with taxes.
3.1 Reminders
People can easily set up reminders for them-selves. There are task lists, reminders can be set up in a digital agenda, personal agenda's etc. But even though there are a lot of sys-tems that people can use to set reminders for themselves, they still fail to do their task in time on occasion, because people simply forget about the task. Therefore procrastination can be reduced by reminding people of their desired behaviour externally. By designing reminders, the limited memory of people must be taken into account, but also how memory interacts with other psychological phenomena that produce behaviour. These mechanisms are explained in section 2.4. about cognitive mechanisms, like planning fallacy and the intention behaviour gap.
The provision of information that a reminder is coming to ful l a task can work in a positive way, but also in a negative way. For calibrated individuals that are time-consistent, it is always optimal to tell them in advance a reminder is coming. The strategies of these individuals will adopt appropriately. The task will be brought back to mind and enable optimal delay if early cost draws are high. However when an individ-ual is present-biased, anticipated reminders may actually lower their welfare by enabling procras-tination. Unanticipated reminders will always provide higher welfare. It will increase the
mem-ory when the reminder is received and it will do nothing to the perceived memory when they get the actual information before the reminder (Marzilli Ericson, 2014).
3.2 Self-imposed
dead-line/commitment
People have dierent sets of preferences at dierent points in time. Think of people that want to start a diet or want to stop smoking. In the morning they want to eat healthy and quit smoking, but later on the day they prefer a piece of cake and smoke a cigarette over their wish to lose weight and live healthy. This can be described as hyperbolic time discounting. Immediately available rewards have a dispro-portionate eect on preferences relative to more delayed rewards, causing a time-inconsistent taste for immediate grati cation. Self-control problems range from smoking, unsafe sex, lack of exercise, drug abuse, procrastination etc.
Well in advance of a task the bene ts seem larger than the costs. In these cases people are willing to take on such a task because the costs seem small compared to the bene ts. But when the deadline comes closer people become increasingly aware of the time needed to complete a task, while the bene ts become increasingly less clear.
Problems with time inconsistent behaviour can be overcome by setting self-imposed deadlines. Examples of this are; saving in non-interest bearing Christmas clubs, buying small packages of cigarettes and avoiding restaurants with a dessert menu. Putting money in a non-interest bearing Christmas fund is not the most pro table way too stall your money, but it helps people save for Christmas presents instead of spending the money on something else. With
buying a small package of cigarettes you set a maximum for yourself with the amount of cigarettes you can smoke.
Ariely and Wertenbroch (2001) found in their study that people are willing to self-impose deadlines to overcome procrastination, even though this comes at a cost. When people are aware of their procrastination behaviour, self-imposed deadlines can be strategic and reasonable.
In the study from Buehler, Peetz and Grif- n (2010) participants where giving instructions for completing an assignment. This assignment involved going to a computer lab and work-ing trough a tutorial program. The tutorial took less than 1 hour and only required the participants to read the text and respond to queries. The participants where informed that they could nish the task at any time in the upcoming 3 weeks. The participants where also asked to predict when they would nish the tu-torial. They got an initial starting point, a card draw between 1 and 21, and they had to adjust their predictions away from this initial starting point. The initial starting point was manipu-lated by Buehler, Peetz and Gri n. One group received as starting point number 4 (early an-chor condition) and the other group received as starting point number 17 (late anchor con-dition).
The result showed that participants who re-ceived the early anchor condition predicted they would nish the task earlier than the late anchor condition. The earlier anchor predicted that they would nish the program 16.84 days before the deadline and the late anchor predicted that they would nish 11.8 days before the deadline1. The participants in the early anchor condition also nished the task earlier compared to the late anchor condition. Respectively 14.26 be-fore the deadline compared to 9 days bebe-fore the deadline2. Giving the participants an early an-chor did work to stimulate the participants to complete the task earlier in the submission pe-riod.
1(M = 16.84 days, SD = 3.29)&(M = 11.80 days, SD
= 4.06), t(32) = 4.01, p < .001
2(M = 14.26 days, SD = 6.14)(M = 9.00 days, SD =
8.01), t(32) = 2.17, p = .04
3.3 Reciprocity
When preferences depend on the fair or unfair behaviour of other agents, it is called reciprocity. An individual who has these kinds of preferences responds to actions he perceives as kind, in a kind manner and actions he perceives as unkind, in an unkind manner (Fehr & Schmidt, 2005). Putting a post-it on a letter is perceived as a solicitation for a personal favour, facilitating a normative compliance response (Garner, 2005). Post-it persuasion
A possible intervention is therefore Post-it persuasion. A post-it can be attached to an invitation letter that requests the target audience to perform a task. In the article by Randy Garner (2005) the in uence of a post-it note is described. He examines with 4 studies what the in uence of a sticky post-it note is. The rst study shows that the participants, who received a survey with an a xed post-it, returned the surveys signi cantly more often than the group that received a survey without a post-it attached to it. 38 of the 50 participants in the Post-it group (76%) returned the survey, as compared to 24 of the 50 participants (48%) in the written-only group and 18 of the 50 participants (36%) in the no-message control group, χ2 = 16.93, p <.01. The written only group received the message on the post-it in the letter instead of on a post-it. The control letter did not receive any extra information. The participants that received the message in the survey returned the survey with the same rate as the non-post-it group. This shows that the post-it note itself was essential.
The second study also compared 3 groups. The groups exited of a group that received a Post-it message, blank post-it and no post-it (control group). After returning the surveys there was no dierence found between the blank post-it, (43%), χ2(1, N= 15) =4.69, p < .05 and non-post-it group, (34%), χ2(1, N = 12) = 8.23, p < .O1. The group that received the letter with the message on the post-it, had a signi cantly higher return rate, (n = 24, 69%). This shows that a message on the post-it is necessary. From the rst 2 studies can be concluded that a post-it and message both contribute to the return rate.
The third study almost did the same com-pared to the second study, with an addition of a follow-up questionnaire as well as an assessment on survey response quality and promptness of return. The participants in the post-it conditions provided signi cantly more comments in the open-ended section of the modi ed survey. 18 of the 32 individuals in the Post-it condition(56%), oered additional information in these sections, whereof 9 of the 21 individuals in the control condition(43%) provided additional comments. From this can be concluded that not only people who receive a survey with a post-it respond more often, they also provided signi cantly more information in the open ended sections compared to the control groups.
The fourth study used a short and a long sur-vey, relatively 5 and 24 pages. Half of the group that received the short or long version received a post-it with the standard message used in the previous studies and the other half received a post-it with a more personalized message that included the name of the participant, as well as a concluding "Thank you!" and the initials "AB".
For the group that received the short version, the personalized post-it had no in uence. How-ever, When the task was more time consuming, the personalization of the Post-it message became an important factor, see table 3.1 (Garner, 2005).
From this study it can be concluded that post-it notes can have a signi cant eect on the re-turn rate. However, there has to be a message on it, because otherwise it does not work. At-taching a post-it to a letter ensures that peo-ple will return information more likely and are more open to cooperate and therefore procrasti-nate less. The intrinsic motivation to cooperate is raised by the reciprocity people feel for the adding of the post-it.
Small gift
Falk and Fischbacher (2001) use the same terminology as Fehr and Schmidt. A reciprocal action is modelled as the behavioural response to an action that is perceived as either kind or unkind. The consequences and the underlying intentions are essential. According to the inequity aversion approach, people will only
exhibit reciprocal behaviour, when it reduces inequality. However Falk and Fischbacher found that people exhibit this kind of behaviour when it is driven by rewarding or punishing, regardless of the inequity between the persons. Falk and Fischbacher state that reciprocal behaviour is primarily driven as a response to kindness, not a desire to reduce inequity.
Giving a small gift can therefore stimulate to co-operate more with the organization who pro-vided the gift. People are more likely to respond on the request to ll in certain information when they receive a small gift. A small gift can there-fore stimulate to reduce procrastination. This is shown in the study of Cialdini and Goldstein (2002). In their study they looked at the or-ganization of disabled American Veterans. The organization of disabled American veterans en-closed a small gift in the envelope that seeks donations. The average response rate rose from 18% to 35%, only by adding personalised ad-dress labels.
Rewarding
Incentives can come in various forms. If you want to stimulate people to perform certain behaviour you can use positive or negative incentives. Negative incentives are for example threats, punishments and incapacitations. Positive incentives are in this case actions by government authorities that are directed to-wards speci c individuals and that are intended to increase their compliance with law. They are intended to reinforce or increase something that the target individual nds pleasant. Positive incentives should not only focus on material rewards, they may also increase or reinforce characteristics as satisfaction, pride, rectitude, recognition from others, feelings of consistency, fairness etc. Actions on an individual level are in general more eective.
Smith and Stalans (1991) make a distinction between two general models by which positive incentives created by authorities' actions may increase tax compliance. Method 1 focuses its attention on the direct link between compliant behaviour and the receiving of the positive in-centive. Method 2 makes the link less explicit and relies upon what has been called peripheral processing.
Table 3.1: Survey Return Rates as a Function of Survey Length and Request
Survey type: Standard
post-it personalisedpost-it Control
Long version 12(40%) 20(67%) 4(13%)
Short or regular version 21(70%) 23(77%) 10(33
Method 1: Directly linking compliance and the incentive
This method emphasises on the possibility of a reward. In the case of tax compliance you can think of rewards that if you submit your taxes earlier, you get paid out earlier. Or if you sub-mit the taxes for a certain date, the participant enters in a lottery with the possibility to win some prices. This is called operant conditioning. The theory of operant conditioning states that a desired behaviour will increase over time if the occurrence of the desired behaviour is paired with reinforcement, that is, an action that the person nds pleasant. Therefore reinforcing the behaviour to come in action at the desired mo-ment can be reached by rewarding the actions of the collaborators when they do not procras-tinate.
This has the possibility to reduce procras-tination because you reward people with an earlier pay-out or the chance to win a price. This has the potential to work for multiple years, because they get rewarded for their early submission and therefore are more stimulated to perform this behaviour the next year.
However this might not be always the best solution because positive reinforcement often comes at a cost, which might be higher than the costs that are associated with procrastination. Method 2: Peripherally processed incen-tives
Some people justify their behaviour of not lling in their taxes because of negative perceptions they have about the tax system. There is evidence that positive feelings and attitudes lead to decisions that are consistent with those feelings. Improving the promptness and respon-siveness of correspondence, resolving taxpayers' problems e ciently and reasonably, providing more assistance and clearer instructions and forms, and complimenting or praising taxpayers for their eorts, are all ways of changing
citizens' perceptions about the tax system. This is called peripheral processing. Persuasion through the peripheral route has two major advantages. Firstly, indirect incentives are not linked to speci c acts of compliance; their implementation does not require authorities to monitor individuals' taxpaying behaviour or to distinguish between compliance and non-compliance (Smith & Stalans, 1991). Secondly, this approach is more discrete with regard to the cost/bene t component of taxpayers' decisions about complying with tax laws and consequently is more likely to invoke normative or aective processing (Carroll, 1987).
Linking the indirect incentive to expected behaviour may involve several dierent types of cognitive and aective processing. Respectful treatment or praise by authorities may increase taxpayers' liking for and satisfaction with the tax authorities, while also inducing compli-ance through classical conditioning. Through further cognitive processing, this increased positive eect may result in an acceptance and internalization of the authorities' message to comply with the law, because the satisfaction with the authorities creates dissonance with tax non-compliance and attitudes that condone tax cheating (Smith & Stalans, 1991).
The section that describes reasons for pro-crastination de nes one of the task properties as unattractiveness of the task. By changing the attitudes toward the task and the organiza-tion the task might become more attractive and therefore people might procrastinate less on the task.
3.4 Social norms
Cialdini and Trost (1998) characterize social norms as 'rules and standards that are under-stood by members of a group, and that guide and/or constrain social behaviour without the
force of laws'. The 4 dierent social norm con-structs identi ed by Cialdini and Trost are in-junctive norms, descriptive norms, subjective norms, and personal norms.
• Injunctive norms: The perception of what most people think other people should do in a given situation. They specify what should be done and which the moral rules of the group are. The social goals of injunctive norms are to build and maintain social re-lationships.
• Descriptive norms: What one perceives that other people do in a given situation. Watching others provide information about what is "normal" in a novel or ambiguous situation. The social goal of descriptive norms is eective action, the desire to be accurate in one's choices and behaviours. • Subjective norms: A person's perception
about what those who are important to him think he should do in a given situation. They are one's perceptions of the injunc-tive norms held by the people whose opin-ions matter most to him. The social goal of subjective norms is to build and maintain social relationships.
• Personal norms: Self-based standards or expectations for behaviour (i.e., what an individual believes he/she should do) in a given situation. These standards arise from internalized values. The social goal of per-sonal norms is to manage self-concept. If these norms are set in such a way that people feel that procrastination is contrary to their beliefs they will procrastination to a lesser extent.
Bobek et al (2012) found that not all the four social norm constructs are signi cantly correlated with tax compliance behaviour, only subjective norms and personal norms directly aected taxpayers' compliance intentions. Descriptive and injunctive norms were related to subjective and/or personal norms; however, they did not have a direct eect on tax com-pliance intentions. They found that personal norms played the biggest role in compliance than any of the other three. This is concluded, because tax returns and compliance decisions are not shared with the public. Injunctive norms demonstrated a strong in uence on both
personal and subjective norms and, descriptive norms shaped both injunctive and subjective norms, so they do have an in uence. (Bobek, Hageman, & Kelliher, 2012).
Reno et al. (1993) performed research on the eects of descriptive norms. They used a sign to make people aware of the fact that taking the stairs is better than taking the elevator. The sign pointed out that most of the people took the stairs. Reno et al. found it only worked when the text that says 'most of the people take the stairs' was on the sign. Without this information, they found no change in the elevator usage. The descriptive norm information was necessary to produce the required eect. Drawing attention to readily available norm information can be an eective strategy for changing behaviour.
This information can be linked to procrasti-nation. By linking a social norm to the desired behaviour it is necessary to include a descrip-tive norm. By doing this people will feel more urge to cooperate in the desired behaviour and people do not want to deviate from the descrip-tive norm. They see this as some sort of anchor to check the appropriateness of their own be-haviour. But using such a norm can lead to certain complications. Descriptive norms can decrease undesired behaviour among individu-als who perform that behaviour at a rate above the norm; the same message may actually serve to increase the undesirable behaviour among in-dividuals who perform that behaviour at a rate below the norm. They performed an experiment and found out that this so called 'boomerang' eect can be prevented by adding an injunc-tive message. By giving approval of their cur-rent behaviour you lose the boomerang eect (Schultz, Nolan, Cialdini, Goldstein, & Griske-vicius, 2007).
Smith and Stalans (1991) found something simular. When you provide people with the norm that most people ll in their tax at the last moment this might result in people drawing the conclusion that it is okay to ll the tax in late, because everybody does this, this is the so called 'chump eect'.
Establishing the feeling of similarity can also aect the behaviour of people. Cialdini and Goldstein (2002) show in their study that the ef-fect can be quite astonishing. One study added
the phrase 'I am a student, too' to the request to contribute money and the contributions dou-bled relative to the request without the line with 'I am a student, too' in it.
3.5 Framing
The early phase of decision making is called framing. It helps the decision maker simplify situations by paying more attention to certain factors and ignoring others. In this early face of decision making, the decision maker has to break the situation up into alternatives, outcomes and references points that form the inputs to the decision process (Tversky & Kah-neman, 1981). It makes a big dierence how tax payers frame their tax behaviour. They can frame their behaviour as compliance but also as avoidance. They will create dierent behaviours when they frame their behaviour dierently. It is a fundamental issue how a person frames taxpaying, involving motivation and moral reasoning. Dierent types of attitudinal types are found among taxpayers; honest taxpayers or utility maximization taxpayers. It is interesting to see how you can change the perception of taxpayers into a more positive view on taxes, so that they have more positive feelings towards taxes and are more likely to cooperate (Carroll, 1987). This can reduce procrastination when the framing is positive towards an organi-zation. An organization can try to display the organisation as more positive in the hope that their clients comply more with their wishes. Tversky and Kahneman (1981) showed with the Asian disease problem that framing can have an in uence on the choices of people. In this study the participants were asked to imagine the outbreak of an unusual Asian Disease, which is expected to kill 600 people. Two alternative programs for the disease were proposed. With the positive framing, the people were told that if plan A is adopted, 200 people will be saved. If plan B is adopted there is only one third probability that all people will survive and two third probability than no one will survive. In this treatment 72% was risk averse, so they chose option A. Interestingly is that when framed negatively the participants became risk loving. In the treatment with negative framing the participants were told that if option A is adopted 400 people will die
and 200 survive and when option B is adopted, there is a one third probability that none of them will die and two thirds probability that all 600 people will die. In this case, 78% chose option B and was risk loving.This shows that the way of framing can lead to very dierent results and can in uence the procrastination behaviour of people.
Tu and Soman (2014) did an experiment with time framing. One group got a deadline in December and the other group a deadline in January. Both the groups had the same amount of time. When the deadline was in the current year instead of in the next year there was greater task initiation, although the available time for completing the task was the same across the two conditions. Also more people n-ished the task in the December treatment. The December group had a return rate of 31.86%; (51/161), compared to the January group who had a return rate of 8.09% (11/136; χ2(1) = 24.84, p < .001) Tu and Soman concluded from this, that people are more likely to initiate a task when the task deadline is classi ed in a like-the-present category than in an unlike-the-present category. They got the same result with a test with performing a 4-hour data entry job. The like-the-present category got a start date in April and the subjects in the unlike-the-present category got a start date in April and an end date in May. All the participants had 4 days to complete the task. Again, they found a decrease in willingness to initiate the task when the deadline just moved from a like-the-present category to an unlike-the-present category. An independent t-test showed that people in the like-the-present condition were more likely to start working now (M=3.72, SD=1.75) than those in the unlike-the-present condition (M=6.10, SD=1.57; t(98)=7.16, p <.001).
The Dutch tax authorities (2016) tested some interventions with framing. They gave the same date to two groups, but they framed the submis-sion date dierently. One group got the message that they had to hand in the information at lat-est by 30 September and the other group got the message they had to hand in the informa-tion prior to the rst of October. These 2 dates are equal to each other; they both have till the 30th of September, 23:59. The Dutch tax au-thorities expected that the group that received the date with the sentence at latest by the 30th
of September would submit the information ear-lier than the group that received the date in a sentence prior to the 1st of October. This is be-cause September feels more closely, a like-the-present setting, because the letter was send in September. They expected that October would feel more distant, an unlike-the-present setting. The Dutch tax authorities expected that the subjects in the unlike-the-present setting would be more likely to procrastinate. There was no eect between the 2 groups, the framing did not show any eect. Comparing the results of Tu and Soman and the Dutch tax authorities gives dierent results. For the Dutch tax authorities framing did not help, but Tu and Soman did nd an eect.
4
|
Casus
Each year RVO (Netherlands Enterprise Agency) has to collect the Gecombineerde op-gave (GO) from agricultural entrepreneurs, in other words: collect information from farmers in the Netherlands. With the GO farmers deliver information about agricultural census, registration of emissions, manure legislation and the farmers apply for (EU) subsidies. The entered information is used, among other European and national statistics, for policy development and at the outbreak of animal and plant diseases. If the farmers are entitled to subsidies, they will get these subsidies based on the information they provide in the GO.
A part of lling in the GO is the subscribing of parcels. In this part of the GO the farmers have to specify which patches of land are theirs and what they do on these patches. They can grow crops, hold animals or leave it fallow. The subscribing of parcels has to be done very speci c and the size of the parcels has to be entered on third decimal accuracy. Most of the subsidies they receive are based on the surface and utility of these parcels. The farmers get 6 weeks to complete this task; however the updating of changes in their parcels can be done throughout the whole year. The 6-week period starts on the 1st of April and ends on the 15th of May. The RVO cannot change the 6-week submission period. This period is de ned by the European Union and every member state must adhere to these 6 weeks. Each member state has some space with the starting and ending of the 6 week period; it has to be between February and May, but it cannot be longer than 6 weeks.
Before the 6-week period starts, all the farmers who are considered by the RVO to do the GO get an invitation letter which says they are required to submit the GO. This letter informs them about the period in which they
Figure 4.1: Submission GO per day 2014 can submit their GO. They will also receive a reminder letter around the rst of May when they did not submit anything at that date. The farmers are not noti ed of this reminder upfront, but most of the farmer already know about the reminder from earlier years.
The problem that RVO has is the fact that most of the farmers submit the GO, in the last 2 weeks of the period. 60% of the submissions are done in the last 2 weeks of the period, of which 40% of the farmers submitted their GO in the nal week (for the year 2014). Figure 4.1 shows the submissions per day in 2014. This gure shows clearly that the end of the period is by far the busiest.
The overload in the last 2 weeks is also a problem for the farmers. The farmers suer from a less available site in the last 2 weeks and they have to wait a long time before they can reach the call center and if they get a call center employee on the phone the chance is high that it is not a specialist. This is because RVO has to employ temporary workers in order to cope with the pressure on the call center in the last 2 weeks. The site does not work as good as in the prior week, because too many people log in at the same time. By postponing the GO
Figure 4.2: Submissions per hour rst week 2014
Figure 4.3: Submissions per hour last week 2014 to the last week it will cost the farmers more time and more stress to ll in the GO. If the farmers spread their submissions more equally, the submission will cost them less time, less stress and they won't have a nagging feeling in their mind when they already nished the GO.
The dierence between the amount of submis-sions of the GO in the rst week and in the last week is enormous. Figure 4.2 shows the amount of submissions on hourly level in the rst week of the submission period in 2014. The busiest hours in the rst week are around 50 submis-sions per hour.
Figure 4.3 shows the submission of GO per hour for the last week of 2014. Almost every hour the submissions are more or close to 100 submissions and the busiest hours are up to 600 submission. The dierence between the rst and last week is huge.
The gures 4.2 and 4.3 indicate how big the dierences are between the weeks. The call center and the ICT systems suer from the overload in the last 2 weeks. This causes the service level to go down. The amount of people that can be deployed in the call center has reached it limits, especially because the quality of the service cannot be ensured, when people are attracted from outside to absorb the work
pressure.
RVO contacted the Behavioural Insight Team (BIT EZ) from the ministry of Economic aairs to help in the search for a solution for this problem. The problem that is being addressed is; trying to reduce the behaviour of procrasti-nation with farmers so that they submit their GO earlier. It is mostly focused on trying to reduce the busyness in the last 2 weeks of the period, but also creating more dispersion along the 6-week period. With more dispersion among the farmers, the call center can provide bet-ter service and the site will be less over crowded. To change the behaviour of the farmers to submit the GO earlier in the submission period in contrast to submitting the GO in the last period, procrastination must be prevented. Looking at the communication that already exist between the farmers and RVO, we noticed that the farmers receive an invitation letter from RVO each year to notify the farmers about the GO. The group consist of more than 70.000 farmers/businesses and keeping the short time period in mind, this invitation letter was a logical choice to implement the interventions in.
To get a clearer image of the target audience a eld investigation was done. The eld investi-gation together with the desk research will give more insights about the target audience, the task they have to perform and the behaviour which is involved in the process. This eld in-vestigation will be brie y described in the fol-lowing section.
4.1 Field investigation
The goal of this eld investigation is to gather information to create interventions that can be placed in letters to persuade farmers to ll in their GO earlier on, instead of waiting until the end of the period. A panel survey was held with 10 farmers: 9 dairy farmers and 1 arable farmer. In addition, 3 accountancy bureaus were visited and 3 farmers at their farms. 1 of the farmers is part-time farmer and part-time accountant, so he possessed knowledge about both the agri-cultural and the bureaucratic aspect of farm-ing. These interviews provided insights about the GO and the opinion of the farmers that have
to ll in the combined statements. Interview accountancy bureaus
The accountancy bureaus need the whole 6 weeks to schedule all their clients. Most of the time they do not start the rst day, because they need 1 or 2 days to go through the program to see the changes with respect to prior years. After that, they start making appointments with the farmers. They start with the easier submissions, because sometimes the program has some bugs in the rst weeks.
Some farmers tell the accountancy bureaus they will call them back to make an appoint-ment, instead of making the appointment when the accountancy bureau calls. Most of the time they call for an appointment when the farmers receive the reminder letter or when it is quieter on their farms. This gives the impression that not much improvement can be made through the accountancy bureaus. An appointment takes approximately 1.5 hours.
The bureaus mentioned that the subscription of the parcels takes the longest. It takes a great deal of precision to get all the parcels in the pro-gram and RVO does not always use the same aerial photographs as the farmers. It can take a long time to get it right. This might be one of the causes of procrastination for farmers that ll in the GO themselves. Another point that the bureaus mention is that the farmers do not understand why some questions are asked and what happens with the information the farmers have to give up. The farmers have distrust and suspicions against the government. The farmers also got rewarded in 2015 with waiting to the end of the period to submit the GO. A question was removed at the end of the period, because it was not clear where the information was needed for. By delaying lling in the GO, the informa-tion about what needs to be submitted became clearer and the farmers that waited did not have to give up information they did not want to re-lease. The farmers that already lled in their GO, did submit this information.
Interview panel
From the panel discussion has been learned that a lot of the farmers see the GO as a di cult task, which becomes harder every
year. Some farmers lose the sense of autonomy on the GO and need somebody to check the GO on errors or ask for advice. Lack of information and job autonomy can be reasons for procrastination. The farmers from the panel also pointed out that the quality of the information that the call center provides is insu cient, especially at the beginning of the period. Sometimes the farmers hang up the phone in the hope they will reach another call center employee that has more knowledge of the subject when they call back. The panel also says that the program becomes slower at the end of the period and when it is rainy outside. When the weather is bad, the farmers cannot work on the land and therefore the program gets used more. The more people make use of the program at the same time, the slower it gets. One of the biggest problems that comes for-ward interviewing the farmers in the panel, is the lack of information and transparency. The time needed to submit the GO ranges from 1,5 hours to over 10 hours. It is not always clear why certain information is asked and the farm-ers fear the consequences. If the farmfarm-ers make a mistake this can lead to big discounts on their subsidies and that is why they become suspi-cious. Similar to the accountancy bureaus the panel says that the parcel submission takes the most time, this takes a lot from the computer and it has to be done very precise. They often do not understand why a parcel changes in the eyes of the RVO from year to year, while noth-ing has changed relative to the previous year in real life/the eyes of the farmers.
Kitchen table interviews
The kitchen table talks led to a dierent conclu-sion about the information that is asked from the farmers. All the farmers that where in-terviewed indicated that they did understand why the government needs certain information. They think it is necessary for a government to know what is happening in the agricultural branch. Maybe a dierent image was sketched in the panel discussion by group con rmation; one farmer said something and everybody went with the notion. One of the farmers that only has one type of crop said it takes him 1,5 hour to ll in the GO. Another farmer said that it takes him more than 10 hours. The farmers did agree that the call center is not su cient and
also the information on the website falls short. They would like to see a better information pro-vision, so that it is easier to get answers on their question. They delay their submission be-cause they are waiting on the right answers from RVO. Before they get this information, or get con rmed that earlier answers that were given are right, they do not want to submit anything. They also think it becomes harder and harder every year. They still submitted the GO them-selves in the past years, but due to increasing numbers of rules and di culties they are slowly losing the feeling of job autonomy.
Conclusion eld investigation
Everyone spoken to in the eld investigation said that if they could get their subsidies paid out earlier, when they submit their GO earlier would be the best method to reduce procrasti-nation. Unfortunately RVO cannot make such a promise. This has to do with certain rules and some subsidies are harder to process than oth-ers. These people that get those 'special' sub-sidies get their money later, even though they complete the GO in the rst week or even the rst day. To compare this with the task and character traits which cause procrastination the following can be concluded:
• The GO is a di cult task. Accountancy bureaus spend approximately 1.5 hour per client. Before the clients come to the bu-reau the farmers need to prepare and gather information. Most of the farmers say they spend more than 3 hours lling in the GO. • The GO becomes harder every year and this leads to a feeling of loss of job autonomy. The farmers have the feeling they are not capable any more of doing it on their own. A lot depends on this and they are afraid to make mistakes.
• They nd it a hard, di cult and boring task. Most of the farmers prefer working on their lands instead of doing administra-tive work.
• They also told the timing is not perfect because it is in the spring, the time they can work on the land in comparison with the winter. But when asked which period would be a better period they could not come up with a better date, because they are always busy.
• The farmers miss information and therefore they nd the task unstructured. When they try to contact RVO with their questions, the farmers do not receive direct answers, call center employees ask more informed ad-visers and then they call back. When the farmers have follow up questions, the pro-cedure has to be repeated and therefore it takes a long time to get the right answers. • The farmers feel they are battling against
RVO, while in reality they should be sup-porting each other.
• Last year they got stimulated in their pro-crastination behaviour. A question was cancelled at the end of the period. A lot of farmers said this might me a reason to wait till the end of the period of this year to see if it happens again.
• The accountancy bureaus nd it a nice task; they see it as an opportunity to catch up with their customers. They face the prob-lem of handling a lot of customers in a short period of time. That is why they would like the submission period to be lengthened.
4.2 Letter interventions
The way the interventions can be done is through the invitation letter the farmers receive every year, containing the call to submit the GO. Other interventions are not possible for this year.
The following section describes previous liter-ature with regard to letter interventions. The behavioural insight team (BIT UK) from the United Kingdom did some research into the writing of successful letters. They made a guide that aims to help write letters that lead to a spe-ci c change in behaviour. Below there is a short summary of their recommendations for writing a letter. They de ne 3 main stages to writing a letter.
4.2.1 Making the core
compo-nents clear
Many letters can be improved signi cantly by making them shorter and clearer. It is impor-tant to consider the situation of the recipient. What are the core pieces of information that
they will need to know to understand the letter and act on it? What are they likely to know already and what do we need to explain? Core components need to be described in a short and simple way. All the information that is not strictly necessary has to be removed. Simplicity can be the most eective strategy. Simplicity can be reached by using short sentences, short paragraphs and 'white space'. The BIT UK set as a rule that a paragraph longer than 4 lines should be examined and if it is possible; make it shorter. It is important to keep in mind that the respondent does not take a long time read-ing the letter. Therefore, the core components need to be upfront in the letter.
Consider how you can use the heading at the top of the letter to summarize the desired action. By describing the core components there should be a clear connection between each part. Make use of linking words like 'but', 'this means' etc. Make the desired behaviour as clear and spe-ci c as possible. The letter should leave the re-cipient with a speci c action that is required. Never leave the respondent in doubt what you would like to happen.
Remove barriers that seem small and not sig-ni cant. Instead of saying please call the num-ber, write 'please call 012 34 56 78'. In this way the recipient does not have to search for the number elsewhere.
Start the letter with the core components but also end with a call to action. People tend to focus on the start and end of letters.
4.2.2 Introducing persuasive
mes-sages
Every element of a letter should be seen as an opportunity to change behaviour. It is also very important to check that the text of a letter is not working against the desired goal.
Checking overall feeling and tone
It is very important to consider if the right tone is used. This means there is a need for a -nal sense check. Letters should be respectful but rm where necessary (Behavioural Insight Team, 2016). Cialdini and Goldstein (2002) ar-gue that in persuading people you always have to be honest with them. They state that when you are dishonest there can be major conse-quences not to your advantage.
