DEPARTMENT OF POLITICAL SCIENCE THE FIGHT AGAINST WELFARE CRIME IN SWEDEN A theoretical mischaracterization of the problem? Johanna Winqvist Master’s Thesis: 30 credits Programme: Master’s Programme in Political Science Date: 2023-05-22 Supervisor: Anna Persson Words: 19 812 Abstract In recent years a series of government agencies have indicated that the Swedish welfare systems are being exposed to an increasingly sophisticated and organized welfare fraud. According to the official analysis, Sweden’s problems are largely due to an inadequate ability to detect and sanction these types of crimes. Even if the government has taken extensive measures to prevent welfare crime and government agencies detect and prosecute more crimes today than previously, Sweden continues to have extensive problems with welfare fraud. According to the many government investigations released in the past decades, the reason for this is either that the measures taken have not been extensive enough or that the crime has adjusted to the new conditions. This thesis aims to investigate a third possible explanation for this puzzling development, namely that the Swedish strategy to counteract welfare fraud is based on a flawed understanding of the problem of welfare fraud and that the measures taken have thus been misguided. By conducting in-depth informant interviews with individuals who have a real insight into welfare fraud and how it manifests in Sweden this thesis finds that welfare fraud is a considerably more complex problem than proposed by the government investigations. While welfare fraud may be due to a lack of control and sanctions this is far from the only reason why Sweden is seeing such vast problems with welfare fraud. If Sweden are to overcome these issues it will take more than just effective control systems and tough sanctions. Keywords: Welfare crime, Sweden, Control, Compliance, Problem representations 2 Contents ABSTRACT ............................................................................................................................................................. 2 CONTENTS ............................................................................................................................................................ 3 1 INTRODUCTION ......................................................................................................................................... 4 2 BACKGROUND ........................................................................................................................................... 7 2.1 CONCEPTS AND DEFINITIONS .................................................................................................................. 7 2.1.1 The Swedish welfare system .............................................................................................................. 7 2.1.2 Welfare Crime and incorrect payments ............................................................................................ 7 2.2 THE EXTENT AND DEVELOPMENT OF WELFARE CRIME IN SWEDEN ........................................................ 8 2.3 THE EXPRESSION OF WELFARE CRIME .................................................................................................. 10 2.4 WHY IS SWEDEN VULNERABLE TO WELFARE FRAUD: ACCORDING TO THE OFFICIAL ANALYSIS? ......... 12 3 PREVIOUS STUDIES AND THEORETICAL FRAMEWORK ........................................................... 15 3.1 PREVIOUS STUDIES OF WELFARE FRAUD .............................................................................................. 15 3.2 THEORIES OF COMPLIANCE AND NONCOMPLIANCE ............................................................................. 16 4 AIM AND RESEARCH QUESTIONS ..................................................................................................... 21 5 RESEARCH DESIGN AND METHODS ................................................................................................. 22 5.1 METHODOLOGICAL APPROACH ............................................................................................................ 22 5.2 SELECTION OF INFORMANTS ................................................................................................................. 23 5.3 DATA ANALYSIS AND ANALYTICAL FRAMEWORK ................................................................................ 25 6 RESULTS ..................................................................................................................................................... 28 6.1 PROBLEM - WHAT TYPE OF PROBLEM DO THE INFORMANTS PERCEIVE WELFARE FRAUD TO BE? .......... 28 6.2 CAUSE - WHY DOES SWEDEN HAVE PROBLEMS WITH WELFARE FRAUD? .............................................. 34 6.3 SOLUTION ............................................................................................................................................. 41 7 SUMMARY AND CONCLUDING REMARKS ...................................................................................... 47 8 REFERENCES ............................................................................................................................................ 51 9 APPENDIX .................................................................................................................................................. 60 3 1 Introduction Sweden has long been known for being a well-functioning country and has consistently received top rankings in comparative indices on everything from state capacity, control of corruption, and government effectiveness to social trust (Holmberg & Rothstein, 2020; Transparency International, 2022; World Bank, 2021). In recent years a series of government agencies have released reports indicating severe malpractices within the Swedish public administration (see e.g., Brå, 2015; Försäkringskassan, 2021; Polismyndigheten et al., 2021; SKR, 2021; SOU 2019:59). According to the reports the Swedish welfare systems are exposed to an increasingly sophisticated and organized welfare fraud which not only cost the government large sums of money but also risk undermining trust in the entire welfare state (Korsell, Hagstedt & Skinnari, 2008, p.34; SKR, 2021, p.9). In the past few years alone, the media has reported on several major welfare fraud scams (Carp, 2022; Wikén, 2020; Klingberg & Hjort, 2021; Asmar & Rundberg, 2016; Karlsson, 2023). In a recent government investigation (SOU 2019:59, p.94), the losses due to incorrect payments from the welfare systems were estimated to be between SEK 10-27 billion a year. Even if far from all incorrect payments constitute fraud around 50 percent of the losses were estimated to be caused by welfare crime (SOU 2019:59, p.98). In the political debate, the problems with welfare crime are linked to organized crime. According to the Swedish police authority criminal organizations usurp large sums of money from the welfare system, they even go as far as saying that organized crime is using the Swedish welfare system as a personal ATM (Polismyndigheten et al., 2021, pp.10-11). According to a statement made by the Swedish Minister of Justice at a press conference earlier this year, welfare fraud and other types of fraud have started to challenge drugs as the main source of income for organized crime (Regeringskasnliet, 2023, 15:57-17:15). Welfare crime is not a new phenomenon in Sweden, in fact, welfare crime has more or less been on the political agenda since the mid-1990s (Korsell, Hagstedt & Skinnari, 2008; Lundström, 2011; SOU 2008:74, p.29-30). Over the years the government has ordered numerous investigations with the specific task to identify the causes of these issues and present suitable policy recommendations (e.g., Dir 2005:52; Dir. 2016:60; Dir 2021:39). The investigations provide a unanimous and clear analysis of Sweden problems. The Swedish welfare systems are based on information. To receive a social benefit the applicant must 4 provide information regarding for example residency, income, or medical condition. Information is easy to manipulate, to prevent crime it is therefore crucial that the agencies can control that the information provided is correct. The Swedish welfare systems were never built to detect or prevent fraud (FUT-delegationen, 2007, p.12). They were built on trust and on the assumption that people generally want to do the right thing and thus provide correct information (Ekobrottsmyndigheten, 2016; Polismyndigheten, et al., 2021; SOU 2020:35, p.153). As a result, many of the issuing agencies lack adequate routines and tools for control and thus detect few crimes (FUT-delegationen, 2007, p.33; SOU 2018:14, p.92; SOU 2020:35, p.131). Even when controls are conducted and crimes are detected few crimes lead to a conviction (SOU 2022:37, p.332-333). In 2021 the Social Insurance Agency, for example, conducted 30 600 control investigations of suspected welfare fraud. Only 5 900 resulted in a report to the police and out of those barely 50 percent led to a conviction (SOU 2022:37, p.23). Based on the official analysis the solution to Sweden’s problems appears to be quite straightforward; according to the government investigations, what is needed to prevent welfare crime is more extensive controls, more efficient investigations and tougher or at least more effective sanctions (See e.g., SOU 2008:74: SOU 2017:37; Sou 2020:35; SOU 2022:37). The idea underlying these proposals appears to be logical. If the risk of getting caught as well as the cost of committing these crimes are increased, then the incentives to abuse the systems will decrease. If the incentives decrease so will the number of people abusing the systems. Unfortunately, this theory seems to have worked better in theory than in practice. Since the mid-1990s, many of the proposals given by the government investigations have been implemented (SOU 2022:37, p.95f). Even if many of the efforts seem to have been effective (SOU 2020:35, p.117) and the government agencies detect more welfare frauds today than previously (SOU 2022:37, p.341), the penalties for welfare fraud have been raised (prop. 2018/19:132) and efforts to prevent fraud have gotten a higher priority within practically all issuing authorities (SOU 2022:37, pp.22-23), welfare fraud remains at similar levels as before (SOU 2019:59, p.29; SOU 2020:35, p117). How can it be that despite all these measures Sweden is still experiencing such extensive problems with welfare fraud? The government investigations provide two answers to this question; either the measures taken have not been extensive enough or the crime has adjusted to the new conditions and found ways to curb the governments control efforts (SOU 2020:35, p.117; SOU 2017:37, 278; 5 SOU 2022:37, p.114f). This thesis aims to investigate a third possible explanation for this puzzling development, namely that the Swedish strategy to counteract welfare fraud is based on a flawed understanding of welfare fraud in the first place. Welfare fraud is far from the only problem against which extensive measures have been taken without achieving any noteworthy results. Efforts to curb corruption have long been plagued by the same problem (Persson, Rothstein & Teorell, 2013). Withing corruption research there is growing consensus that the reason why anti-corruption reform fail is not because the measures taken have not been extensive enough but rather that they have been based on a flawed understanding of the problem. Corruption has traditionally been seen as a so-called principal-agent problem, that is, a problem of individual incentives. What this group of scholars argues is that corruption is not primarily an incentive-based problem but rather a problem of the system, a so-called collective action problem (Persson, Rothstein & Teorell, 2013, p.457; Mungiu-Pippidi, 2013). Because the reforms implemented have been based on this flawed understanding of the problem the measures taken have been misguided and therefore ineffective, in some cases even counterproductive (Karklins, 2005; Persson, Rothstein & Teorell, 2013; Mungiu-Pippidi, 2006). Like corruption welfare fraud is largely argued to be a problem of incentives best counteracted by deterrence reforms in terms of increased and more effective control measures as well as tougher sanctions (see e.g., SOU 2008:35; SOU 2017:37; SOU 2019:59; SOU 2002:35 & SOU 2022:37). Is the Swedish analysis, correct? Can it be that the Swedish conceptualization of welfare crime, like the one of corruption, is flawed and that there are other ways of understanding the development of extensive welfare fraud in Sweden? By interviewing people that have first-hand knowledge and insight into the problems with welfare fraud and how they manifest in Sweden, this thesis aims to investigate exactly that. Is the conceptualization of the problem with welfare fraud in the context of the Swedish welfare state accurate if not; how can we understand the increased problems with welfare crime in Sweden? The remainder of this thesis is structured as follows. It proceeds with a background on the Swedish welfare system and welfare crime. After that, the Swedish analysis of welfare fraud is outlined. In the third section, previous studies of welfare fraud are reviewed, and the theoretical framework is introduced. The thesis then precedes with a discussion of the 6 methodological approach. Thereafter the results are presented and discussed. The thesis is concluded with a summary and concluding remarks. 2 Background 2.1 Concepts and definitions 2.1.1 The Swedish welfare system It is hard to talk about welfare crime without first defining the welfare state and the welfare system. The Swedish welfare state consists of several welfare systems. Each system contains social benefits that are issued by different government agencies such as The Swedish Public Employment Service, The Swedish Board of Student Finance, or a municipality or an unemployment insurance fund (ESV, 2022). From here on the government agencies in charge of issuing social benefits will be referred to as issuing agencies. The payments from the welfare systems are different and have different purposes. Some aim to compensate for a loss of income in case of, for example, unemployment or illness, others aim to remove economic obstacles in order to enable individuals to for example study. Many social benefits also aim to strengthen the economic conditions for vulnerable groups like pensioners and students (ESV, 2022, p.7). The welfare systems are important to many individuals. In 2020 almost 4 million people received payments from the national dental care subsidy alone (Försäkringskassan, 2021, p.15). The costs for the welfare systems are large, in 2020 the welfare systems (excluding the retirement system) accounted for almost two-thirds of total government expenditures (ESV, 2022, p.8). 2.1.2 Welfare crime and incorrect payments The term welfare crime is used to describe crimes aimed at the welfare systems. In many cases, the government investigations and agency reports written on the subject are not mainly focused on welfare crime but rather a broader phenomenon referred to as incorrect payments from the welfare systems (e.g., SOU 2008:75; Sou 2019:57). Another term commonly used in these reports is welfare fraud. Because these terms are often used parallel in the different reports and investigations it is important to initially clarify the meaning of these different concepts. 7 Although different reports and investigations use slightly different definitions of the term, incorrect payments are generally understood as payments that are decided or paid out on incorrect grounds or to an amount that is too high or too low (SOU 2008:74, p.43; SOU 2020:35, p.108; SOU 2019:59, p.63). In contrast to welfare crime, incorrect payments do not solely refer to deliberate attempts to cheat the welfare system. Rather incorrect payments encompass everything from small unintentional errors caused by either the applicant or the issuing agency, to systemic and organized welfare fraud (ESV, 2022; SOU 2008:74, p.43; SOU 2020:35, p.108). Unlike incorrect payments which refers to both intentional as well as unintentional errors, the terms welfare fraud and welfare crime is often used to describe intentional errors caused mainly by the applicant (see e.g., SKR, 2021; SOU 2019:59; SOU 2020:35). Welfare fraud is a specific criminal classification within the Swedish criminal code and refers to all crimes that violate the welfare fraud act (Bidragsbrottslagen SFS 2007:612). According to the welfare fraud act, a person is guilty of welfare fraud if he or she provides incorrect information or fails to report changed circumstances that he or she is obligated to report by law, and thus contribute to the risk of a social benefit being paid out wrongfully or with a too high amount (2 § SFS 2007:612). In contrast to welfare fraud, which is a very specific term defined by law, the term welfare crime is broader and more loosely defined. In many cases it is used as an umbrella term for all types of fraud aimed at the welfare systems, that is both welfare fraud and general fraud aimed at the welfare systems (Brå 2015, p.15; Brå, 2022, p.19; SKR, 2021; SOU 2017:37, p.94, SOU 2022:37). Because welfare fraud does not necessarily have to imply intent (negligence on the part of the applicant is enough to be convicted of welfare crime (4§ SFS 2007:612)), neither welfare fraud nor welfare crime does, by definition solely refer to deliberate attempts to cheat the welfare systems. For the sake of clarity, this thesis will use the terms welfare crime and welfare fraud to describe deliberate attempts to cheat the welfare systems. 2.2 The extent and development of welfare crime in Sweden Welfare crime is often described as, not only a large but also an increasing, problem in Sweden. As late as in December last year, the Swedish minister of finance stated that economic crime has “taken a quite deep root in our welfare systems” and that “the more 8 systematic cheating, is increasing” (Regeringskansliet, 2022, 03:09-04:43). Determining the extent of welfare crime has however proven to be a difficult task, no one knows how extensive the problem actually is. According to the official crime statistics the reported welfare fraud has successively increased over time. Since 2008 when the welfare fraud act (SFS 2007:612) was first initiated, the number of reported welfare frauds has increased by 133 percent. In 2021 the total number of reported welfare fraud was 21 600. 1 Graph 1: The development of reported welfare fraud, 2008-2021 Number of reported welfare fraud 30000 25000 20000 15000 10000 5000 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Source: The Swedish National Council of Crime Preventions statistical database Trying to determine the extent of welfare crime by looking at reported crime is, however, problematic. Welfare fraud is often hard to detect and only a small share of crimes committed are detected (SOU 2017:37, p.25; SOU 2022:37, p.24). As a result, the actual number of reported crimes is likely significantly higher than indicated by the official crime statistics (SOU 2023:8, p.222). Reported welfare fraud is also affected by the agencies’ control activities and their propensity to report crimes to the police. An increase in reported welfare crimes may thus be the result of an increase in actual crimes it may however also be the result of more effective control measures as well as an increased propensity to report crimes (SOU 2023:8, p.222). Over the years, there have been several attempts to estimate the extent of welfare crime using other methods such as expert assessments, randomized controls, and self-declarations studies (ESV, 2022, p.18). According to the most recent estimate made by the delegation for correct 1 All of the statistics referred to in this section are gathered from the Swedish National Council for Crime Preventions statistical database. 9 payments (the KUT-delegation) in 2019, 5,5 percent of all payments from the welfare systems were estimated to be incorrect payments, and about 50 percent of them were argued to constitute welfare crime (SOU 2019:59, p.94;98). The estimated extent of welfare crime has however varied quite largely between different reports and the uncertainty surrounding these estimates is large (ESV, 2011, p.86; FUT-delegationen, 2007, p.87; SOU 2019:59, p.94;98). Even if welfare crime makes up a small percentage of the total number of payments from the welfare systems it still amounts to large economic losses. According to the KUT delegation’s calculations, a welfare fraud of 3,1 percent of the total payments corresponds to an economic loss of SEK 9 billion every year (SOU 2019:59, p.21). 2.3 The expression of welfare crime Welfare crime comes in different forms and can occur at different levels (SOU 2020:35, p.118). To illustrate the different types of welfare fraud that occur in Sweden as well as the frequency with which they occur, the issuing agencies often use the so-called compliance pyramid (ESV, 2022; SOU 2020:35, 119; SOU 2023:8, p.233). The further up the pyramid the less common but also more serious, deliberate, and organized the welfare fraud is (2020:35, p.118). As illustrated by Figure 1 below, the pyramid contains four types of incorrect payments: unintentional errors, intentional errors, welfare fraud, and organized crime (ESV, 2022a). The different types of welfare fraud differ greatly in terms of both motive, severity, and criminal structure (SOU 2020:35, p.118-121). Figure 1: The compliance pyramid Organized crime Welfare fraud Intentional errors Unintentional errors Source: ESV, 20222 2 The figure constitutes a translated version of the compliance pyramid illustrated in The Swedish National Financial Management Authority’s (ESV) report from 2022, p.11. Similar figures are also used in government investigations for example SOU 2020:35, p.119. 10 Intentional errors are a minor type of welfare fraud that occurs when an applicant intentionally provides incorrect information or fails to alter changed circumstances in order to obtain unwarranted social benefits. Examples of intentional errors can be everything from taking out a day’s sick pay without being sick to withholding information regarding certain incomes in order to obtain more benefits than warranted (SKR, 2021, p.10) or regularly requesting temporary prenatal benefits even though the kids are healthy (SOU 2020:35, p.120). While the compliance pyramid indicates that intentional errors are the most common type of welfare fraud in Sweden, many government investigations indicate that advanced welfare fraud is an increasing problem (SOU 2020:35, p.117; SOU 2022:37, p.679). Advanced welfare fraud is defined as a type of fraud that includes one or more elements that are considered more advanced and harmful than normal (SOU 2017:37, p.23). Signifying of these frauds are often the use of companies as a means to cheat the systems, falsified documents such as employment records or medical certificates, or the use of false or hijacked identities (SOU 2020:35, p.121-122). Examples of advanced welfare fraud are the use of double identities in order to obtain double benefits from a particular welfare system, dental care companies that systematically charge the Social Insurance Agency for treatments that have not been performed, or students that falsify certificates in order to obtain student grants even if they are not studying (Brå, 2015, p.32-49; Polismyndigheten et al., 2021; SOU 2017:37, p.250-257; SOU 2020:35, p.121). How widespread the advanced welfare fraud is, is difficult to determine. An attempt to estimate the extent of advanced welfare fraud was made by the government investigation on advanced welfare fraud (SOU 2017:37) in 2017. By looking at the total number of reported welfare crimes from 9 different issuing agencies between 2011- 2015, the government investigation concluded that 84 percent of all of the reported crimes constituted advanced fraud (SOU 2017:37, pp.187; 263). As mentioned in the introduction welfare crime is often linked to organized crime. When the Swedish Minister of Social Insurances, in a recent interview, was asked about the government’s strategy to prevent welfare fraud, the first thing she replied was “Organized crime poses a serious threat to the welfare systems” (Ceasar, 2023). In a press realize issued by the government in December last year the Swedish Minister of Finance made a similar connection and argued that “one of the most serious threats to our welfare is organized crime that exploits the systems for their own gain” (Finansdepartementet, 2022). When welfare 11 fraud is connected to criminal organizations can, however, be difficult to determine. There is currently no clear definition of what organized crime against the welfare system entails (SOU 2017:37, p.196). Actors associated with organized crime often move across the entire compliance pyramid. It is not uncommon that criminal individuals commit welfare fraud as a way to finance a criminal lifestyle (SOU 2017:27, p.471). As a result, even minor intentional errors can be connected to organized crime (SOU 2020:35, p.118). According to the police authority (Polismyndigheten et al., 2019) there are clear indications that the use of companies is common and occurs systematically within organized crime. Companies enable access to various forms of employment and employer supports. After the companies have gone bankrupt, they are often used for wage guarantee fraud. According to the law enforcement agencies, it has also become more common for organized crime to commit fraud by using false identities. False identities are for example used as decoys in companies or to apply for social benefits from the welfare systems (Polismyndigheten et al., 2017). As mentioned above, organized crime also uses welfare companies to commit welfare fraud. According to The Swedish Work Environment Agency, personal assistance and home care are considered to be risk industries for financial crime (Arbetsmiljöverket, 2019, p.19). There are examples of elaborate cases connected to organized crime where assistance providers acquire clients by identifying disabled individuals abroad and bringing them to Sweden (see e.g., SOU 2012:6, pp. 165–166 and ISF & Brå, 2011:12, pp. 61–62). According to the compliance pyramid, welfare fraud involving organized crime is the least common type of welfare fraud in Sweden (SOU 2020:35, p.117-122). This type of fraud is however also the hardest to detect which means that estimations are surrounded by large uncertainties (SOU 2020:35, p.22; SOU 2022:37, p.391). 2.4 Why is Sweden vulnerable to welfare fraud: according to the official analysis? Since the mid-1990s the Swedish government has ordered numerous investigations with the aim of identifying why the Swedish welfare system is vulnerable to abuse and how the problems could best be prevented (See e.g., Dir 2005:52; Dir 2015:96; Dir 2015:52; Dir 2016:60; Dir 2021:39). Even if many of the investigations were conducted years apart, they offer a surprisingly unanimous analysis of why the Swedish welfare systems are so exposed to welfare crime. The Swedish welfare systems are exposed to welfare fraud because they are easy to abuse. 12 While the official analysis highlights several reasons why the systems are easy to abuse the main problem is typically described to be that the fact that Swedish welfare systems are largely based on trust rather than control (Polismyndigheten et al., 2021; Ekobrottsmyndigheten, 2016). Decisions on social benefits are often based on information provided by the applicant or on certificates issued by a third party such as an employer or a doctor (Brå, 2015, p.9; SOU 2022:37, p.72). Although it is well known that falsified certificates and incorrect information constitute one of the main sources of incorrect payments, most authorities do not sufficiently control or verify the information provided (FUT-delegationen, 2007, p.33; SOU 2018:14, p.92; SOU 2020:35, p.131). As observed by several investigations and reports, control activities have tended to be deprioritized in favor of service and efficiency (Brå, 2015, p.88-89; FUT-delegationen, 2007, p.37; KUT-delegationen, 2018, p.34). The extent to which controls are issued varies largely between agencies. While some agencies conduct controls both before and after a payment has been conducted some social benefits are issued without any prior controls (Brå, 2015). Applications regarding dental care subsidies are for example generally conducted digitally and the decision on support is almost always made automatically without any prior verification of the reported information. If controls are conducted, they are done after the payment has been issued which makes it difficult to reclaim the money (Polismyndigheten et al., 2021). This type of process is not unique to the dental care system. The importance of more preventive controls has been stressed by several government investigations (SOU 2017:37, p.121; SOU 2020:35, p.22). The issuing agencies' inability to detect fraud and errors is further excavated by the complexity of the welfare system. Even if the issuing agencies are responsible for their welfare system the different systems are connected and dependent on each other. Decisions made by the Swedish Tax Agency regarding income and residence for example affect decisions on payments from other agencies. Despite this, there are sharp boundaries between the agencies that complicate the control process. In many cases, agencies’ ability to conduct efficient controls is dependent on information from other agencies. Their ability to access this information is however limited by the current legislation which makes it difficult to perform efficient controls (SOU 2020:35, p.125; SOU 2022:37, p.319). The Social Insurance Agency is for example not able to access records of school attendance, something that makes it hard for them to control if temporal parental leave has been issued correctly (SOU 2022:37, p.319). 13 The legislation can also create other obstacles. The Swedish Social Insurance Agency, for example, has limited opportunities to control that personal assistance has been carried out to the extent stated. In some cases, the Public Employment Service has a limited ability to act when they suspect or know that an actor has abused the system in the past. Likewise, the Swedish Tax Agency cannot conduct risk-based controls regarding population records (Polismyndigheten et al., 2021). Together, these shortcomings make it hard for government agencies to detect errors and fraud, something that in turn makes the risks of being detected while committing welfare fraud rather low. According to the government investigation on advanced welfare fraud (SOU 2017:37, p.272), the complex systems and the lack of control do not only entail a risk that welfare fraud will not be detected and investigated it also creates incentives to commit welfare fraud. Another source of concern is the fact that the cost of committing crimes against the welfare system is too low compared to the potential benefits. Several reports argue that the sanctions against welfare fraud have not been sufficient to deter crime and call for stronger and more efficient penalties (SOU 2017:37; SOU 2019:59, p.246; SOU 2022:37, p.357). Many of the reports also state that the law enforcement agencies' ability to investigate and prosecute has been inadequate. In many cases law enforcement agencies have lacked the competence and resources necessary to investigate these crimes, as a result, welfare crime cases have tended to pile up, less severe crimes have been put on hold and many cases have been statute-barred before they have been properly investigated. As a result, the clearance rate has stayed low despite an increasing number of reported crimes (SOU 2022:37, p.357f). To sum up, the many government-agency reports, and government investigations provide a very unanimous and clear analysis of Sweden’s problems with welfare fraud and what is required to overcome them. The government agencies’ ability to detect and sanction welfare crime has been too weak, as a result, the incentives to commit crimes against the welfare systems have been high (e.g., SOU 2017:37, p.271-278; SOU 2022:37, p.361-363). To overcome the problems the current handling of suspected welfare fraud needs to become faster and more efficient, the control needs to increase and society’s reaction to these violations needs to become stronger and tougher (SOU 2017:37, p.25-31; SOU 2020:35, p.21; SOU 2022:37, p.21-25). As stated by the former Minister of Justice, the social democrat Morgan Johansson “It has been way too easy to be a welfare criminal” (Svenska Dagbladet, 2017). 14 The image of welfare fraud as an incentive-based problem best solved by increased control and tougher sanctions is not only apparent in government investigations and government agency reports but also in the political debate. In a debate article from 2021, representatives from one of the current incumbent parties, the Christian Democrats stressed that “Our welfare system is best protected by a functioning legal apparatus that can detect and denounce fraud” (Bush, Carlson & Eklind, 2021). In an interview with the Swedish public service television company (SVT), the economic political spokesperson for the Swedish democrats drew similar conclusions and stated that “trust is very nice when it’s working but what we need now is more of a control system” (SVT Nyheter, 2022). The need for more extensive control is not only emphasized by the parties to the right. In an interview with SVT, the leader of the social democratic party and former prime minister Magdalena Andersson stressed the importance of welfare crime being met by a strong reaction and argued that “if you try to cheat or get money on false grounds you shall be met by strict controls and punishments” (SVT Nyheter, 2021). 3 Previous studies and theoretical framework 3.1 Previous studies of welfare fraud Although welfare fraud has gained considerable political and media attention in several countries (Dobson, 2019) the academic interest in this phenomenon has been rather modest (Tunley, 2011; Yaniv, 1996). Most of the studies conducted have focused on the discourse surrounding welfare fraud (e.g., Button & Tunley, 2015; Korsell, Hagstedt & Skinnari, 2008; Lundström, 2013) and the effect of government anti-fraud efforts (e.g., Appelgren, 2019; Dobson, 2019; Gustafson, 2009; McKeever, 1999; Shonhadju & Maulidi, 2021). While there are some studies that have tried to identify the underlying causes of welfare fraud most of them have focused on the individual motives behind welfare fraud (Dean & Melrose, 1997; Regev-Messalem, 2013; Tunley, 2010). There is a lack of comparative research (Sainsbury 2003, p.286) and little scholarly attention has been spent on identifying the factors explaining aggregated compliance with welfare regulations or why the level of welfare fraud seems to vary between countries as well as over time. However, there are other acts of non-compliance that have to a greater extent been theorized. In the remainder of this section, these different theories of non-compliance will be discussed. 15 3.2 Theories of Compliance and Noncompliance Acts of noncompliance like tax evasion or draft evasion, have traditionally been studied through the prism of compliance theories (e.g., Graetz & Wilde, 1985; Levi, 1988; 1977; Kirchler, Kogler & Muehlbacher, 2014). Compliance theories offer theoretical mechanisms explaining why and under what conditions people choose to comply or refrain from complying with various state-demanded obligations or legislations (Levi, 1997). All states need some degree of compliance to be successful. It would for example be hard to uphold a generous welfare system if people did not comply with government tax demands and refrained from contributing to the public funds needed to finance the system (Zhang et al., 2015). While compliance can be achieved by force, continual coercion through auditing or the threat of sanctions is a costly and often ineffective method to enforce compliance and thus not seen as a desirable solution to many states (Grimes, 2006; Levi, 1988). Instead, it has been argued that the best way to make people comply with state demand is to mobilize them into some type of voluntary compliance (Levi, 1988). The important question thus becomes, what makes people comply with state demands, and how does a state mobilize its people to voluntary compliance? According to Margaret Levi (1997, p.17-19), compliance is a behavioral response to government policy, however, that response can be interpreted in different ways. The research on compliance offers different interpretations of that response and thus also varies in terms of what factors they perceive motivate it. 3.2.1 Compliance as a rational choice Within studies of taxation, tax compliance has traditionally been analyzed using economic models of compliance in which compliance is perceived as a rational choice (Allingham & Sandmo, 1972; Becker, 1968; Graetz & Wilde, 1985; Kirchler, Kogler & Muehlbacher, 2014). Like the Swedish analysis of welfare fraud, these models perceive compliance largely as a response to incentives. From this perspective, compliance is assured by altering the incentives so that compliance becomes the most profitable behavior. This is best done by increasing the probability of detection through the implementation of control regimes and audit schemes and by increasing the cost of getting caught by inducing hard punishments (Allingham & Sandmo, 1972). While the economic model offers a parsimonious explanation of noncompliance, the empirical evidence in favor of this model is ambiguous. Although a lot of studies have found that an increased probability of tax audits has a positive effect on tax compliance (see for 16 example Alm, Sanchez & De Juan, 1995; Pommerehne & Weck-Hannemann, 1996), the effect is often weaker than would be expected (Fischer, Wartick & Mark, 1992; Kirchler, Kogler & Muehlbacher, 2014). When it comes to fines and other economic sanctions most studies find that these types of policies have little effect on compliance (Ali, Cecil & Knoblett, 2001; Pommerehne & Weck-Hannemann,1996; Webley et al., 1991). Furthermore, several scholars have pointed out, that people often comply even when the cost of compliance exceeds the cost of non-compliance, something that should not happen if the decision to comply was solely based on material considerations as suggested by the economic model (Graetz & Wilde, 1985; Levi, 1997). According to some scholars, policies based solely on incentivizing behavior, are not only insufficient but can in fact be counterproductive. In a case study of tax compliance in Tanzania, Fjeldstad and Semboja (2001) for example found that while an increased probability of prosecution enhanced compliance, oppressive tax enforcement reduced the citizens’ willingness to pay taxes. Similarly, Strümpel (1969) illustrated that unfair penalties had a negative impact on attitudes toward the tax office as well as taxes in general. Similar results have also been found in studies of welfare fraud. By conducting 35 interviews with claimants who fiddle Hartley and Melrose (1997) found that fiddlers were rarely deterred by the risk of getting caught. From the fiddler’s perspective, low income was a bigger concern than the probability of being caught fiddling. 3.2.3. Compliance as contingent consent3 Another theory that offers a completely different interpretation of compliance is the model of contingent consent offered by Margaret Levi (1988; 1997). According to Levi (1997), compliance is informed by both ethical and material considerations. While people want to make sure that their personal interests are being satisfied as far as possible, they generally also want to contribute to the common good. However, people are not altruists that comply out of the pure goodness of their hearts. If people are to comply, they need something in return, compliance should thus be seen as a bargain or an implicit social contract between the government and its citizens (Levi, 1991). Like all bargains compliance is contingent on trust. After all, if one party does not believe that the other one is going to keep their part of the 3 This section is partly based on the paper QoG and abuse of the welfare state that I wrote for the course SK2212 during spring 2022 17 bargain there is not much point in making a deal at all. As a result, compliance is heavily influenced by the perceived trustworthiness of the government (Levi, 1997). The citizens’ evaluation of a government’s trustworthiness is dependent on both procedural as well as substantial fairness (Rothstein & Samanni & Teorell, 2012, p.10). If people are to comply, they want to make sure that their compliance results in valuable and qualitative social goods (Levi, 1991, p.135; Rothstein & Samanni & Teorell, 2012, p.8). As indicated by several studies the perceived quality of public goods as well as the effectiveness of government institutions increase both trust in government as well as support for the welfare system (Kumlin, 2007; Munroe & Duckett, 2016; Schneider & Popic, 2018). That the government produces valuable public goods does however seem to be neither a necessary nor a sufficient condition for compliance. As pointed out by Levi (1997) people often comply with policies they do not prefer. What seems to be important is instead the process through which these policies are made and implemented. If people are to comply, they need to know that the policy-making process is fair and that the policies are implemented in a fair way. If a group’s interests are systematically ignored, or policies are implemented in a discriminatory way the citizens are unlikely to perceive the government as trustworthy, and as a result, they will be less inclined to comply with its demands (Levi, 1991, p.135; Rothstein & Samanni & Teorell, 2012, p.8). Several studies have found that if people believe that government decisions are fair and impartial, they are more inclined to trust and support the government as well as comply with state demands (Marien & Werner, 2019, p.72; Linde, 2012, p.410). While institutional trust is a necessary condition for compliance it is not a sufficient one. If people believe that others are failing to keep their part of the bargain by avoiding taxes or abusing the welfare system, they are likely to do the same. Compliance is thus not only contingent on institutional trust but also on social trust, or as Levi (1997, p.24-26) calls it, ethical reciprocity. In other words, compliance is not based on one but rather a pair of implicit social contracts, a vertical contract between the state and its citizens and a horizontal contract between the citizens (Persson, 2008, p.22). Like institutional trust, social trust is also largely affected by the government’s actions. According to Rothstein and Stolle (2008, p.446), public servants are not only representatives of the state but also people in general. The behavior of the public servants is therefore not only an indicator of the moral standards of the public administration but also of the society at large. If people perceive public officials as 18 trustworthy and impartial it signals that citizens are in general trustworthy (Thisted Dinesen & Mannermar Sønderskov, 2021, p.5-6). The trustworthiness of others can also be endured by coercion. By making sure that transgressors are sanctioned the government reassures citizens that compliance will not be “a sucker choice” (Levi, 1991, p.134). 3.2.3. Norms and compliance A third set of theories links compliance to broader cultural norms and values. Like Levi (1997), these theories argue that people are inherently moral beings that want to do the right thing. Their behavior is thus largely dependent on their moral values, that is their perception of right and wrong. From this perspective, people will engage in noncompliant behavior if they believe that such behavior is morally defendable or because it is perceived as acceptable within the community (Grasmick & Green, 1980; Meares & Kahan, 1998; Paternoster & Simpson. 1996; Robinson & Darley, 1995, p.201-203; Zhang et al., 2016). There are some empirical supports for this theory. Studies on tax evasion have for example found that tax compliers generally have more positive attitudes toward taxation and hold stronger social norms regarding the moral correctness of adhering to the law than do tax evaders (Kirchler, 2007). Similar results have been found in studies of welfare fraud. In an interview study of 35 individuals convicted of welfare fraud Dean and Melrose (1997) found that most respondents did not perceive fiddling to be dishonest. Furthermore, they found that nearly all interviewees believed that fiddling was common. While few mentioned this as a motivating factor behind their decision to fiddle it made them more comfortable with their own actions. As pointed out by Tunley (2011, p.309) this could indicate that welfare fraud might be more normalized in some micro-communities compared to others, in other words, the propensity to commit welfare may vary between certain social and cultural groups. There is some evidence in favor of this hypothesis. Dean and Melrose (1997) for example found that the attitudes towards welfare fraud differed between the interviewees in their study. Older people, women, and black people were more anxious about their fiddling and exhibited a stronger attachment to the idea of the welfare state than the younger respondents, men, and white people. That welfare fraud may be more accepted within certain social groups is also indicated by Shiri Regev-Messalem (2013). By conducting 49 in-depth interviews with Jewish women in Israel she found that most of the interviewees perceive welfare fraud to be morally defendable, in fact, they encouraged it. Even if the women clearly stated that they knew that welfare fraud 19 was a criminal offense they argued that it was legitimate because the social security system denied them what they were morally entitled to. Within this community welfare fraud was thus seen as a political act, a way to claim their right and thus morally defendable. 3.2.4. Welfare Fraud and economic need A fourth and final theoretical perspective on noncompliance is offered by Dean and Melrose (1997). In their interview study with 35 ‘petty fiddlers,’ they found that the main motive behind welfare fraud is economic need. People fiddle because they have to in order to obtain a sufficient income. According to their findings fiddling is more strongly connected to values of work and consumption than values of welfare citizenship as suggested by Regev-Messalem (2013). Similar findings are offered by Tunley (2011). By reviewing 66 high-profile welfare fraud cases in the UK, he found that the need resulting from the gap between economic means and goals is a motivating factor of welfare fraud. The goal for most welfare fraudsters is to obtain an income that enables them to meet their basic needs. 3.2.5. Summary To sum up, the literature on compliance show that there are more ways of explaining problems with noncompliance than the one offered by the official analysis of welfare crime in Sweden. While the literature indicates that acts of noncompliance such as welfare fraud may be the result of insufficient control systems and weak sanctions (Allingham & Sandmo, 1972) it offers several alternative explanations for this type of aggregated noncompliance. From Levi’s (1997) perspective the increased problem of welfare fraud in Sweden could indicate two things: either the citizens do not (or no longer) perceive the government as trustworthy, or the citizens believe that the level of noncompliance has increased and are thus no longer willing to cooperate. As pointed out by Levi (1997) the perceived level of noncompliance, as well as the perceived trustworthiness of the government, is affected by multiple factors: the government’s ability to sanction noncompliers and ensure the citizens about the honest behavior of others, the perceived fairness of policies, as well as the perceived fairness of the political process. The increased problems with welfare crime could thus be the result of a change in either one or several of these factors. As suggested by for example Dean and Melrose (1997) as well as Regev-Messalem (2013) compliance is also affected by norms as well as economic need. Another possible explanation for the increased problems with welfare crime is thus that there has been a normative shift in society and that welfare fraud is more normally accepted today than it used to be, alternatively that welfare fraud has become more 20 normalized in certain groups in society. As suggested by Dean and Melrose (1997) it could also be the result of an increased poverty. 4 Aim and research questions During the past decades, the Swedish government has taken a number of measures to prevent welfare fraud. Even if many of these efforts seem to have been effective (Prop. 2018/19:132; SOU 2020:35, p.117; SOU 2022:37, p.341; pp.22-23) Sweden continues to have not only extensive but also increasing problems with welfare fraud (SOU 2019:59, p.29; SOU 2020:35, p117). As highlighted by previous studies of corruption, reforms sometimes fail, not because they are not extensive enough but because they are based on a mischaracterization of the problem (Persson, Rothstein & Teorell, 2013). A possible explanation for the puzzling development currently observed in Sweden is thus that the Swedish analysis of the problem with welfare fraud is based on a flawed understanding of the problem and that the efforts taken therefore have been, at least partly, misguided and by extension ineffective. This thesis aims to investigate this possible explanation. More specifically it strives to answer the following research questions: • Is the Swedish conceptualization of the problems with welfare fraud accurate? • If not, how can we understand the increased problems with welfare crime in Sweden? Answering these questions is important. As illustrated by the scholarly debate surrounding anticorruption reform measures based on a flawed understanding of the problem risk becoming ineffective or in worst case counterproductive (Karklins, 2005; Persson, Rothstein & Teorell, 2013; Mungiu-Pippidi, 2006). Since welfare fraud is not only contributing to large economic losses but also risk harming the legitimacy of the welfare system (Dir 2005:52; SOU 2019:54, p.266; SOU 2022:37, p.21) it is crucial that the efforts taken are effective. By investigating if the official analysis of welfare fraud is accurate this thesis can contribute to knowledge that is needed to effectively prevent welfare fraud. 21 5 Research design and methods 5.1 Methodological approach The thesis combines a deductive and inductive approach. While it aims to test whether the official analysis of welfare fraud is accurate it also strives to capture other potential perspectives of welfare fraud to generate deeper knowledge. This is done by conducting in- depth informant interviews with government officials and local authority employees who work directly with welfare fraud. Because interviews allow the informants to give rich descriptions of their experiences and perceptions it is an effective data-collection method for studies with an explorative aim (Esaiasson et al., 2017, p. 260). The interviews conducted were semi-structured. In semi-structured interviews, the questions are generally open-ended which invites the informants to speak freely from their own experiences. While these types of interviews are guided by a pre-determined interview guide, they also allow the interviewer to deviate from the script and dig deeper into specific areas based on the interviewee’s responses (Magnusson & Marecek, 2015). This flexibility enables a more comprehensive understanding of the informants’ perceptions and thus allows for unexpected insights. The interview guide is an operationalization of the thesis analytical framework, which builds on Carol Bacchi’s (2009) ‘What’s the problem represented to be’ (WPR) approach. In accordance with the analytical framework, the pre-determined interview questions were organized around three central themes: problem, cause, and solution. The interview guide thus allowed the informants to reflect upon different components of the problem something that according to the WPR approach help elicit the informants’ understanding of welfare fraud as well as the underlying assumption on which the understanding is constructed (Bacchi, 2009). In order to generate a conversational tone and make the informants feel comfortable opening up and reflecting on their perceptions and insights the interviews were initiated with so-called ‘easy questions’ where the informants were asked to describe the welfare fraud aimed at their agency, before the more sensitive topics regarding the nature of the problem and their reflections regarding the government’s strategy to prevent welfare fraud were introduced. To save the more sensitive topic for later is preferable because it gives the informant time to become more comfortable with the interview situation which makes them more likely to reflect freely (Magnusson & Marecek, 2015). In order to accommodate the diverse viewpoints and perspectives of the informants the interview guide was adjusted for each informant 22 dependent on their particular experiences (the two interview guides that functioned as a base for all interviews are included in the appendix). Since the informants spoke Swedish, all interviews were conducted in Swedish. All of the data has thus been manually translated by the author. Data translation can be tricky, there is always a risk that the original meaning is lost or alternated through the translation process (McKenna, 2022). To reduce the risk of translational errors the informants were asked to confirm and validate the translations of their quotes before the thesis was submitted. Because many of the informants were situated in other parts of Sweden the interviews were mainly conducted via video calls on Zoom. After approval from the informants, all interviews were recorded. Each interview generally took between 60 and 75 minutes. 5.2 Selection of Informants When conducting informant interviews there are two important selection criteria, centrality and perspective. The aim is to find informants that have relevant knowledge and insight into the phenomenon under study while also making sure that all relevant perspectives of the problem are covered (Magnusson & Marecek, 2015). Since government agencies are the main target of welfare crime as well as a central part of the government’s efforts to prevent it, employees working with welfare crime at these agencies were expected to have firsthand knowledge of welfare fraud. Because it is reasonable to believe that employees from the different agencies have different experiences of welfare fraud the sample of informants includes employees from all of the issuing agencies. Because the Police Authority and the Economic Crime Authority are responsible for investigating these crimes, the sample of informants also included one representative from each of these agencies. While first-hand experience working with welfare fraud can be regarded as a necessary condition to make a well-founded analysis of welfare fraud, it is not the only type of knowledge that could be considered relevant. As highlighted by Levi (1997) noncompliant behavior such as welfare fraud can be motivated by several different factors. If welfare fraud is motivated by for example a lack of trust in the government or a decline in ethical reciprocity, government officials may not be the ones who have the best knowledge of this. To capture these alternative perspectives on welfare fraud the sample also included a number of informants who through their work are in contact with a large number of residents and 23 therefore can be expected to have good insight into the residents’ trust in government institutions as well as their attitudes towards welfare crime. Because welfare crime is at least indirectly connected to socioeconomically vulnerable neighborhoods (Polismyndigheten et al., 2021), the informants were selected based on their proximity to the residents in these types of areas. In total 15 interviews were conducted with 18 different informants from 10 different government agencies as well as two municipalities, one non-profit organization, and one university (a table of all informants is available in the appendix). The informants from the government agencies were selected based on centrality, that is their proximity to and experience of working with issues relating to welfare fraud. The informants were identified by calling or e-mailing the customer service at each agency who then identified an appropriate interview subject. In some cases, the informant was handpicked based on their previous experience. The investigator at BRÅ, Johanna Skinnari, and the docent in peace and development studies Hans Abrahamsson were for example selected based on their previous research on welfare crime in Sweden (e.g., Brå, 2015; Brå 2016; Korsell, Hagstedt & Skinnari, 2008) and social contracts (e.g., Abrahamsson, 2012; 2015). The representative from the non-profit organization was identified through the recommendation of the municipality of Gothenburg. The field assistants were identified based on a recommendation from the local police in that municipality. Since the informants are treated as sources of information it is important to evaluate the trustworthiness of the informants in the same way you would a written source. There is always a risk that the informants are biased, that they have an interest in presenting the phenomenon in question in a particular way or withholding particular information (Esaiasson et al., 2017, p, 272). It is for example unlikely that the police authority would argue that Sweden’s problems with welfare fraud are due to the fact that the police do not sufficiently prioritize or investigate these crimes. Since the government agencies follow the government’s strategy to prevent welfare fraud it is reasonable to assume that the informants from these agencies may be overly positive towards control measures since this is an important part of their job and they want to be loyal to their employer. In order to reduce the risk of bias and create a situation in which the informants felt comfortable with speaking openly they were offered the opportunity to be anonymous, after approval from the informants all of them (except for the researchers who were selected based on their previous work) were anonymized and cited with title and workplace. While the risk of bias has further been eliminated by 24 choosing informants from a wide variety of perspectives and employers the risk of bias should be kept in mind when reading the results. 5.3 Data Analysis and analytical framework In order to elicit the essential aspects of the informants’ problem descriptions analytical concepts are required. These concepts aid in organizing and illuminating aspects of the informants’ statements that may not be directly observable (Esaiasson et a., 2017, p.213). After the interviews were transcribed, the data were therefore coded around a few central analytical concepts. The concepts were selected based on Carol Bacchi’s WPR approach. The WPR approach is a method for identifying how problems are represented and constructed in policies. The approach involves a number of questions designed to reveal the problem representations embedded in policy discourse, including: • How is the problem represented? What kind of problem is X represented as, what is X an expression of? • How is the cause of the problem represented? Who or what is represented as the cause of the problem? • How is the solution to the problem represented? What policy proposals are enabled by the problem representation? What the approach suggests is that questions regarding the nature of a problem, its causes, and solutions help illuminate the specific understanding of a problem, in other words, what the problem is perceived to be. Since the aim of the thesis is to identify the informants’ understanding of the problem of welfare fraud, these questions were regarded as suitable concepts. After reading the transcriptions closely the informants’ answers to the three questions posed in the WPR approach were identified in the material. The identified answers were then coded into different categories using the software Nvivo. The parts of the interviews that constituted an answer to the first question were coded as ‘problem descriptions’. Sections that indicated an answer to the second question were coded as a ‘cause’ and so on. In this way, revenant parts of the interviews were identified and highlighted. 25 When the relevant parts of the transcripts had been identified and coded, they needed to be categorized into different perceptions or problem representations. In studies of perceptions or framings of a particular phenomenon, ideal type analysis is a commonly used tool (Esaiasson, 2017, p.137). Since the phenomenon of interest in this thesis is understandings or analyses of welfare fraud the ideal types constructed for this study constitute ideal type descriptions or understandings of welfare fraud. An ideal type does not capture real perceptions or analyses of a phenomenon but rather an extreme version of a certain type of perception or analysis. The aim is to capture the distinctive feature of that perception or analysis in order to distinguish it from other types of perceptions (Esaiasson et al., 2017, p.140). The different theoretical perspectives offered by the literature on compliance constitute different ways of framing and understanding noncompliance such as welfare fraud. While the theories are somewhat overlapping, they rest on opposing assumptions of the nature, causes, and solutions to non-compliant behavior and thus function as a good base for the construction of ideal types. Based on the different theories of compliance, four ideal types were constructed. An economic model of welfare fraud based on the economic theories of compliance, a contingent consent model based on Levi’s model of contingent consent, a normative model based on the normative theories on compliance, and a model of welfare fraud as a necessity based on theories framing welfare fraud as an act motivated by economic need (For an illustration of the analytical framework see table 1 below). The ideal types functioned as the analytical framework of this thesis and were used to interpret the responses given by the informants. The informants’ understandings of welfare fraud were identified through a qualitative assessment of how well they correspond to the different understandings brought forward by the theories of compliance, that is the ideal types. Important to note is that it is not the informants but rather the ideas or perceptions they express that are of importance. As a result, it was the specific statements and expressions made by the informants that were interpreted and not the informants as such. To reduce the risk of overlooking interesting perceptions of welfare fraud not predicted by previous research the analytical process was guided by an inductive approach meaning that the material was analyzed a second time with the aim of identifying additional understanding of welfare fraud not included in the analytical framework. In order to strengthen the credibility of the study the results of the analysis were presented using ‘thick descriptions’ which means that the 26 interpretations made were shown using illustrative quotes which allow the reader to draw their own conclusions (Tracy, 2010, p.843). Table 1: Analytical framework Ideal types Problem Cause Solution How is the problem represented? What are the causes of the What solutions are What is presumed to be the problem? Who or what is presented? Who is problem? What is it an expression represented as the cause of responsible for solving of? To whom is it assumed to be a the problem? the problem? problem? Ideal type 1: The • Incentive based problem • Opportunity and • Deterrence regimes economic model (the • Criminality incentive structures • Control systems Swedish analysis) • Deliberate malicious act • Lack of monitoring and • Sanctions • Act of greed control (low risk of getting caught) To whom is it a problem: theft from • Lack of sufficient the taxpayers sanctions (low cost of getting caught) • Generous welfare systems (high profits) Ideal type 2: The model • A breach of the social contracts • Distrust in government • Measures aimed at of contingent consent • A Reaction to political policy • Dissatisfaction with both formal and • An implicit withdrawal of consent government policy informal institutions • An expression of discontent • Perception of • Impartial and fair • A reaction to the expected government as unfair implementation of behavior of others • Lack of ethical policies reciprocity/ social trust • Inclusive political To whom is it a problem: the process government and the citizens • Deterrence regimes (society) that reassure the citizens about the ethical reciprocity of others Ideal type 3: The • A normatively defendable act • A normative shift • Measures aimed at normative model • A withdrawal of consent/ a regarding the acceptance the informal political protest of welfare crime institutions uphold norms of non- 27 To whom is it a problem: no one, compliant/fraudulent it’s a legitimate act behavior Ideal type 4: The • An act of survival/Necessity • Economic need • Higher allowances survival model • Need rather than greed • Poverty alleviation • Efforts aimed at To whom is it a problem: the creating legitimate offender that is forced to commit opportunities for fraud self-sufficiency Other potential perceptions or understandings 6 Results 6.1 Problem - what type of problem do the informants perceive welfare fraud to be? As highlighted earlier in this thesis, the many government investigations and government agency reports released in the past decades paint a very clear image of welfare fraud, its underlying causes, and solutions. Like the economic model, the Swedish analysis largely holds that welfare fraud is an incentive-based problem. The reason why Sweden is seeing such extensive problems with welfare fraud is because the welfare systems have been easy to abuse. The control functions have been insufficient and too few crimes have been met by a reaction from society (e.g., SOU 2019:59, p.246; SOU 2020:35, p. 131; SOU 2022:37, p.23- 27). When the informants were asked to give their picture of Sweden’s problems with welfare fraud and describe the welfare fraud aimed at their specific agencies, many of them provided descriptions similar to the one offered by the official analysis. A controller at the Social Insurance Agency for example argued that: Most people know that they don’t really have a right to do this. I think mainly it’s financial incentives that guide people to do it. You simply want more, you want more money (…) I also don’t see that it’s about some kind of poverty, that you do this because you have to. It’s more about the fact that you just want more. You want to be able to live beyond your means or have expenses or buy things that you can’t really afford and then you do this [commit welfare fraud]. 28 A similar view is offered by the operative coordinator at the Western police region’s fraud unit. In accordance with the Controller at the Social Insurance Agency, he perceives welfare fraud to be a problem of economic incentives. He argues that there are three types of welfare criminals: It’s about those who do this really carelessly and aren’t that careful and maybe apply for something and then they get money, so they think, yes but it’s pretty good, so they don’t care that much more about it, and maybe don’t take it so seriously. And then there are those who actually do this to make money from it, a little bit more elaborate. Then we have the organized part. It’s like this is their way of making money and they do it as much as they can and find systems for this and organize around it. This is typical for assistance fraud; you see that there is a gap where we might make money and where we can take advantage of the issuing agencies. In many cases, the informants connect economic incentives to opportunities for crime. They argue that the reason why Sweden is exposed to welfare fraud is because the systems create opportunities and incentives to commit crimes. A social service worker at the unit for incorrect payments in Gothenburg explained it like this: It is economic incentives. I mean the risk of getting caught has historically been low and even if you get caught, the penalties are unfortunately not that noticeable. It’s very difficult to be convicted of this type of crime. I think you have the answer right there. This is no, like organized civil disobedience or like silent protest. That’s not how I would describe it, but it’s more something that everyone does. It’s a smart way. You’re stupid if you don’t do it. The problem description highlighted by these quotes lies close to ideal type one (the economic model) as well as the official analysis of welfare fraud. It highlights an understanding of welfare fraud in which welfare fraud is understood as an incentive-based problem. People know that they are doing something wrong, they know that they are committing a felony, but they do it anyway. Not because they need to but because they can, because the system creates opportunities to do so. In accordance with the economic model, they hold that, if the profits are high and the risk is low there will be incentives to abuse the 29 system. This type of economic perspective is perhaps best summarized by the head of the unit for correct payments at the Swedish National Financial Management Authority (ESV) who stated: If you back up, why is this even happening at all? It is of course because it is worthwhile, because it’s possible, and because the control is not sufficient. That welfare fraud is a problem motivated by economic incentives is however far from the only understanding of welfare fraud offered by the informants. A completely different understanding of welfare fraud is for example offered by the docent in peace and development studies, Hans Abrahamsson. From his perspective welfare fraud is “In the end, a matter of trust”. He argues that: There is a mutual lack of trust. That is, we know that the citizens have a decreasing, in some neighborhoods in particular (…) there is very little trust, and it has to do with the fact that the state’s modern institutions do not reach. The modern institutions are retreating. And then we have a lack of what I call vertical trust between the citizens and the authorities which is connected to the unequal development. The state does no longer deliver what the social contract once stipulated (…) and then people don’t feel bad for committing welfare fraud, because they believe that it’s totally reasonable. What Abrahamsson is saying is in other words that welfare crime is the result of a breach in the vertical social contract. That is however not the only contributing factor. In accordance with ideal type two, he argues that Sweden’s problem is also connected to a decreasing social trust: With the increased inequality in Sweden the horizontal inequality increase. When people start to compare themselves to other social groups, they become convinced that other people are cheating and then they become inclined to cheat themselves. You go on rumors; this person is doing this and this, and then the entire social trust is eroded He summarizes his analysis by saying that: 30 I mean that welfare fraud, regardless of who commits them, whether they are criminals or not, it’s a lack of the social contract, it is a lack of social belonging (…) it is the changing role of the state and increasing inequality The understanding of welfare fraud offered by Abrahamsson is almost identical to the one offered by Margaret Levi (1997) and thus fits perfectly within ideal type two, the model of contingent consent. From his perspective, Sweden does not have problems with welfare crime because its welfare systems create opportunities and incentives to commit fraud. Welfare fraud is not the result of economic incentives it is a silent protest, a reaction to the state’s failure to live up to their part of the bargain and decreasing levels of social trust. It is a sign that the social contracts have been breached. Abrahamsson is not alone in his analysis. The operations coordinator at a local non-profit organization in one of Gothenburg’s vulnerable neighborhoods offers a similar analysis. She even questions the entire definition of welfare crime and argues that: Who defined welfare fraud? Because if you think about how I misinterpreted the definition from the beginning, I think that’s an interesting take on it. Because for me, a welfare crime can also be dismantling the welfare. I’m partly thinking about our public housing companies (…) we talk about it all the time, that there are dire challenges in our suburbs. People are having a hard time right now, but we’re still raising the rents (…) It’s the same with overcrowding. It is a terrible social problem, absolutely, but what happens if you send a family with children out on the streets because they hold a black lease? Who takes care of them in homelessness then? Is that not also a crime? The quote highlights a completely different perspective on welfare crime than the one offered by the official analysis. From this perspective, the real welfare crime is not the fraud committed by individuals but rather the actions taken by the state. By dismantling the welfare system, the state forces people to take other actions to ensure their basic needs, or as the coordinator at the non-profit organization herself put it: You yourself do not choose to become a criminal but rather you have to live off what you can get (…) If we as a state cannot ensure that the basic things that we 31 need as humans to survive are satisfied, then I think it is hard to expect people to have faith in the system was supposed to solve it but didn’t. Because I think that survival comes first and then you can try to feel a sense of obligation towards others, and I think that’s universal. The understanding of welfare fraud illustrated by the quote is well in line with ideal type two. It indicated that welfare fraud is a reaction to policy, if the state fails to provide what it promised to provide the citizens are no longer obligated to comply with state demands. It does however also align with ideal type four, the survival model. What the informant is saying is not only that welfare crime is a response to government policy but also that it is an act of survival. That welfare fraud is something people do because they must. This perspective is highlighted by several of the informants. A field assistant at the municipality of Växjö put it like this: I don’t think you do it [commit welfare fraud] because you are disappointed at the state and feel a certain grudge or anger towards the authorities and the state but simply because you need to have a better standard for your family and children. I think that’s it. Then it creates this, I mean you know that you’re cheating. So, you’re worried all the time. When will I be discovered? (…) you get very very little [financial aid] to be able to live. It is very minimum, minimum, minimum compensation. And I’m not advocating cheating, but I still think it is humiliating that you should receive this little when you have children and teenagers. Some families are forced to live this way [to live on cheating], because of our system. That welfare fraud is a problem of need rather than greed is also highlighted by informants from the issuing agencies. A legal practitioner at the Swedish Pensions Agency for example stated that: The most common reason [for why you commit welfare fraud] is simply that you are short on money, and you understand that if I tell you [the issuing agency] that I’m married or cohabiting, I will get less money. It’s better if I say that I live alone for example, or that I don’t have this extra income and so on. If you look at the most common cases, It’s clear that it’s an economic issue. 32 The quotes above thus highlight a very different understanding of welfare fraud than the one put forward by the official analysis namely that welfare fraud is a problem of economic inequality or poverty that put people in a position where they feel that they have to engage in welfare fraud to survive. In addition to the perspectives listed above, many of the informants mentioned that people do not seem to perceive welfare fraud to be so serious and imply that there may have been a normative shift in society, that people take welfare crime more lightly today than they did in the past. The controller at the Social Insurance Agency for example made the following reflection: Over the years we have been able to see a shift in terms of attitudes towards both the social insurances and welfare crime in general. That you may not see it as a crime. If you deliberately withhold information from the Social Insurance Agency to get compensation, you have committed a crime. As a rule, you get reported to the police (…) but I don’t feel that people see it as a crime or see it as that serious. These quotes highlight a fourth understanding of welfare fraud that in accordance with the ideal type three, the normative model, which holds that the reason why Sweden is seeing such extensive problems with welfare fraud may be because the attitudes towards welfare crime have become more accepting. People do not believe that they are doing something wrong but rather that cheating the welfare system is acceptable. From this perspective welfare fraud is mainly a problem of norms and attitudes even if opportunities and punishments regimes are likely a contributing factor. While the perspectives highlighted above offer very different views regarding what type of problem welfare fraud is they all agree that welfare fraud is in fact a problem. This view is not shared by all informants, on the contrary, several informants are hesitant that welfare crime is a problem at all. The head of the unit for correct payments at ESV for example, reasoned like this: You [the interviewer] said that we could start by asking about the problems with welfare fraud, and then I thought like this, yes but is it a problem? We actually ask this question in our latest report. How big of an error is, ok? (…) How much can we put up with? And we usually don’t ask politicians or the authorities that 33 question but it’s the one we’re trying to actualize now. 2 percent, as we say in the study, it might be ok. 14.6 billion may also be okay. But is it okay with 18 percent within temporary parental allowance? Maybe not (…) Let’s work on that and leave something else that might have 5 percent incorrect payments be because it’s such a small benefit or it’s irrelevant for some other reason. So, it’s sort of a priority discussion. How much error can it be and where? The question she poses is in other words, is welfare fraud a problem? Or rather when does it become a problem? The decision-making officer at the Swedish Migration Agency offers a similar perspective: If you compare the Migration Agency’s incorrect payments with the Pensions Agency’s, it’s peanuts. They pay higher sums to individuals than we do (…) We’re talking about incorrect payments which are very important for the state, and it should not happen and at the same time when you’re talking about such low sums and the applicant, maybe get SEK 500 more in his wallet. There’s no organized crime there, maybe they bought a chocolate bar that time because the kids wanted it last time, but they couldn’t afford it. What the analysis indicates is thus not only that there are other ways of understanding welfare fraud than put forward by the official analysis but also that the entire fundament on which the official analysis is built, namely that welfare fraud is in fact a problem, may be off. 6.2 Cause - why does Sweden have problems with welfare fraud? As suggested by Bacchi (2009) the perceived causes of a particular problem reveal something regarding what the problem is represented to be. What the informants believe is the causes of welfare fraud can thus both contribute to a more elaborate picture of how welfare fraud is understood and help identify additional understandings. In accordance with the official analysis, many of the informants argue that the source of Sweden’s problems mainly lies within the welfare system itself. The reason why Sweden is exposed to welfare fraud is that the welfare systems are based on trust rather than control which makes them easy to abuse. As 34 stated by the crime prevention specialist at the Economic Crime Authority, the Swedish welfare systems were never built to prevent welfare fraud: We have a system that relies heavily on self-assurance. The systems are built on people not cheating. When it becomes obvious that people are cheating the systems, we try to seal the holes and rebuild them, but we can’t keep up with that, it’s hard to catch up (…) we need to develop the control systems. What the informant is saying is in other words that the trust-based Swedish systems were not built to prevent fraud, for that reason, it is not surprising that they are vulnerable to it. In order to overcome the problems of welfare fraud the welfare systems need to be rebuilt and adjusted to the new conditions, more specifically they need to be complemented with better control systems. That the main reason why Sweden is vulnerable to welfare fraud is the lack of control in combination with a trust-based application process is a perception that is shared by many of the informants. As highlighted by a social worker at the unit for incorrect payments in Gothenburg: For almost a year now, it is possible to apply for financial aid digitally or electronically. You don’t even have to come here, which means that you could basically sit abroad and send in your application. You don’t even have to submit any documents, so it’s 100% trust-based. What the informant is saying is that the application process today is very easy and very remote, as highlighted by the researcher at BRÅ, this creates opportunities for cheating: Sweden is a ‘distance land’. We have our Bank IDs with which we sign, and we get the money into your account and then we can use it abroad and so on. Of course, that creates unique opportunities for crime. In addition to the fact that the design of the welfare system makes them easy to cheat several of the informants also emphasize that the agencies in many cases lack the tools they need to conduct efficient controls. The thing that seems to be particularly problematic is the limited ability to collect information from other agencies. Several of the informants argue that the strict secrecy between agencies is a large obstacle to 35 detecting and preventing crime. An illustrative example of this is provided by the operative coordinator at the western police regions fraud unit: The secrecy act limits us from conducting simple investigations. I really believe that the secrecy needs to be more generous. (…) sometimes we can’t get the information we know we need. Say, for example, that we need to locate a person that has committed a crime to bring him in for questioning. We know that this person is living in a home that the social services are responsible for, then we cannot get that information from the social services [because they are not allowed to share that information]. In accordance with the official analysis the informants thus seem to agree that the lack of control is an important contributing factor to Sweden’s problems with welfare fraud. A lack of control is however not a problem in itself. As highlighted by the Economic Crime Authority’s crime prevention specialist “If less resources are devoted to the control systems (…) fewer welfare crimes are discovered and fewer welfare crimes are reported.” In other words, the lack of control is a problem because it prevents the authorities from detecting crime and thus reduces the risk of getting caught, something that in turn, increases the incentives to commit crime. This type of reasoning lies close to the assumptions in ideal type one, which hold that compliance is mainly influenced by the risk of getting caught as well as the severity of the punishments. In addition to the lack of control, several of the informants highlight the lack of quick and tough sanctions as a contributing factor to Sweden’s problems with welfare fraud. The investigator at the Public Employment Service for example argues that: The whole idea with the welfare fraud act, when it came in 2007, was precisely that it should have a norming effect. It was supposed to signal, that it is not okay to cheat the welfare systems. But it hasn’t really had that effect because the police do not have time to prioritize these crimes, so there is no major penalty anyway. A stronger penalty could lead to a stronger, or more deterring effect. What the informant is saying is that more convictions and higher penalties would have a deterring effect, in other words, the modest number of convictions and the low penalties today, encourage welfare fraud. This perception is also close to ideal type one which holds that a lack of sufficient sanctions creates incentives for noncompliant behavior. 36 Another potential cause for Sweden’s problems with welfare fraud that is highlighted by several of the informants is the fact that Swedish has a very generous welfare system. While many of the informants are careful to point out that they believe that a generous welfare system is something positive that should be cherished they also believe that it enables fraud. The re-examiner at the union IF Metall’s unemployment insurance fund for example put it like this: The basic idea of welfare in Sweden looks very different compared to the rest of the world. Our parental insurance is for example much more generous even compared to like Denmark’s. So, on the one hand, there are more situations when there is helping to get, but also more situations where it is like, easy to end up wrong, both concisely and most unconsciously. The coordinator at the western police regions fraud unit puts it more frankly: If we are generous in giving out social benefits, then we will have people who take advantage of it, that’s how it is (…) It will decrease if you reduce the payments, it goes without saying. These quotes highlight a different cause of welfare fraud than the ones presented earlier. They indicate that at least a part of the explanation for why Sweden is exposed to welfare fraud is because Sweden has such generous welfare systems. If you have more benefits, you also have more criminal targets. In other words, that generous welfare system creates more opportunities for crime. This type of reasoning is also in line with ideal type one which holds that noncompliance is in large connected to opportunity structures, where there are opportunities there is crime. Opportunity structures, inadequate controls, and weak sanctions are not the only causes of welfare fraud that the informants point out. As highlighted in the previous sections, some of the informants perceive welfare fraud to be a normative issue and that one of the causes of Sweden’s issues with welfare fraud is that there has been a normative shift. As the controller at the Social Insurance Agency put it “I don’t feel that people always see it as a crime or see it as that serious”. According to the re-examiner at IF Metalls unemployment insurance fund, 37 one reason for the more relaxed attitudes towards welfare fraud could be that people do not understand the severity of welfare fraud, that the system is dependent on their compliance to work: People don’t see the whole picture, that I cannot do this because this is a trust- based system, that it is tax funds, and it affects the collective as a whole (...) that everyone has to contribute for the welfare system to function so that it doesn’t erode. The researcher at BRÅ is on the same track and argues that the attitudes towards welfare crime are probably not a generational issue but rather a relations issue, that one’s attitudes towards welfare crime depend on one’s relationship with the welfare systems: Both the KUT-delegation and ESV have done attitude surveys and they show roughly the same thing, older people perceive welfare fraud to be more serious than younger people, but as the younger gets older (…) their attitudes seem to sharpen. So maybe it has more to do with, like, one’s own relationships with the welfare system than anything else. What she is saying is in other words that if you have had interactions with the welfare state and you have seen the benefits of the welfare system then you will be more skeptical of welfare fraud. It is not primarily about the fact that people do not feel loyally towards the systems as proposed by ideal type two (the model of contingent consent) but rather that you do not consider or understand that you yourself are a part of the system and that the system is dependent on your compliance. Another potential explanation for welfare fraud is put forward by the docent in peace and development studies Hans Abrahamsson: I mean that welfare crime, regardless of who commits them, whether they are criminals or not, it is a lack of the social contract it’s a lack of social belonging. This perspective lies close to ideal type two in which welfare fraud is perceived to be the result of a breach of the vertical social contract between the state and the citizens. As 38 highlighted by ideal type two, the breach of the vertical social contract can be the result of several factors: that the state does not provide the services promised or that process through which policies are decided and implemented is not perceived to be fair. According to the coordinator at the non-profit organization in Gothenburg, Sweden’s problems with welfare fraud are mainly connected to the former: What creates welfare crime? I think you have to look at the dismantling of the welfare to understand this part (…) There are saving requirements and cutbacks in all of the institutions that create a stable and safe society (…) If we as a state cannot ensure that the basic things that we need as humans to survive are satisfied, then I think it is hard to expect people to have faith in the system This line of reasoning goes well in line with ideal type two. What she is saying is that if the state does not deliver what it promised the citizens will have incentives to retract their compliance. From this perspective welfare fraud is not caused by a lack of control but rather a lack of public services. Several informants also highlight the second dimension of ideal type two which hold that noncompliance is caused by a lack of ethical reciprocity. Hans Abrahamsson for example argues that: We have a horizontal inequality. When people start to compare themselves with other social groups, they become convinced that other people are cheating, and then you yourself start to cheat. A similar perspective is offered by the controller at the Social Insurance Agency: One reason [for why welfare fraud is increasing] could be that you think that other people are also doing it [cheating the system]. That’s why it’s so important for us to show that we conduct controls (…) because if it decreases and everyone believes that everyone is cheating the systems then the willingness and trust in the system will decrease. From this perspective, the main cause of welfare fraud is primarily a lack of ethical reciprocity or social trust. The reason why welfare fraud seems to have increased is because people believe that other people are cheating, when people believe that they are the only ones 39 contributing their willingness to contribute declines. If you look closely at the quotes cited above, you realize that the two informants have somewhat different perceptions of why the ethical reciprocity seems to have declined. From Abrahamsson’s perspective, the declining trust in Sweden is connected to the increased horizontal inequality. The investigator at the Social Insurance Agency has a slightly different perspective and connects the lack of ethical reciprocity to the authorities’ ability to exercise control. More specifically he is saying that if the authorities do not show that they are in control, people will believe that cheating is widespread. This perception is well in line with the assumptions of ideal type two, which emphasizes that the state, by sanctioning rule breakers, assures other residents that people can generally be trusted which in turn fosters compliance. In addition to the causes put forward above several of the informants argue that the main cause for welfare fraud is poverty. People commit welfare fraud because they have to in order to survive, to support themselves and their families. As highlighted by the coordinator at the non-profit organization in Gothenburg: It’s about people’s vulnerability, vulnerability creates crime (…) I think that if you had to choose between committing welfare fraud or being able to have a salary, I think most people would have chosen to be able to solve it on their own. From this perspective the primary cause of welfare fraud is poverty. Not everyone agrees that welfare fraud is connected to vulnerable people. As highlighted by the crime prevention specialist at the Economic Crime Authority: Economic crime exists within all groups in society (…) some have very little money and so they want more money, but we also see that some actually have a lot of money but just want more. (…) So, I think it’s super important to remember that the fact that you have little money or are in a vulnerable social situation that can explain a part, but it can’t explain all of it. A similar understanding is offered by the re-examiner at IF Metall’s unemployment insurance fund. She points out that the individuals who turn to the unemployment insurance fund and thus become investigated by the agency are individuals that find themselves in a difficult and stressful situation where they need money. On the other hand, she argues that the perception 40 of economic need can be rather subjective. That the line between need and greed is not always as clear as one might think: People are under more pressure today. They have higher mortgages and higher demand for the standard of living (…) when you’re unemployed, the rug is pulled out from underneath you (…) I don’t think people are really prepared to lower their standard of living as much as can be required. Most people are able to adapt to a new situation, but some don’t (…) they don’t see that if I do this [commit fraud] for 10 days or 2 weeks, it shouldn’t be a big problem because it’s like, yeah but I have to sort out my own desperate financial situation. I see a benefit here and I need this money, I can pay it back later or something like that (…) basically I think that society promotes individualistic thinking and when you are individualistic you also risk becoming egoistic. What the informant emphasizes is that people often perceive that they have a need, that they are forced to commit welfare fraud. People thus commit welfare fraud because they need money. This need is however not necessarily the result of an increased poverty or socioeconomic inequality but rather an increased standard of living or increased materialistic demands. It is not that people can’t survive and commit welfare fraud to buy food, they commit welfare fraud to live up to their own and society’s expectations of how you should live and what you should be able to have or buy. What the informant puts forward is thus some form of capitalist analysis of welfare fraud where the cause of the problem is the increased economic standard, and the increased materialist demands rather than the fact that people have become more vulnerable. 6.3 Solution How a problem is understood and what the main cause of the problem is perceived to be affects which solutions are considered to be suitable. While the analysis has shown that the informants offer several understandings of welfare fraud and point out various potential causes the solutions offered by the informants are largely focused on control. The crime prevention specialist at the Economic Crime Authority for example put it like this: 41 We need to develop the control systems and make sure that those who can control that these people don’t get into the systems do that. We need to increase the knowledge in this area, that this crime exists, and that we need to strengthen the systems around it. So, it’s both the legislation that needs to be adapted to counteract this, but we also need to strengthen the control authorities and municipalities that can act. According to many of the informants, the agency’s ability to conduct controls is highly dependent on its ability to exchange information. It is thus not surprising that many of the informants argue that relaxations of the secrecy act are one of the most important measures in preventing welfare crime. The re-examiner at IF Metalls unemployment insurance fund put it like this: An increased exchange of information as well as a closer collaboration, that’s what I see as the most effective (…) the tighter you work together and the closer to each other, and more transparency, the easier I think it is to access any irregularities. The operative coordinator at the western police regions fraud unit provides a similar view: Above all, it is about secrecy. The secrecy is something that makes our work very difficult, if we had been able to talk openly with for example the issuing authorities, we would have had a completely different situation. Several of the informants also request tougher and quicker sanctions. The investigator at the Public Employment Service put it like this: I mean, we need some kind of change and then I think that a sanction system can be a good system. Precisely to get to the bottom of the problems faster and more efficiently As pointed out by the crime prevention specialist at the Economic Crime Authority: 42 Harder punishments are not an end in itself (…) the important part of investigating and prosecuting crime is actually that it becomes known that these crimes exist and that we need to prevent it. The fact that you are prosecuted for it also has a good preventive effect, that we actually act against these crimes. Both control measures and increased penalties are suggestions that adhere to ideal type one. In accordance with ideal type one, they indicate that the best way to hamper welfare fraud is to implement measures that increase the risk of getting caught as well at the associated cost of being detected. In other words that welfare fraud is an incentive-based problem best solved by measures that reduce the individual incentives to commit crime. While several of the informants argue that greater control is needed and that this will require relaxations of the secrecy act, many of them also acknowledge that this type of measure is not entirely unproblematic. The re-examiner at IF Metalls unemployment insurance fund for example reasoned like this: I believe in more transparency, cooperation, and exchange of information. Individual rights versus how much one should violate personal integrity in order to prevent crime, it’s always a clash (…) It’s not an easy balancing act. Being a member of an unemployment insurance fund is no secret but it’s still sensitive information. So, you still have to have some sort of focus on protecting the individual. The legal practitioner at the Swedish Pensions Agency reflects on the same question but offers a different perspective: If you want to deal with this [welfare crime], you have to accept greater surveillance, and do we want that? Yes, maybe we do. If you have nothing to hide, you’ll have no problem appearing [at the agency] once every year to show that you’re still alive. Unlike the informant from the unemployment insurance fund the legal practitioner argues that while control might be problematic, it is necessary, that the end justifies the means. The quotes thus illustrate that there are two sides to the discussion surrounding control and 43 especially the ability to exchange information. A third perspective on the personal integrity issue is offered by the researcher at BRÅ. From her perspective control may trump personal integrity in some cases but not in others: If someone manages to show that there is a problem, you solve the problem there without realizing that the problem may also exist here and here. Here and here it may be about larger amounts and more serious crime and therefore it’s more reasonable to implement this measure here than here, especially when interest is directed at the individual’s integrity (…) Over the years we’ve felt some concern that large control resources are put on the least serious crimes because they are easier to control. What the informant highlights is that controls may be justified in some cases, but not in others and that it is important to analyze when and where controls extensive controls are justified. As the crime prevention specialist at the Economic Crime Authority put it: It doesn’t always have to be about more controls, it should be expedient controls That extensive control measures will entail an infringement of personal integrity that we may not want is not the only objection against control measures that are lifted by the informants. A fourth perspective on control is offered by the head of the unit for correct payments as ESV who highlights that the changes needed to reap the rewards of a changed secrecy act are rather costly: Most authorities have a kind of digital debt, they are stuck with old systems. They need to replace platforms and change the rules (…) The investments you need to make in order to take advantage of the benefits that a changed secrecy legislation would allow is thus rather large From her perspective control is, therefore: a question of resources, how much do you want to prioritize this and how much does it actually cost? 44 The informant from ESV also highlights another interesting perspective on control, namely that increased control does not only entail a restriction of personal integrity but in many cases also a more difficult application process which may not always be positive. From her perspective we thus have to ask ourselves: What is the purpose [of the welfare system]? And what is it that we want to protect? (…) take housing allowance for pensioners for example. There you need to declare 3 different income concepts I think (…) Many pensioners think that it is too complicated so they don’t apply even if they would be eligible. I mean that this is one such example where it is actually too complicated, and then you lose some of the point [with having the allowance] (…) If something is too complicated to control or perhaps cannot be controlled, then is it a task for which the state should set aside state funds? From her perspective it might be better to ask ourselves: What is the purpose of this allowance, and can it be achieved in some other way? (…) general allowances like child allowance might be better than means-tested allowances. What the informant is stating is thus that it might not be possible to control everything and that the best solution may therefore not be to conduct more controls or put up more obstacles for the applicant but rather to change the structure of the allowances from means-tested to general benefits. While this type of solution is different from the ones generally offered by the official analysis, it is still in line with ideal type one. It is about eliminating the possibilities and thus the incentives to cheat. After all, if you don’t have to apply, you can’t provide incorrect information, in other words, you cannot cheat. According to some informants increased control is not perceived to be a desirable solution at all. On the contrary, they argue that increased controls may in fact be counterproductive and increase rather than decrease the incentives to commit welfare crime. This perspective is for example offered by a field assistant in Växjö who argued that: 45 I don’t think you will accept it [higher controls]. Because there is already a lot of questioning within financial aid. It can take up to 3 months before you get help (…) so if they start questioning things, I really believe that it will create more, what should I say, distrust against the authorities and that you may not want to do right by yourself, simply because you are pissed off. What the field assistant is saying is that people are already dissatisfied with the service provided by the social services, that people believe that it takes too long for the agency to issue the allowances. From the applicant’s perspective, the agencies are thus not in a position where they can increase their demands on the applicants. A similar view is offered by the coordinator at the non-profit organization in Gothenburg, from her perspective extensive control measures would be: Devastating, the trust in the system will decrease drastically. Because you will feel like ‘you have no trust in us so why should we trust you?’ That will be the consequence. Because you won’t sit there and feel like ‘how nice that you control us and everything that we do’, it’s apparent on all levels. (…) and you have to devote incredible resources for this [controls]. And where will that money come from? Well, then there will be further cuts in those sectors of society that allow us to actually move forward and make society better. From her perspective welfare fraud needs to be met by a completely different type of measure: Eradicate the vulnerability. If we can’t change the conditions under which people live and start working with the soft values first it’s not going to work out, it’s not going to get better. As long as we constantly try to treat the symptoms of people’s vulnerability nothing will happen. We need to do a terribly big societal shake-up if we are to solve these problems. The docent in peace and development studies Hans Abrahamsson offers a similar perspective: To fight welfare crime, it’s important that we find a way to create a new local social contract where people feel a sense of belonging and social cohesion 46 These solutions are more in line with ideal type two. From this perspective it is not about hindering people from abusing the systems but rather about giving them a reason to want to contribute. According to the informants who hold this perspective trust is the key to hampering welfare crime. As highlighted by a field assistant in Växjö, trust is largely connected to people’s ability to make a living. If we want people to comply, we must make sure that people can provide for themselves: I think it’s important that you [the state] try to help them find something to do so they can earn some money so that they can reach a level, well that’s a good standard of living. So that’s what’s needed because it’s the need that makes you cheat. A similar view is offered by the decision-making officer at the Migration Agency who states that: I really hope that the government raises the daily allowance instead. Because if they raise the daily allowance to a level that is reasonable, that a person can actually live on, then they might not apply for so many special allowances. These quotes highlight a perspective in line with ideal type four and hold that people only cheat because they have to cheat. From this perspective, the most reasonable solution is thus to make sure that people can make a living because if they can, they will not cheat. 7 Summary and concluding remarks The aim of this thesis was to investigate if the official analysis of welfare fraud is correct or if there are other possible ways of understanding the puzzling development of increasing welfare fraud in Sweden. The thesis was guided by a hypothesis that the reason why Sweden’s efforts to curb welfare fraud seem to have been ineffective is that the Swedish analysis is based on a flawed understanding of welfare fraud. That welfare fraud is not, at least not solely, an incentive-based problem and that measures aimed at reducing the individual incentives to commit welfare fraud will thus never be enough to hamper the negative development. If we want to solve the problem of welfare fraud the deterrence 47 schemes advocated by the official analysis will have to be complemented by other types of measures. As highlighted by the analysis the individuals who work closely with welfare fraud offer a very broad and multifaceted understanding of welfare fraud. From their perspective welfare fraud is a problem of economic incentives but it is also a trust-based problem, a poverty problem, and a problem of inequality. According to some, it may not even be a problem at all. The reason why Sweden is having such a vast problem with welfare fraud is, according to the informants, due to a lack of control systems and weak sanctions which have created a situation where the individual incentives to cheat the welfare systems are high. It is however also the result of a normative shift in society. According to the informants’ people do not perceive welfare fraud to be a serious felony. While the reasons for this may be connected to the design of the welfare systems it may also be the result of something else, of a social development that has made people more individualistic and egoistic or an increased socioeconomic inequality. As pointed out by several of the informants Sweden’s problems with welfare fraud may also, as suggested by Margaret Levi (1997), be the result of a breach in the social contracts on which compliance is built. From this perspective, welfare fraud is not the result of economic incentives or opportunity structure. The reason why people seem to have become more inclined to commit welfare fraud is because the state has failed to provide the citizens with the public services that they need to live a comfortable life, the public services they were promised by the state. As a result, they no longer feel obligated to comply with state demands, after all, if the state does not live up to their part of the bargain, then why should they? In accordance with Levi’s (1997) theory of noncompliance, several of the informants also argue that welfare fraud is, at least in part a problem of ethical reciprocity. From this perspective welfare fraud is the result of a decreasing social trust, people do not believe that other people are complying as a result they choose to not comply themselves. As pointed out by Bacchi (2009) one’s understanding of a problem and its underlying causes determine what you think should be done about it. Considering the broad range of problem descriptions presented by the informants it is thus not surprising that the informants offer a wide variety of solutions. While many of the informants seem to be in favor of the government’s strategy for preventing welfare crime and argue that tougher sanctions and 48 more extensive controls are likely needed, they are also careful to point out that these types of measures are not entirely unproblematic. As highlighted by several of the informants control measures are expensive and will most likely entail an invasion of personal integrity, it is thus important to ask ourselves, does the end justify the means? Are the extensive controls justified considering the size of the problem and the potential negative consequences the efforts may have? After all, we do not know how large of a problem welfare fraud actually is or how much money it generates (FUT-delegationen, 2007; ESV, 2011; SOU 2019:59). According to some informants, welfare fraud may not even be a problem at all. Even if welfare fraud is a problem, it may not be a problem of individual incentives. If it is not and it is instead a problem of decreased trust in the government, control measures might, as pointed out by multiple informants, be counterproductive and increase the distrust that created the problem in the first place. To sum up, this thesis indicates that the official analysis of welfare fraud is correct but only partly and that the policy proposals suggested in the many government investigations are most likely needed but also insufficient. According to the individuals who work directly with welfare fraud, welfare fraud is a complex problem with multiple potential causes. If Sweden are to overcome its issues with welfare fraud it is reasonable to assume that the solutions have to reflect the complexity of the problem and include more measures than just extensive control schemes and tougher punishments. By interviewing individuals with a real insight into welfare fraud and its potential causes, this thesis has generated a deeper understanding of welfare fraud and contributed with new insights regarding the potential causes influencing people’s willingness to comply with state demands. It has also provided multiple potential hypotheses regarding why Sweden is having such vast problems with welfare fraud and what motivates individuals to commit welfare crimes. To determine the degree to which these hypotheses correspond to reality and to what degree they have contributed to Sweden’s development more studies would be needed. An important task for future studies would for example be to investigate: if the attitudes towards welfare fraud have actually become more accepting as indicated by some informants, if people who commit welfare fraud actually do so because they are dissatisfied with the government, and if the typical welfare criminal is a vulnerable citizen or an ordinarily middle- class citizen looking to finance his ‘luxury lifestyle’. Investigating these questions would generate an even deeper knowledge of welfare fraud which in turn would likely enable even 49 more accurate and effective preventive measures, something that is needed to hamper this development. 50 8 References Abrahamsson, H. (2015). Vår tids stora samhällsomdaning. Mistra Urban Futures (Rapport 2015:10). https://www.mistraurbanfutures.org/files/hans_forskningsessa_skl_.pdf Abrahamsson, H. (2012). Städer som ett nav för globalt hållbar samhällsutveckling eller slagfält för sociala konflikter. Malmö Stad. http://cmspres-vir- 2.it.gu.se/digitalAssets/1443/1443822_stader_som_nav_juni_2012.pdf Ali, M. M., Cecil, H. W., & Knoblett, J. A. (2001). The effects of tax rates and enforcement policies on taxpayer compliance: A study of self-employed taxpayers. Atlantic Economic Journal, 29, 186-202. Allingham, M. G., & Sandmo, A. (1972). Income tax evasion: A theoretical analysis. Taxation: critical perspectives on the world economy, 3(1), 323-338. Alm, J., Sanchez, I., & De Juan, A. (1995). Economic and noneconomic factors in tax compliance. KYKLOS-BERNE-, 48, 3-33. Appelgren, L. (2019). Optimal auditing of social benefit fraud: a case study. Empirical Economics, 56, 203–231. Asmar, L., & Rundberg, H. (2016, 10 June). Huvudmannen döms i Kyrkohärvan. SVT. https://www.svt.se/nyheter/lokalt/sodertalje/dom-i-kyrkoharvan Arbetsmiljöverket. (2019) Lägesrapport 2019 – för det myndighetsgemensamma arbetet mot fusk, regelöverträdelser och annan brottslighet i arbetslivet. https://www.av.se/globalassets/filer/nyheter/lagesrapport_2019_det_myndighetsgemensamma _arbetet.pdf Bacchi Lee, C. (2009). Analysing policy: what's the problem represented to be? Pearson. Becker, G. S. (1968). Crime and punishment: An economic approach. Journal of political economy, 76(2), 169-217. https://www.nber.org/system/files/chapters/c3625/c3625.pdf Brå. (2022). Välfärdsbrott mot kommuner och regioner: fel och oegentligheter bland företag och föreningar (Rapport 2022:1). https://bra.se/download/18.161d181f17db3c8d91d4ee7/1648139354068/2022%201%20Valf %C3%A4rdsbrottslighet%20mot%20kommuner%20och%20regioner.pdf Brå. (2016). Bedrägeribrottsligheten i Sverige – kartläggning och åtgärdsförslag (Rapport 2016:9). https://bra.se/download/18.358de3051533ffea5ea2ec64/1458044205141/2016_9_Bedr%C3% A4geribrottsligheten_i_Sverige.pdf 51 Brå. (2015). Intyget som dörröppnare till välfärdsystemet: En rapport om välfärdsbrott med felaktiga intyg (Rapport 2015:8). https://bra.se/download/18.779f51ff14b83989644160a/1427460033099/2015- 8_Intyg_som_dorroppnare_till_valfardssystemet.pdf Bush, E., Carlson, A., & Eklind, H. (2021, 18 May). Låt ny myndighet ansvara för alla välfärdsutbetalningar. Dagens Nyheter. https://www.dn.se/debatt/lat-ny-myndighet-ansvara- for-alla-valfardsutbetalningar/ Button, M., & Tunley, M. (2015). Explaining fraud deviancy attenuation in the United Kingdom. Crime, Law and Social Change, 63, 49–64. Carp, O. (2022, 7 February). Felaktiga utbetalningar av assistans för mångmiljonbelopp. Dagens Nyheter. https://www.dn.se/sverige/felaktiga-utbetalningar-av-assistans-for- mangmiljonbelopp/ Ceasar, J. (2023, 22 Mars). M och S i duell: så ska välfärdsbrotten stoppas. Dagens Nyheter. https://www.dn.se/ekonomi/m-och-s-i-duell-sa-ska-valfardsbrotten-stoppas/ Dean, H., & Melrose, M. (1997). Manageable discord: Fraud and resistance in the social security system. Social Policy & Administration, 31(2), 103–118. Dir. 2021:39. Stärkta möjligheter att bekämpa bidragsbrott för myndigheter som omfattas av bidragsbrottslagen. https://www.regeringen.se/contentassets/16a694287ddc4183898f43b41269a5b7/dir2021_39.p df Dir. 2016:60. En delegation mot överutnyttjande av och felaktiga utbetalningar från välfärdssystemen. https://www.regeringen.se/contentassets/edaf50651f6949d38dd67ba848e032c6/en- delegation-mot-overutnyttjande-av-och-felaktiga-utbetalningar-fran-valfardssystemen-dir.- 201660 Dir. 2005:52. Delegation för åtgärder mot felaktiga utbetalningar från trygghetssystemen. https://www.regeringen.se/contentassets/d89c5b2a54e1455bade1ac7ecbb4ba61/delegation- for-atgarder-mot-felaktiga-utbetalningar-fran-trygghetssystemen/ Dobson, K. (2019). Welfare fraud 2.0? Using big data to surveil, stigmatize, and criminalize the poor. Canadian Journal of Communication, 44(3), 331–342. Ekobrottsmyndigheten. (2016). Ekobrottsmyndighetens lägesbild om ekonomisk brottslighet i Sverige 2016–2018. https://www.ekobrottsmyndigheten.se/wp- content/uploads/2021/06/ekobrottsmyndighetens-lagesbild-2016-2018.pdf 52 Esaiasson, P., Gilljam, M., Oscarsson, H., Towns, A., & Wängnerud, E. (2017). Metodpraktikan. Wolter Kluwer AB. ESV. (2022). Underlag inför 2022 års omfattningsstudier (Rapport 2022:37). https://www.esv.se/contentassets/f88f15625bad43ed92227a65d3c651d0/2021-37-underlag- infor-2022-ars-omfattningsstudier.pdf ESV. (2011). Samverkansuppdrag mot felaktiga utbetalningar från välfärdssystemen 2010 (Rapport 2011:11). Finansdepartementet. (2022). Utbetalningsmyndigheten inrättas för att minska fusk och bidragsbrott i välfärdssystemen. https://www.regeringen.se/pressmeddelanden/2022/12/utbetalningsmyndigheten-inrattas-for- att-minska-fusk-och-bidragsbrott-i-valfardssystemen/ Fischer, C. M., Wartick, M., & Mark, M. M. (1992). Detection probability and taxpayer compliance: a review of the literature. Journal of Accounting Literature, 11(1), 1- 46. Fjeldstad, O. H., & Semboja, J. (2001). Dilemmas of fiscal decentralization: A study of local government taxation in Tanzania. Forum for Development Studies, 27(1), 7-41. https://doi.org/10.1080/08039410.2000.9666122 FUT-Delegationen. (2007). Varför blir det fel? Orsaker till felaktiga utbetalningar från trygghetssystemen (Rapport 5). https://www.regeringen.se/contentassets/c5720cc424934b94ac3286c2a6c3973a/varfor-blir- det-fel-orsaker-till-felaktiga-utbetalningar-fran-trygghetssystemen/ Försäkringskassan. (2021). Försäkringskassans årsredovisning. https://www.forsakringskassan.se/download/18.3a5418591814e228e441633/1662209557701/ forsakringskassans-arsredovisning-2021.pdf Graetz, M. J., & Wilde, L. L. (1985). The economics of tax compliance: fact and fantasy. National Tax Journal, 38(3), 355-363. https://www.journals.uchicago.edu/doi/pdf/10.1086/NTJ41792032?casa_token=GdgIO0rrWd wAAAAA:hkJrgFmA0AajQP14moLeLPvkAo7ODCkWlURuNe5j3qx6f4Rg6760gipAp96Cb 117bkBv3UMEMFCS Grasmick, H. G., & Green, D. E. (1980). Legal punishment, social disapproval, and internalization as inhibitors of illegal behavior. Journal of Criminal Law & Criminology, 71(3), 325-335. Grimes, M. (2006). Organizing consent: The role of procedural fairness in political trust and compliance. European Journal of Political Research, 45(2), 285-315. 53 Gustafson, K. (2008). The criminalization of poverty. Journal of Criminal Law & Criminology, 99(3), 643-716. Holmberg, S., & Rothstein, B. (2020). Social trust – The Nordic gold? https://gupea.ub.gu.se/bitstream/handle/2077/63131/gupea_2077_63131_1.pdf?sequence=1&i sAllowed=y ISF & Brå. (2011). Bidragsbrott och skattebrott: Välfärdens dubbla kriminalitet (rapport 2011:12). https://isf.se/download/18.6e75aae16a591304896caa/1565330433010/Bidragsbrott%20och% 20skattebrott-ISF-Rapport%202011-12.pdf Karklins, R. (2005). The System Made Me Do It: Corruption in Post-Communist Societies. ME Sharpe. Karlsson, R. (2023, 3 Mars). Misstanken – bluffade till sig 900 000 kronor. Norra Halland. https://norrahalland.se/nyheter/misstanken-bluffade-till-sig-900-000-kronor/33882 Kirchler, E., Kogler, C., & Muehlbacher, S. (2014). Cooperative tax compliance: From deterrence to deference. Current Directions in Psychological Science, 23(2), 87-92. Kirchler, E. (2007). The economic psychology of tax behavior. Cambridge, England: Cambridge University Press. Klingberg & Hjort. (2021). Hemtjänstmaffian. Miljonsvindeln som förändrade äldreomsorgen. Atlas. Korsell, L., Hagstedt, J., & Skinnari, J. (2008). Från kelgrisar till styvbarn-fusket med välfärdssystemen. Nordisk tidsskrift for kriminalvidenskab, 95(1). Kumlin, S. (2007). Overloaded or Undermined? European Welfare States in the Face of Performance Dissatisfaction In S. Svallford (Eds.), The Political Sociology of the Welfare State: Institutions, Social Cleavages, and Orientations (pp.80–116). Stanford: Stanford University Press. KUT-delegationen. (2018). Risker för felaktiga utbetalningar från välfärdssystemen (Rapport 1). https://www.regeringen.se/contentassets/25cf60884d0e443a8582c32b56cac9dc/rapport-1- risker-for-felaktiga-utbetalningar-fran-valfardssystemen.pdf Levi. M. (1997). Consent, Dissent, and Patriotism. Cambridge University Press. Levi, M. (1991). Are there limits to rationality? European Journal of Sociology/Archives Européennes de Sociologie, 32(1), 130-141. 54 Levi, M. (1988). Of rule and revenue. University of California Press. Linde, J. (2012). Why feed the hand that bites you? Perceptions of procedural fairness and system support in post‐communist democracies. European Journal of Political Research, 51(3), 410-434. Lundström, R. (2013). Framing fraud: Discourse on benefit cheating in Sweden and the UK. European Journal of Communication, 28(6), 630–645. Lundström, R. (2011). Den kalkylerande medborgaren: bidragsfusk i svensk välfärdsdebatt 1990–2010 (Doctoral dissertation, Sociologiska institutionen, Umeå universitet). https://www.diva-portal.org/smash/get/diva2:468450/FULLTEXT01.pdf Magnusson, E., & Marecek, J. (2015). Doing interview-based qualitative research. Cambridge University Press. Malmström, Å. (2021, 19 april). Tolv myndigheter larmar: Kriminella dränerar välfärdssystemen. Tidningen Näringslivet. https://www.tn.se/ekonomi/10522/tolv- myndigheter-larmar-kriminella-dranerar-valfardssystemen/ Marien, S., & Werner, H. (2019). Fair treatment, fair play? The relationship between fair treatment perceptions, political trust, and compliant and cooperative attitudes cross‐ nationality. European Journal of Political Research, 58(1), 72-95. McKeever, G. (1999). Fighting fraud: An evaluation of the government's social security fraud strategy. The Journal of Social Welfare & Family Law, 21(4), 357-371. McKenna, L. (2022). Translation of research interviews: Do we have a problem with qualitative rigor? Nurse Author & Editor, 32(1), 1–3. Meares, T. L., & Kahan, D. M. (1998). Law and (norms of) order in the inner city. Law & Society Review, 32(4), 805-838. Mungiu-Pippidi, A. (2013). The quest for good governance: How societies develop control of corruption. Cambridge University Press. Mungiu-Pippidi. A. (2006). Corruption: Diagnosis and Treatment. Journal of Democracy 17(3), 86–99. Munro, N., & Duckett, J. (2016). Explaining Public Satisfaction with Healthcare Systems: Findings from a Nationwide Survey in China. Health Expectations 19(3), 654– 66. 55 Ostrom, E. (1998). A Behavioral Approach to the Rational Choice Theory of Collective Action. American Political Science Review, 92(1), 1–22. Paternoster, R., & Simpson, S. (1996). Sanction threats and appeals to morality: Testing a rational choice model of corporate crime. Law and Society Review, 30(3), 549-583. Persson, A., Rothstein, B., & Teorell, J. (2013). Why anticorruption reforms fail—systemic corruption as a collective action problem. Governance, 26(3), 449-471. Persson, A., (2008). The Institutional Sources of Statehood: Assimilation, Multiculturalism, and Taxation in Sub-Saharan Africa (Gothenburg Studies in Politics, 111) [Dissertation, University of Gothenburg]. Polismyndigheten et al. (2021). Myndighetsgemensam lägesbild – organiserad brottslighet, 2021. https://polisen.se/om-polisen/polisens-arbete/organiserad- brottslighet/myndighetsgemensam-satsning-mot-organiserad-brottslighet/ Polismyndigheten. (2019). Myndighetsgemensam lägesbild om organiserad brottslighet 2019. file:///Users/johannawinqvist/Downloads/Myndighetsgemensam%20l%C3%A4gesbild%20om %20organiserad%20brottslighet%202019.pdf Polismyndigheten et al. (2017). Myndighetsgemensam lägesbild om organiserad brottslighet 2016-2017. https://docplayer.se/11969122-2016-2017-myndighetsgemensam-lagesbild-om- grov-organiserad-brottslighet.html Pommerehne, W. W., & Weck-Hannemann, H. (1996). Tax rates, tax administration, and income tax evasion in Switzerland. Public Choice, 88(1), 161–170. Prop. 2018/19:132. Ett starkare skydd för välfärdssystemen. https://www.regeringen.se/contentassets/dcde79c6c214402ba73774cb6752d376/ett-starkare- skydd-for-valfardssystemen-prop.-201819132.pdf Regeringskansliet. (2023, 17 februari). Pressbriefing med justitieminister Gunnar Strömmer [Video]. Youtube. https://www.youtube.com/watch?v=FfckiNCp6W0 Regeringskansliet. (2022, 19 december). Pressträff med finansminister Elisabeth Svantesson 19 december 2022 [Video]. Youtube https://www.youtube.com/watch?v=ZYFc7SFVmKw&t=292s Regev-Messalem, S. (2013). Claiming citizenship: The political dimension of welfare fraud. Law & Social Inquiry, 38(4), 993-1018. Rothstein, B., Samanni, M., & Teorell, J. (2011). Explaining the welfare state: power resources vs. the Quality of Government. European Political Science Review 4(1), 1-28. 56 Rothstein, B., & Stolle, D. (2008). The State and Social Capital: An Institutional Theory of Generalized Trust. Comparative Politics 40(4), 441–59. Sainsbury, R. (2003) Understanding social security fraud. In: J. Millar (ed.) Understanding Social Security: Issues for Policy and Practice. The Policy Press, 277–295. Schneider, S. M., & Popic, T. (2018). Cognitive Determinants of Healthcare Evaluations—a Comparison of Eastern and Western European Countries. Health Policy 122(3), 269–78. SFS 2007:612. Bidragsbrottslag. https://www.riksdagen.se/sv/dokument- lagar/dokument/svensk-forfattningssamling/bidragsbrottslag-2007612_sfs-2007-612 Shonhadji, N., & Maulidi, A. (2021). The roles of whistleblowing system and fraud awareness as financial statement fraud deterrent. International Journal of Ethics and Systems. SKR. (2021). Välfärdsbrott och otillåten påverkan – från bidragsfusk till systemhotande brottslighet. https://skr.se/download/18.583b3b0c17e40e303846fdce/1642754384149/7585- 953-8.pdf SOU 2023:8. Arbetslivskriminalitet: arbetet i Sverige, en bedömning av omfattningen, lärdomar från Danmark och Finland. https://data.riksdagen.se/fil/972289E7-C8A5-48CB- B575-E5676994F96C SOU 2022:37. Stärkt arbete med att bekämpa Bidragsbrott: Administrativt sanktionssystem och effektivare hantering av misstänkta brott. https://www.regeringen.se/contentassets/019c22f771544f64bad7842d5617ee6b/sou- 2022_37.pdf SOU 2020:35. Kontroll för ökad tilltro - en ny myndighet för att förebygga, förhindra och upptäcka felaktiga utbetalningar från välfärdssystemen. https://www.regeringen.se/contentassets/d62c8a6857ab49e4b21111009897f819/kontroll-for- okad-tilltro--en-ny-myndighet-for-att-forebygga-forhindra-och-upptacka-felaktiga- utbetalningar-fran-valfardssystemen-sou-202035.pdf SOU 2019:59. Samlade åtgärder för korrekta utbetalningar från välfärdsystemen. https://www.regeringen.se/contentassets/25cf60884d0e443a8582c32b56cac9dc/samlade- atgarder-for-korrekta-utbetalningar-fran-valfardssystemen-sou-201959.pdf SOU 2018:14. Bidragsbrott och underrättelseskyldighet vid felaktiga utbetalningar från välfärdssystemen – en utvärdering. https://www.regeringen.se/contentassets/bdebbfbfc706476eb1169ee4d6dacec5/bidragsbrott- och-underrattelseskyldighet-vid-felaktiga-utbetalningar-fran-valfardssystemen--en- utvardering-sou-201814-.pdf 57 SOU 2017:37. Kvalificerad välfärdsbrottslighet – förebygga, förhindra, upptäcka och beivra. https://www.regeringen.se/contentassets/8df6c84af7a549f6be4326e0114b2ce6/kvalificerad- valfardsbrottslighet--forebygga-forhindra-upptacka-och-beivra-sou-201737/ SOU 2008:74. Rätt och riktigt – Åtgärder mot felaktiga utbetalningar från välfärdssystemen. https://www.regeringen.se/contentassets/e61958f846134c3b84e83b53d7591e23/sou-200874- ratt-och-riktigt---atgarder-mot-felaktiga-utbetalningar-fran-valfardssystemen/ Strümpel, B. (1969). The contribution of survey research to public finance. In A. T. Peacock (Ed.), Quantitative Analysis in Public Finance (pp. 13-22). New York: Praeger. Svenska Dagbladet. (2017, 5 May). Utredning: Kriminella hotar välfärdssystemen. Svenska Dagbladet. https://www.svd.se/a/or8m7/utredning-kriminella-hotar-valfardssystem SVT Nyheter. (2022, 1 June). SD vill ha bidragsstopp för dömda fuskare. SVT Nyheter. https://www.svt.se/nyheter/inrikes/sd-vill-ha-bidragsstopp-for-domda-fuskare SVT Nyheter. (2021, 6 September). Ny myndighet ska motverka välfärdsbrott. SVT Nyheter. https://www.svt.se/nyheter/inrikes/ny-myndighet-ska-motverka-valfardsbrott Thisted Dinesen, P., & Mannerman Sønderskov, K. (2021). Quality of Government and Social trust In A. Bågenholm, M. Bauhr, M. Grimes, & B. Rothstein (Eds.), The oxford handbook on Quality of Government. Oxford University Press. Tracy, S. J. (2010) Qualitative Quality: Eight ”Big Tent” Criteria for Excellent Qualitative Research. Qualitative Inquiry, 16(10), 837-851. Transparency International. (2022). Corruption perception index. Gathered 2023-02-22 from https://www.transparency.org/en/cpi/2022 Tunley, M. (2011). Need, greed, or opportunity? An examination of who commits benefit fraud and why they do it. Security Journal, 24, 302-319. Webley, P., Robben, H. S. J., Elffers, H., & Hessing, D. J. (1991). Tax evasion: An experimental approach. Cambridge: Cambridge University Press. Wikén, J. (2022, 17 Maj). Här är granskningen om assistansfusk – i korthet. SVT. https://www.svt.se/nyheter/inrikes/har-ar-granskning-om-assistansfusket-i-korthet World Bank. (2021). Worldwide Governance Indicators. Gathered 2023-02-22 from http://info.worldbank.org/governance/wgi/Home/Reports 58 Zhang, N., Andrighetto, G., Ottone, S., Ponzano, F., & Steinmo, S. (2016). "Willing to pay?" Tax compliance in Britain and Italy: An experimental analysis. PLoS One, 11(2). doi:10.1371/ journal.pone.0150277 Yaniv, G. (1997). Welfare fraud and welfare stigma. Journal of Economic Psychology, 18(4), 435–451. 59 9 Appendix 9.1. List of informants Workplace/Agency Title 1 IF Metalls unemployment insurance fund Re-examiner If Metalls A-kassa 2 Non-profit organization in Gothenburg – operating Operations Coordinator in some of Gothenburg’s vulnerable neighborhoods Ideell förening i Göteborg med verksamhet i några av stadens utsatta områden 3 The Economic Crime authority Crime prevention specialist Ekobrottsmyndigheten 4 The Municipality of Gothenburg (Social services) Social worker at the unit for Göteborgs stad (Socialförvaltningen) incorrect payments in Gothenburg 5 The Municipality of Gothenburg (Social services) Social worker at the unit for Göteborgs Stad (Socialförvaltningen) incorrect payments in Gothenburg 6 The Municipality of Växjö Field Assistant* Växjö kommun 7 The Municipality of Växjö Field Assistant Växjö kommun 8 The Municipality of Växjö Field Assistant Växjö kommun 9 The Police Authority (Fraud Unit in Gothenburg) Operative coordinator Polismyndigheten (Bedrägerienheten I Göteborg) 10 The Police authority (police district Växjö/Alvesta) Area police (local police officer Polismyndigheten (Lokalpolisområde focused on preventive work) Växjö/Alvesta) 11 The Public Employment Service Investigator Arbetsförmedlingen 12 The Social Insurance Agency Controller Försäkringskassan 13 The Swedish Board of Student Finance (CSN) Controller 60 Centrala studiestödsnämnden (CSN) 14 The Swedish Migration Agency Decision-making officer Migrationsverket 15 The Swedish National Council for Crime Researcher Prevention Brottsförebyggande rådet 16 The Swedish National Financial Management Head of the unit for correct Authority (ESV) payments Ekonomistyrningsverket 17 The Swedish pensions agency Legal practitioner Pensionsmyndigheten 18 University of Gothenburg (The School of Global Docent in peace and Studies) development studies Hans Göteborgs universitet (Institutionen för globala Abrahamsson studier) * Fields assistants are a common profession in municipalities. Their job is mainly to move in places with a lot of people, especially youths, to build relationships. They often work as a bridge between the residents in socioeconomically vulnerable areas and the municipalities. 61 9.2. Interview guide 1 (public employees)4 The interview will focus on the problems with welfare crime in Sweden. Welfare crime is a broad term, in this setting welfare crime mainly refers to all incorrect payments that are or are suspected of being the result of intentional actions by the applicants. It thus includes both systematic and organized fraud, but also smaller intentional errors made by individuals once or over a longer period. Opening questions • How long have you worked for this organization/agency? • What is your position in the organization/agency? • Did you have any other positions before this? o A) If yes: where did you work before this? • How come you started working with these issues? • Can you describe your daily work? Theme 1: The organization's preventive work • Can you describe the welfare fraud aimed at your organization/agency? o How do the problems manifest? • How does your organization/agency work to prevent welfare fraud? o Can you give any concrete examples? • Do you feel that the preventive work has changed in recent years? o A) if yes, how has it changed? o B) is there anything in particular that has gained more attention today compared to previously? o C) What do you think about that? (Do you believe it is positive, why/why not)? • Do you think anything is missing in your preventive work as it is performed today? Is there anything you think should be changed or added? o A) if yes, what? o What do you think that would have contributed to/added? 4 A version of this interview guide was used when interviewing the informants working directly with welfare fraud. 62 Theme 2: the Central government’s preventive efforts In the public debate and according to the government agencies the problems with welfare crime in Sweden are extensive. The problems have also been described as increasing. • Why do you think Sweden has such extensive problems with welfare fraud? • Why do you think that the problems with welfare fraud have increased? • How would you describe the problem with welfare crime? • What type of problem is welfare crime? (What is it an expression of or for?) • Why do you think people commit welfare fraud? In the public debate welfare fraud is largely connected to organized crime. • How do you perceive the connection between organized crime and welfare fraud? • Do you see that the welfare crimes directed at your agency are connected to organized crime? o A) if yes, in what way? What does the connection look like? o B) can you give any examples? • What proportion of the welfare crime directed at your agency would you say is committed by or connected to organized crime? • Based on the cases you come across, would you say that there is a specific type of person who commits welfare fraud? o A) if yes, how would you describe that person? • Would you say that welfare crime mainly is organized crime? • Can welfare crime be perceived as something else than organized crime? Theme 3: the respondent’s own perception of the problem with welfare crime • What do you think we need to do in order to overcome the problems with welfare fraud? • What do you think about harder punishments as a measure to prevent welfare fraud? o A) if positive effect, in what way do you think such a measure would contribute to preventing welfare fraud? o B) if no effect, why do you think that the measure will be ineffective? 63 o C) if negative effect, what negative effects do you think such a measure will have and why? • What do you think about increased opportunities for government authorities to control information and payments as a measure to prevent welfare fraud? o A) if positive effect, in what way do you think such a measure would contribute to preventing welfare fraud? o B) if no effect, why do you think that the measure will be ineffective? o C) if negative effect, what negative effects do you think such a measure will have and why? • What do you think about ‘contribution ceiling’ or similar measures for preventing welfare fraud? o A) if positive effect, in what way do you think such a measure would contribute to preventing welfare fraud? o B) if no effect, why do you think that the measure will be ineffective? o C) if negative effect, what negative effects do you think such a measure will have and why? Theme 4: The respondent’s own perception of the solution to the problem of welfare crime Welfare crime has been on the political agenda for a long time. During the past 20 years, both the government and the government agencies have taken a number of measures to increase the agencies’ ability to detect and prevent these crimes. The government has for example introduced a welfare fraud act; several cooperation projects have been launched in order to enable the agencies to exchange information. Multiple measures have also been taken in order to make it easier for the applicants to provide correct information. The government agencies have, furthermore, been given a clearer mission to prevent fraud and several agencies have gotten an increased mandate to issue controls. Despite all of these measures the problems with welfare fraud seem to remain at a similar level as previously. • Why do you think that is? According to the evaluations made many of the measures taken seem to have been effective. More crimes are discovered, reported to the police, and prosecuted, yet the problems remain. • What is your reflection on that? 64 Closing questions • Is there anything you would like to add? 9.3. Interview guide 2 (additional informants)5 The interview will focus on the problems with welfare crime in Sweden. Welfare crime is a broad term, in this setting welfare crime mainly refers to all incorrect payments that are or are suspected of being the result of intentional actions by the applicants. It thus includes both systematic and organized fraud, but also smaller intentional errors made by individuals once or over a longer period of time. Opening questions • How long have you worked for this organization? • What is your position in the organization? • Did you have any other positions before this? o A) If yes: where did you work before this? • How come you started working with these issues? • Can you describe your daily work? Theme 1: The residents’ relationship with and view of government agencies • How would you say that the residents in X (the area where you work) see government agencies? • Would you say that the residents generally trust the agencies? • Do people believe that the agencies do a good job? Do the agencies live up to the residents’ expectations and deliver the services they are supposed to? • Do the residents generally believe that they are being treated fairly by agencies such as the social services, the Social Insurance Agency, and so on? The social services have been subject to a lot of debate recently. From a national level, it is believed that there is a disinformation campaign against the social services where 5 A version of this interview guide was used when interviewing informant 2, 6, 7, 8 and 18. These are the informants who were selected based on their proximity to the residents in vulnerable areas and their knowledge regarding trust in government institutions and attitudes towards welfare crime in these areas. 65 international forces spread false information about the social services kidnapping children, and so on. This information is assumed to have taken hold in many of the suburbs. It is believed that the residents in these areas do not trust the social services. • Would you say that the residents here are skeptical of the social services? o A) If yes, what is it that people express? What is the main reason? Theme 2: The residents' view of the opportunities to make a living • Is financials, that is, how one should be able to make a living a common topic of conversation among the people you meet? o A) if yes, in what way? What do people generally say? • Do you think many people find it hard to get their finances together, to put food on the table so to speak? o A) If yes, why? • Do you hear people talking about how social benefits (for example financial aid and sick pay) are too low to live on? • Do many people talk about the difficulty of getting a job? o A) If yes, what do people say is the biggest obstacle to getting a job? Theme 3: Welfare fraud • Is there a lot of talk about welfare crime and benefit fraud? o A) If yes, what do people say? • Do you know if welfare fraud is common? o If yes, how and what type of welfare fraud do you believe is common? There are those who believe that the acceptance of welfare fraud is higher in vulnerable areas than compared to others. It is for example not uncommon to hear that people talk rather openly about how they cheat the welfare system. That your stupid if you do not do it. • Is this anything you’ve heard? o If so, where did the information come from? (The residents or some other source). • Do you think that this image is correct? 66 • Why do you think people commit welfare fraud? • Do you think welfare fraud is more common in vulnerable areas than in other types of areas? o A) If yes, why do you think that is? o B) If no, why do you not believe that that is the case? Theme 4: Measures to prevent welfare fraud • What measures do you think are most effective in preventing welfare fraud? • What do you think about harder punishments as a measure to prevent welfare fraud? o A) if positive effect, in what way do you think such a measure would contribute to preventing welfare fraud? o B) if no effect, why do you think that the measure will be ineffective? o C) if negative effect, what negative effects do you think such a measure will have and why? • What do you think about increased opportunities for government authorities to control information and payments as a measure to prevent welfare fraud? o A) if positive effect, in what way do you think such a measure would contribute to preventing welfare fraud? o B) if no effect, why do you think that the measure will be ineffective? o C) if negative effect, what negative effects do you think such a measure will have and why? • What do you think about ‘contribution ceiling’ or similar measures for preventing welfare fraud? o A) if positive effect, in what way do you think such a measure would contribute to preventing welfare fraud? o B) if no effect, why do you think that the measure will be ineffective? o C) if negative effect, what negative effects do you think such a measure will have and why? Closing questions • Is there anything you would like to add? 67