Willingness to pay taxes through mutual trust: The effect of fairness, governability, tax-enforcement and outsourcing on local tax collection rates

This study extends our understanding about the inter-relationship between taxation policy, democracy and financial performance in local governments. Every government, particularly local ones, would like to make their residents more willing to pay taxes. Based on the trend of making local authorities less reliant on national budgets and the ever-increasing expectations that they provide high quality services, we investigated how public perceptions about the local government's fairness, tax enforcement, governability and the out-sourcing of local tax collection affect the percentage of local property taxes collected. We used a survey of 607 Israeli residents randomly sampled from municipalities that do and do not outsource, and independent data about tax collection rates. The results indicate that perceptions about tax enforcement and governability mediate the relationship between perceptions about fairness and tax collection rates. However, only within local authorities that outsourced tax collection was there a positive relationship between perceptions about tax enforcement and tax collection rates and a positive relationship between perceptions about governability and tax collection rates. Implications of the findings

are discussed in the era of local democracy and local governance.

| INTRODUCTION
Residents do not like to pay local taxes.This reluctance should not be a surprise for any local politician, manager, resident or scholar.Even though the local authorities are closer to their people and theoretically can more accurately identify their needs, residents are still unlikely to want to pay local taxes.Thus, the challenge of the local leadership and administration is to maximize tax revenues without losing the trust and support of local residents.This challenge epitomizes the connection between local politics, local tax policies and the local democratic climate.We call this political challenge the willingness to pay taxes through mutual trust.
Local authorities directly affect residents' quality of life, as they provide a range of essential services (Fjeldstad, 2001).As such, researchers have repeatedly examined the relationship between residents' opinions and municipalities' performance.Local authorities have experienced two major changes in government-New Public Management in the 1980s and a shift from local government to local governance in the 2000s (Stoker, 1998).Yet, the outcomes of these changes are still questionable (e.g., Koch, 2013).The ongoing pursuit of better performance and better financial management has led national governments to push local authorities to meet ever-increasing demands and expectations independently (Torgler, 2005).They are now expected to co-produce and co-create public value, offer diverse services and deal with economic, environmental and community development issues (Cornforth & Paton, 2004;Osborne, 2010Osborne, , 2018)).
One of the major side effects of the New Public Management reformssome would say a cure and others would say an illnessis privatization and one of its operationsoutsourcingand its effect on mutual trust.Outsourcing has been touted as a tool for making government more efficient (Bel et al., 2007;Wassenaar et al., 2010).While its effect is still questionable, internal stakeholdersparticularly the local leadershipregard it as an important factor in local performance.
We consider both external and internal factors in the success in collecting taxes.We regard outsourcing as an external factor, and public opinion about the local government's fairness, governability and tax collection rates as internal factors.We assume that local authorities cannot provide all local services and products alone at a satisfactory level.Therefore, they tend to rely on external providers of local services, prompting them to outsource (Fjeldstad, 2001).Nevertheless, they take into account public opinion regarding governance, democracy and taxation.Modern outsourcing emphasizes the inter-organizational relationships and partnerships of multiple actors and sectors that collaborate rather than compete.It also emphasizes collaboration with the public rather than a top-down, one sided, technocratic method of service provision.Therefore, based on Haus and Sweeting's (2006a) typology of local democracy, we maintain that outsourcing in the era of local network governance also has the potential to advance local democracy, enhance mutual trust, fairness, legitimacy, governability, and local performance.
Balancing democracy and bureaucracy and utilizing local outsourcing in local authorities is delicate and depends heavily on the managerial qualities of the local leadership and its relationship with the public (Mgonja & Poncian, 2019).As Basoli (2010) stressed, local governance arrangements may affect local democracy, both positively and negatively.Local collaborations may enhance the relationship between institutions and individuals.At the same time, they might involve conflicts and tensions that increase the imbalance of power between the partners (Grimshaw et al., 2002), complicate their interaction and reduce local performance (Klijn & Edelenbos, 2008;Sørensen & Torfing, 2008).In the era of shifting from new public management to local network governance, it is still questionable how the public's views on local performance and outsourced services affect the relationship between local democracy and local governability.
We explore how public opinion about the local government's fairness, governability and tax collection rates and their interrelations impact the success of local authorities in collecting taxes.Thus, this study extends our understanding about the interrelationship between local taxation policy, local democracy and local financial performance and answers the call to deepen the explicit discussion of the causal effects of outsourcing mechanisms on local democracy and performance related to local taxation (e.g., Geys & Sørensen, 2016).Our main research question is, how do public perceptions about the local government's fairness, tax enforcement governability and outsourcing of local tax collection affect tax collection rates?
To answer this question empirically, we used a quantitative research design.We avoided common source bias by using (i) a survey of 607 Israeli residents randomly sampled from 12 municipalities that do and do not outsource who reflected on local democracy and local governability, and (ii) independent data about tax collection rates supplied by the Israeli Ministry of the Interior Ministry and the Israeli Bureau of Statistics that reflect the financial performance of local authorities.

| LITERATURE REVIEW AND HYPOTHESES
It goes without saying that local authorities need adequate and sustained revenue flows, which come both from the central government and money that the authorities raise themselves (Fjeldstad, 2001).The property tax collection rate is the percentage of property taxes that a municipality collects from residents and businesses in its jurisdiction out of the potential taxes.Compliance with the payment of property taxes is important because it affects many aspects of local life.The funds it provides give local leaders greater flexibility in managing the quality and quantity of services provided to all.These taxes also shape the decisions of residents and businesses about where to locate.Finally, they allow local leaders to make investments in the community and the local economy.Hence, local authorities have a dual, some may say contradictory, role of collectors and planners (Fainstein, 1991).
Nevertheless, what factors prompt residents to pay or avoid paying local property taxes?We maintain that there are internal and external factors that promote or deter people from complying with the laws in this area.

| Feelings of fairness and tax collection rates
One of the major approaches to examining property tax collection rates is the attitudinalbehavioral approach, meaning the views, feelings, and actions of residents.One attitudinalbehavioral element is fiscal exchange, a contractual relationship between ratepayers and the local government.The basis of this exchange is people's sense that the local government is trustworthy and provides reasonable service to residents.This approach, sometimes called tax morale, deals with the intrinsic motivation to pay taxes.However, it correlates with extrinsic factors (Cullis & Lewis, 1997;Filippin et al., 2013;Fjeldstad, 2004;Haj, 2007;Levi, 1998;OECD, 2013;Schnellenbach, 2006;Torgler, 2005;Torgler & Schneider, 2007;Vigoda, 2000;Zeedan, 2017).
The residents' sense that the local government is fair means they believe that its decisions are effective, efficient, and ethical (Hassan et al., 2014).Levi (1988) suggested that residents trust the government if the government is regarded as fair.Thus, when decisions and information are transparent and accessible, residents tend to see it as procedurally fair.Alongside municipal decisions, activities such as the rational and fair use of the municipal budget and high quality local services promote the sense that the local government is fair.Such feelings may result in residents' assumptions that the local bodies feel responsible for them, increasing a sense of mutual trust and the willingness to pay taxes (Beeri et al., 2019;Fjeldstad, 2004;Torgler & Schneider, 2007).Furthermore, Schnellenbach ( 2006) associated fairness and legitimacy with tax compliance.According to this argument, people have an intrinsic motivation to fulfill their obligations such as participation in collective decision-making and paying taxes.In return they expect fair treatment from the authorities.
These relationships also accord with Fox and Miller's (1995) loop model of democracy.According to this model, democracy starts with citizens' needs and preferences, which influence the policies of leaders and institutions.Successful government bodies manage reality in a way that increases perceptions that they are trustworthy and fair, on the basis of which residents judge their performance and reward or punish them in the next electoral round (Reingewertz & Beeri, 2018).Moreover, policies that rely solely on deterrence and punishment to obtain taxes are often problematic and even counterproductive (Leonardo, 2011).In this sense, fulfilling civil duties such as paying taxes may be a way of rewarding local leaders for positive feeling of fairness, whereas evading taxes is the punishment for feelings of unfairness.On the other hand, local tax evasion and refusals to pay are more likely when the local authority does not meet a certain level of minimum standards and when residents feel that the municipality is unfair and untruthful.This sense might be reflected in unreasonable expenses and local debt, which will probably result in poor financial performance that, in a vicious cycle, will reduce residents' sense of fairness, leading to lower tax collection rates (Wintrobe, 2001).
Therefore, the baseline of our model (see Figure 1) and our first hypothesis posit that: Hypothesis 1.The more residents feel that the local government is fair, the higher the local tax collection rates.
2.2 | Perceptions about tax enforcement and governability as mediators of the relationship between perceptions about fairness and tax collection rates By perceptions about tax enforcement, we mean the residents' views of the local authority's efforts and success in enforcing tax collection (Haj, 2007;Scholz, 1998).By perceptions about governability, we mean the residents' views of the local authority's ability to govern, to make decisions and carry them out, or, in other words, to design local policy and implement it.In our study, perceptions about governability reflect the extent to which residents see the local authority as capable of designing and implementing local tax policy.Two other elements of the attitudinal-behavioral approach related to perceptions about tax enforcement and governability are coercion and social norms.Coercion refers to the enforcement activities and penalties imposed on non-payers and the public's assessment about the likelihood of being caught or fined.According to Allingham and Sandmo (1972), people will avoid paying taxes as long as the benefits of tax evasion outweigh the costs of being caught.
Social norms in our context are the norms that shape compliance behavior and the normative commitment to comply with the law (Fjeldstad, 2004;Haj, 2007;Levi, 1998;OECD, 2013;Traxler, 2010;Vigoda, 2000;Zeedan, 2017).This element is also in line with the theory of planned behavior, according to which tax compliance is the result of attitudes, subjective norms, and perceived behavioral controls such as tax morale, the fairness of the tax, trust in government and perceptions about the power of the authorities (Taing & Chang, 2021).
The relationship between perceptions about fairness and tax collection rates is based on two assumptions, both related to good governance, governability and democracy.First, that it is the local authority's duty to collect taxes in order to guarantee the supply of quality local services.Second, that if needed, the local authority has the power and legitimacy to punish tax evaders.To accomplish these goals, the local authority is expected to operate an efficient system that maximizes tax collection with minimal investment in tax enforcement.It is expected to design and implement a policy and techniques of local budgeting that increases the amount of its revenues and resources (Fjeldstad, 2001).
However, the rationale of an efficient tax system does not necessarily go hand in hand with the interests of taxpayers.Allingham and Sandmo (1972), who created a model that explains tax evasion, stressed that residents do not pay taxes voluntarily but follow an alternative rationale involving cost-benefit considerations about doing so.In contrast to the attitudinal-behavioral approach, this economic approach assumes that people are rational and that this rationality is the basis of the tax evasion theory.According to this theory, people do not want to pay taxes.The only reason they do so is out of fear of being caught (Allingham & Sandmo, 1972;Brunori, 2007;Filippin et al., 2013;Savas, 1987).If the municipality does not take strict measures against tax evaders, taxpayers are more likely to feel that the system is unfairly burdening them.Therefore, they are more likely to free ride, to evade tax duties while enjoying the services that are provided to all.Thus, the willingness of taxpayers to pay taxes is a result of their assessment of the system as fair.This assessment, in turn, leads to perceptions of the extent to which the local authority enforces tax collection, perceptions that others are also paying taxes and calculations about the chances of being punished if they do not pay.This calculation takes into consideration the efforts made by tax authorities to prevent evasion and the risk that tax authorities will employ coercive measures beyond warning letters to debtors.For example, tax authorities may impose economic sanctions such as fines, foreclosures of bank accounts and private property and even jail on those who evade taxes (Allingham & Sandmo, 1972;Fjeldstad, 2001;Haj, 2007).
Taking this argument one step further, one should bear in mind that the relationship between feelings of fairness and local taxes does not end with a simple calculation of threats and risks.Additional factors that were discussed earlier and that are related to perceptions about local democracy, governability and performance play a role in deciding whether to convert a certain level of feeling of fairness on the part of the local authority into the payment or non-payment of local taxes (Levi, 1988).These factors may include the residents' perceptions about local governability.Thus, when people regard a political leader as fair, and as a legitimate governor and regulator, he or she can enforce social norms and civil duties such as paying local taxes more easily and efficiently (Fjelsdstad, 2004).Taking all of these factors into account, we can conclude that a local authority that treats its community members fairly is likely to increase the residents' beliefs that the local authority is quite capable of designing and implementing a tax policy and can enforce it.This assessment increases the taxpayers' tendency to pay local taxes.The opposite is also true.If residents feel that the local government is unfair, cannot govern effectively, and cannot create a tax policy that it can enforce, they are more likely to avoid paying local taxes (Cheibub, 1998;Filippin et al., 2013;Haj, 2007;Taing & Chang, 2021;Talit, 2012).
Accordingly, our mediation hypotheses state that: Hypothesis 2. The more residents feel that the local government is fair, the more likely they are to believe that tax enforcement is efficient, which, in turn, leads to higher local tax collection rates.
Hypothesis 3. The more residents feel that the local government is fair, the more likely they are to believe that the local government can govern effectively, which, in turn, leads to higher local tax collection rates.

| Outsourcing of tax collection as a moderator of the relationship between perceptions about tax enforcement, perceptions about governability and tax collection rates
Privatization is defined as the act of reducing the role of government or increasing the role of the private sector in an activity or in the ownership of assets (Savas, 1987).Privatization has unambiguous political origins and objectives, which is the effort to withdraw, at least to some extent, from the public sphere and thus shrink government.It may take many shapes, from letting public services decay through deregulation, to the active transfer of public responsibility, authority, processes or property to private and NGO hands (Ramamurti, 2000;Starr, 1988).
Our study focuses on one specific technique of privatization, outsourcing, sometimes called contracting-out.Privatization and outsourcing are not identical.Privatization refers to the general transfer of the ownership of assets from public to private ownership and the importing of private values and managerial mechanisms into public organizations.Within this study's context, outsourcing is one possible operationalization of privatization.However, outsourcing is not reserved exclusively for public organizations, as all sectors may outsource activities to all sectors.Outsourcing involves opening up a set of activities that were previously immune from competition, usually through bids for contracts to provide particular services (Domberger & Jensen, 1997).
In our context outsourcing results in a private company, which has been asked to or won a contract to provide local services, making a profit from doing so (Domberger & Jensen, 1997;Warner & Bel, 2008).Municipalities have made extensive use of outsourcing for various services (Boyne, 1998;Kakabadse & Kakabadse, 2001), a leading one of which is tax collection (Mgonja & Poncian, 2019;Talit, 2012).The local authority is still responsible for funding, regulating and supervising such services.However, taxation is a delicate policy action that potentially conflicts with privacy and proprietary rights (Bloch & Demange, 2018).Therefore, the outsourcing of local tax collection leaves a great deal of space for politics, governance and democracy to influence it and to be influenced by it.
Scholars have reported mixed evidence regarding the costs and benefits of outsourcing (e.g., Boyne, 1998;Brudney et al., 2005).This debate is usually between those concerned about public policy and democracy and those who focus on economics (Domberger & Jensen, 1997).Opponents of outsourcing have asserted that outsourced tax collection simply does not achieve its main goals.Due to transaction cost economics, the limits to competition or contract failures (Crouch, 2015;Domberger & Jensen, 1997), it does not reduce costs or increase efficiency as expected.Furthermore, it does not automatically increase tax collection rates (e.g., Talit, 2012;Yusuf & O'Connell, 2014).Critics concerned about public policy and democracy maintain that outsourcing limits governmental control, leading to negative effects on employment, the fragmentation of the delivery of public services, the displacement of the traditional coordination between governmental arms, and the relegation of local policy makers to public service auditors and inspectors (Brown et al., 2006;Martin, 2010;Milward & Provan, 2000).Other scholars claimed that it hurts performance and results in poorer quality local services (Boyne, 1998;Joassart-Marcelli & Musso, 2005;Pollitt & Bouckaert, 2004), or even destroys them completely (Geys & Sørensen, 2016).Another criticism of outsourcing is that it bypasses local government, stripping away its functions, and limiting its autonomy (Martin, 2010).Thus, it may erode local mutual trust and the public's sense of the local authority's integrity, transparency, and determination to treat all of its residents equally, which, in turn, increases negative attitudes about the local authority (Rosenbloom & Piotrowski, 2005).
Moreover, the outsourcing of tax collection may lead to corruption, increased tariffs that shift tax burdens to residents, and losses in organizational autonomy, flexibility, knowledge, technology, and the ability to recognize the community's needs (Mgonja & Poncian, 2019).Critics maintain that the outsourcing of tax collection will undermine local accountability and governability (Domberger & Jensen, 1997;Kakabadse & Kakabadse, 2001;Mgonja & Poncian, 2019;Starr, 1988;Vigoda-Gadot et al., 2014).Locally elected officials who wield power are held publicly accountable to the residents for their decisions and performance.However, outsourcing may detach responsibility from accountability and transfer responsibility for and control over tax policy that were solely in the hands of elected leaders to a for-profit company.As a result, responsibility is transferred into the hands of a third party, divorcing operational functions from fiscal accountability and duties (Domberger & Jensen, 1997;Starr, 1988).In addition, private tax collectors who seek to maximize their profits tend to employ disproportionate tools to collect debts.Thus, the erosion of local governability might weaken the perceived fairness and transparency of the municipality, and in turn, increase negative feelings among the residents, invade their privacy and fracture the norms of integrity, compliance and equality (Brown et al., 2006;Milward & Provan, 2000;Rosenbloom & Piotrowski, 2005).
In contrast, there are strong arguments about the benefits of outsourcing.Those who point to the potential benefits of outsourcing provide a mirror image of the logic described above.From this point of view, outsourcing is not about abandoning the public sphere.It is a tool that seeks to create new kinds of market relations that are superior to conventional public administration (Ramamurti, 2000;Starr, 1988).According to the public choice school, the public sector is irredeemably inefficient and public enterprises necessarily perform less efficiently than private enterprises.Therefore, the public would prefer to limit the increasing expenditures and excessive budgets that serve only self-interested coalitions of voters, politicians, and bureaucrats (Starr, 1988;Tang, 1997).Outsourcing can eliminate monopolies and the expenses for services and improve the quality of those services (Fjeldstad, 2001;Zafra et al., 2014).Those who favor the outsourcing of tax collection aver that it increases efficiency and effectiveness because it results in economic development, enables flexibility in operation, allows a quick response to dynamic changes and transforms the immediate incentive of high collection rates into large profits (Pollitt & Bouckaert, 2004).Furthermore, the outsourcing of tax collection may improve local governability and perceptions about local governability.This claim relies on several arguments.First, that it allows local leaders to focus on decision-making rather than technical actions.Second, that it enables a stricter audit and inspection process on an external for-profit supplier and thus, inter alia, reduces corruption (Mgonja & Poncian, 2019).Third, that compared to a municipal collector, an external supplier adopts a more economic, impartial, impersonal, depoliticized procedure of tax collection (Talit, 2012).In accordance with these arguments, Mgonja and Poncian (2019), who researched the successful implementation of outsourcing revenue collection, concluded that support from the central government and the strength and quality of the local management in terms of its political commitment to support fiscal decentralization and transparency, all strengthened governance mechanisms and increased outsourced tax collection rates.Thus, outsourcing is just and needed.
Taken as a whole, our last two moderation hypotheses are as follows: Hypothesis 4. The outsourcing of tax collection moderates the relationship between perceptions about tax enforcement and tax collection rates, such that when the local authority outsources tax collection, the correlation between perceptions about tax enforcement and tax collection rates becomes more positive and stronger.
Hypothesis 5.The outsourcing of tax collection moderates the relationship between perceptions about local governability and tax collection rates, such that when the local authority outsources tax collection, the correlation between perceptions about local governability and tax collection rates becomes more positive and stronger.

| The Israeli case
According to Torgler and Schneider (2007), differences in compliance behavior may be a result of sociological, cultural, institutional, and ethnic identities.Thus, it is crucial to describe the context in which our local authorities operate.
Everyone in Israel who lives in a property, whether they are owners or renters, is required to pay property taxes that are used to fund the activities of the municipality (Israel State Comptroller, 2015).Although property tax systems are based locally, they are strongly affected by the dominant Ministry of Finance that traditionally takes a centralist, conservative view towards local governments.Together with the Ministry of the Interior, since the first decade of the 2000s, these ministries have taken a strict approach to local finances, especially but not solely with less sound and under-performing municipalities (Beeri, 2021;Reingewertz & Beeri, 2018).For example, the central government determines local tax tariffs and local municipalities have no authority to grant property tax exemptions or discounts independently, leaving less room for localism and local democracy (Beeri & Yuval, 2015).
Moreover, Israeli local authorities rely on the money they raise more so than is common in other countries.Official data confirm that in Israel about 66% of the local authorities' budgets depend on revenues that they raise themselves, primarily from property taxes (compared with European local authorities that raise 33% of their budget by themselves) (Ministry of the Interior, 2012-2016).In a broader sense, Israeli local authorities' fiscal dependency makes the relationship between the central and local governments extremely politicized and centralized (Beeri & Razin, 2015;Kimhi, 2011;Zeedan et al., 2017).Moreover, the inspection and control of the central government over local taxes are rather loose and fuzzy (Dery, 2002), and local tax systems vary widely in terms of their efficiency and performance (Brender, 2005;Reingewertz & Beeri, 2018).
Over the last three decades, the Israeli central government has used strict national indicators to determine which municipalities have failed to meet financial standards, gone deeply into debt or did not achieve adequate tax collection rates.To deal with these local authorities, the central government has issued warnings, made grants and loans conditional, expropriated local powers, imposed recovery plans, brought in external accountants and even established committees to bring about financial recovery.Thus, local authorities were required to enforce the collection of taxes, prompting them to outsource this activity (Reingewertz & Beeri, 2018).
The Israeli case is useful for exploring the research question for several reasons.First, given that all local authorities function under common rules regarding tax policies, property tax collection rates can be compared.Second, given the limited autonomy of local authorities in setting property tax levels, indirect factors related to local politics, governance and administration may be key explanations for possible differences in tax collection rates.

| Municipal data
Sampling local authorities is challenging because outsourcing does not occur randomly.Municipalities that did or did not outsource their tax collection differ in various ways.Following Yamada and Arai (2020), we overcame this limitation by using stratified random sampling.We stratified the municipalities into several strata, making sure that the differences in municipality-level variables were not statistically significant.Then, from each stratum, we randomly sampled municipalities using a simple random sampling design and matched those with similar structural characteristics.Most importantly, our data reflect the 120 local Israeli authorities that have more than 10,000 residents.Of these, we sampled 12, of which 6 outsourced and 6 did not outsource their tax collection (see Table 1). 1

| Sample and procedure
In 2015, we developed a survey that was conducted by telephone through a well-known professional panel company with about 50 randomly sampled participants from each of the 12 local authorities.We designed the questionnaire and carefully guided the company about data collection.The company conducted the survey during all accepted Israeli workdays (Sunday to Thursday) and during, before and after work hours 7:30 am-5:30 pm).
We approached 880 residents and interviewed 607 in total (N = 607), resulting in a relatively large response rate of 69% (response rates in local authorities ranged between 65% and 73%).Residents were asked to report their perceptions about the local authority's fairness, tax enforcement and governability.Of the 607 residents, 59% were females and 41% males.Sixtyseven percent were Jewish and 33% were non-Jewish.The participants ranged in age from 18 to 97 and on average were 51 years old (SD = 16.8).On average, they had 13.7 years of education.(SD = 3.6).Regarding marital status, 20% percent were unmarried, 42% married, 31% married with children under 18 years old, 3% were widowed and 4% were divorced.Fifty-seven percent of the participants indicated that they earned a monthly income below the national average ($$2500), 25% made a monthly income close to the national average and 18% earned more than the national average.Though not perfectly match, this sample is very representative of the adult population in Israel who live in communities with more than 10,000 residents (ICBS, 2020).

| Main research variables
First, we calculated the dependent variable, the tax collection rates, measured as the average of rates of property taxes collected by the local authority at the household level over the five years prior to the survey, between 2010 and 2014 (see Table 1).We assumed that residents had formulated opinions about the municipality's tax performance during these five years, opinions that were measured in 2015.We used charges for the legal payment and clearing of tax rebates noted in the Israel Central Bureau of Statistics' (ICBS) data.We used an average of five years to avoid bias resulting from economic conditions, such as the sensitivity of one tax year tariffs, fluctuations stemming from immigration, the real estate market and the value of apartments.We argue that this measure is appropriate for capturing possible one-time changes in tax collection rates.In addition, to provide a broader picture, we also considered the earlier tax collection rates between 2005 and 2009.
Second, we conducted a confirmatory factor analysis (CFA) with varimax rotation, which provided goodness of fit indices indicating the adequacy of the factors' solution.This method enabled us to determine which factors exist, how they relate to the variables, and how they relate to each other.The factorability of the CFA was good.The Kaiser-Meyer-Oklin measure of sampling adequacy showed an adequate fit (KMO = .803),exceeding the recommended value of 0.6, and Bartlett's test of sphericity reached statistical significance (p < .001).Both tests indicated compact patterns of three clusters of items.We grouped the items (see Appendix A) into a three-factor solution.The variances explained by the factors were 29.0, 17.2 and 15.1, respectively, and the total variance explained by the factors was 61.3%.The eigenvalues for the three factors were 4.8, 1.9 and 1.2, respectively.The first factor consisted of six items, with factor loadings ranging from .49 to .79 (Cronbach's alpha .84).The second factor consisted of six items, with factor loadings ranging from .38 to .70 (Cronbach's alpha .69).The third factor consisted of three items, with factor loadings ranging from .46 to .91 (Cronbach's alpha .75).Then, we used several techniques such as sum scores, above a cut-off values, standardized variables and weighted sum scores (DiStefano et al., 2009;Suhr, 2006) to compute scores for the factors.Since the various series of factor scores were highly correlated with the average scores of the scales (.211 to .867) and revealed very similar results in the statistical procedures such as Pearson's inter-correlations, hierarchical multiple regressions, and the PROCESS macro for SPSS (Model #14) that we used to test the hypotheses, we continued to use the average scores of the scales (Kilic, 2018).Thus, we measured the next three variables using the average responses to five items.The participants were asked to rate their agreement with these items on a 5-point Likert scale from 1 (totally disagree) to 5 (totally agree).The full scales appear in Appendix A. Perceived fairness reflects the feelings of the residents about their local authority's fairness, honesty, ethical behavior, and moral actions (Haj, 2007;Vigoda, 2000) (sample item: "My local authority is fair to all its residents" (Cronbach's alpha = .76).Perceived tax enforcement reflects the residents' evaluation of the local authority's efforts, capabilities and success in collecting local taxes and acting against evaders (Fjeldstad, 2004;Haj, 2007) (sample item: "The local authority handles violations of the law, such as evading tax payments" (Cronbach's alpha = .70).Perceived local governability reflects the residents' attitudes about the local authority's capability to design and implement local policy (Kooiman et al., 2008) (sample item: "The mayor and council members design and implement local public policy" (Cronbach's alpha = .76).An exploratory factor analysis confirmed the dimensionality of these three factors (KMO = .85*,Total variance explained = 61.52 for the three components with eigenvalues>1.0).
In addition, we controlled for whether the local authority outsourced tax collection (=1) or not (=0).Local authorities that were coded as outsourcing their tax collection had done so repeatedly and continuously for at least seven years from 2008 through 2014.We based this coding on financial reports published by the Ministry of the Interior and direct contacts we made with the treasurers of the local authorities.Local authorities that were coded as not outsourcing their tax collection had never done so between 2005 and 2014.The outsourcing of all tax collection services and enforcement was to private providers (see Table 1).

| Control variables
We controlled for variables at the individual level, including age, education and salary, and for municipal-level variables for each participant according to the local authority in which he/she lived.Thus, we had 607 observations here as well.The ICBS ranks local authorities on a 10-point scale with regard to their socio-economic status and geographic closeness to the center of the country, with higher numbers indicating wealthier communities closest to the center.In our sample, socio-economic status ranged from 2 to 8 and averaged 5.1 (SD = 2.22) and peripheral status ranged from 4 to 9 and averaged 6.2 (SD = 1.53).

| Statistical procedures
We used the IBM SPSS 23.0 software package to analyze the data.After checking for internal consistency using Cronbach's alpha and the descriptive statistics, we calculated inter-correlations and the linear relationships among the variables using Pearson's correlations.In order to determine whether mediation and moderation effects can predict local tax collection rates, we used the PROCESS macro for SPSS (Model #14). 2 Model #14 allowed us to estimate the regression coefficients and examine each conditional indirect effect at the value of the moderator.The PROCESS macro uses bootstrap sampling, which is a very large number of random resamplings that produce confidence intervals for the parameters (Hayes, 2017(Hayes, , 2020)).Model #14 allowed us to test H2, H3, H4 and H5 at the same time and determine the unique effects above and beyond other effects.

| FINDINGS
In Table 2 we present the descriptive statistics and correlations for the focal variables and the control variables at the individual and municipal levels.The ranges of the variance inflation factor (VIF), which quantifies the severity of multicollinearity, were very low (1.130 to 1.582).Therefore, we conclude that there is no concern about multicollinearity in this data.
In H1, we expected that the more residents feel that the local government is fair, the higher the local tax collection rates.H1 was confirmed, as the findings point to a significant, direct, positive correlation between perceptions about fairness and tax collection rates (r = .190,p < .001).In addition, in order to estimate this finding in a multilevel model that accounted for the structure of our data, we also performed a hierarchical regression.The results of this regression supported H1 (see Table 3).They indicated that the significant positive correlation between perceptions about fairness and tax collection rates (b = .06,p < .05)was solid (R 2 = .523,F = 95.31***),above and beyond control variables at the individual (i.e., age, education and salary) and municipal levels (i.e., socio-economic status and peripheral status).
Interestingly enough, the direct bivariate correlations between outsourcing and the indicators of perceptions were small and statistically insignificant, which makes the moderation effects in H4 and H5 even more relevant.
We also used Model #14 to estimate the conditional indirect effects, that is, the effect of perceptions about the local government's fairness on collection rates, using two mediators: perceptions about tax enforcement and governability, and using outsourcing as a moderator (see Table 4).This model estimated three equations in three steps, including the prediction of perceptions about tax enforcement, governability, and tax collection rates.Tax collection rates were also predicted by two terms: for outsourcing and non-outsourcing local authorities.According to the full model, variations in the independent variables explained 86% of the variation in tax collection rates (R 2 = .86,F = 257.7*).Some predictors at the individual and municipal level were significantly related to tax rates.Older, less educated residents and those who earned more tended to live in local authorities that introduced higher tax collection rates.Local authorities populated mostly by Jews and high-income residents, located close to the center of the country, and that had not outsourced their tax collection also introduced higher tax collection rates.In addition, the research variables and interactions significantly predict tax collection rates above and beyond predictors at the individual (i.e., age, education, salary) and municipal level (i.e., outsourcing, majority [Jewish], socio-economic status, peripheral status and outsourcing of tax collection).We tested the indirect effect of perceptions about fairness on tax collection rates through perceptions about tax enforcement (H2) and through perceptions about governability (H3) and found a positive indirect effect (b = .009*,95% [LLCI = .005,ULCI = .013] 3; [b = .012*,95% [LLCI = .006,ULCI = .017]],respectively).Thus, H2 and H3 were confirmed: the more residents feel their local authority is fair, the more they think that their local authority enforces tax collection effectively and their local authority's governability is better, respectively, and, in turn, the higher their local authority's tax collection rates.
In H4 and H5, we expected that the outsourcing of tax collection would moderate the relationship between tax collection rates on one hand and perceptions about tax enforcement, and perceptions about governability, on the other hand.We posited that outsourcing would strengthen these relationships in a positive way.H4 and H5 were confirmed.The data demonstrated significant moderation effects (see Table 4).In local authorities that had not outsourced tax collection, there was a significant negative relationship between perceptions about tax enforcement and tax collection rates (b = À.003*, 95% [LLCI = -.006,ULCI = -.001]), and a significant negative relationship between perceptions about governability and tax collection rates (b = À.006*, 95% [LLCI = -.011,ULCI = -.002]).
In contrast, in local authorities that had outsourced tax collection there was a significant positive relationship between perceptions about tax enforcement and tax collection rates (b = .005*,95% [LLCI = .003,ULCI = .009])and a significant positive relationship between perceptions about governability and tax collection rates (b = .005*,95% [LLCI = .001,ULCI = .010]).Thus, only in the local authorities that had outsourced their tax collection did more residents regard tax enforcement as strong and governability as high.In turn, their tax collection rates were higher.Furthermore, in local authorities that had not outsourced their tax collection, stronger perceptions about tax enforcement and governability were related to lower tax collection rates above and beyond our expectations (see Figure 2). 4  T A B L E 3 Hierarchical multiple regression analysis (standardized coefficients) for the relationship between perceptions about fairness and tax collection rates Dependent variable = tax collection rates

Variables
Step Beyond the statistical significance of the residents' opinions, the results also have economic implications that are important from a tax policy perspective.Tax collection rates (the dependent variable, 2010-2014) for all sampled local authorities averaged 80% (SD = 15.6), averaged 85% (SD = 9.3) for non-outsourcing local authorities, and averaged 75% (SD = 18.7) for outsourcing local authorities.During the period that preceded the dependent variable (2005)(2006)(2007)(2008)(2009), tax collection rates for all sampled local authorities averaged 74% (SD = 21.8),78% (SD = 18.8) for non-outsourcing local authorities, and 71% (SD = 24.1)for outsourcing local authorities.Some trends can be observed from these calculations.First, on average, local authorities increased their tax collection rates in the beginning of the 2010s, compared with the end of 2000s, regardless of how they collected their taxes.Second, initial tax collection rates of those who outsourced their collections were lower than those that did not, which may explain their incentive to outsource.Third, beyond the examined period (2010-2014) and compared to the preceding period (2005)(2006)(2007)(2008)(2009), the increase in tax collection rates of nonoutsourcing municipalities was larger.Thus, the gap between the non-outsourcing and outsourcing local authorities widened over time.In addition, on average, during 2010-2014 the outsourcing local authorities had not reached the level of tax collection rates of nonoutsourcing local authorities during 2005-2009.However, one should not underestimate the economic benefits of outsourcing.First, we cannot estimate what would have been the tax collection rate of the outsourcing local authorities if they had not outsourced tax collection.Second, according to the 2018 ICBS data, the 4% increase in household level tax collection rates by outsourcing local authorities contributed to an increase of about 2.5% in the municipalities' yearly property tax revenues, an increase of 1.7% of their yearly self-revenue, and an increase of 0.7% in their yearly total income.

| DISCUSSION AND CONCLUSIONS
Using data from Israeli local authorities, we investigated the relationship between residents' perceptions of their local governments as fair and tax collection rates.We also explored whether F I G U R E 2 Two moderating effects of outsourcing tax collection on the relationship between (i) perceptions about tax collection and tax collection rates and (ii) perceptions about governability and tax collection rates, based on the conditional full model (mediation-moderation PROCESS model #14 in Table 4) this relationship is mediated by perceptions that the local authority enforces the collection of taxes and governs appropriately, and whether the outsourcing of local tax collection moderates this relationship.As expected, the results indicated that perceptions about fairness are directly related to tax collection rates (H1).Furthermore, perceptions about fairness are indirectly related to tax collection rates through perceptions about tax enforcement and governability (H2 and H3).Additionally, in local authorities that outsourced their tax collection, the more the residents regarded tax enforcement and governability as efficient, the better their local authority performed, measured as collecting a higher rate of property taxes (H4 and H5).These results underscore the dynamics of market relations (Ramamurti, 2000;Starr, 1988), where the partnership with private organizations improves residents' opinions about the ability of their local authority to operate a fair, efficient and effective tax system (Talit, 2012).However, in local authorities that did not outsource their tax collection, the more the residents regarded tax enforcement and governability as efficient, the less successful the local authority was in collecting property taxes.One explanation for this interesting finding is the complex dynamics between residents and non-outsourcing local authorities.The latter's ability to collect taxes may hint at the ability and standards of their leaders.Perhaps they had made extensive public efforts recently to increase tax compliance.As a result, their residents who probably had a limited ability to compare their tax collection rates with those of other local authorities, believed that their leaders were committed to improving governability and tax enforcement.
These relationships are meaningful and important in several ways.They have several theoretical and practical implications for the study of local democracy, governance, outsourcing, and local tax policy.First, above and beyond individual socio-economic characteristics and the structural characteristics of the local authority, and above and beyond shifts in economic developments and popular managerial trends of governance (Haus & Sweeting, 2006a;Osborne, 2010), the unwritten contract of local democracy is still stable.We found that regardless of managerial technique or ideology about service delivery, the local democratic framework between residents, local bureaucrats and politicians is far beyond a rational calculation.Rather, it is rooted in and depends on residents' feelings about the fairness of the local tax collection and its enforcement, as well as assessments about the ability of the local authority to govern.Moreover, while most studies maintain that the centralist policies of the Ministries of Finance and Interior leave less room for localism (Beeri & Yuval, 2015;Reingewertz & Beeri, 2018), our findings indicate that there is more room for local responsibility, autonomy and tax policy.Above and beyond the effect of residents' perceptions about their leaders' fairness, governance and enforcement capabilities, the decision to outsource local tax collection increased household-level property tax revenues in economic terms.
Accordingly, both conceptually and empirically, our model extends Fjelsdstad's ( 2004) claim that enforcement is related to the willingness to pay.We found that regardless of individuals' age, education or salary, or the communities' national majority, socio-economic and peripheral statuses, the residents' sense that the local authority was fair was the basis of an efficient local tax system and high tax collection rates.When residents felt that their local authority was fair, they were more likely to regard tax enforcement as efficient, effective and ethical.They were also more likely to regard their local leadership as fulfilling its duty to govern, make decisions and carry out those decisions.In turn, they were willing to reward this legitimate local leadership by doing their civic duty and paying their taxes, at least as a community.
Our findings accord with the loop model of democracy (Fox & Miller, 1995).Feelings about fairness are not only the basis but also the outcome of higher rates of tax collection.When residents believe that their local authority enforces tax collection, punishes those who do not pay, and maximizes municipal revenues effectively, they feel that their local authority treats taxpayers equally and that it is ethical and fair.Nevertheless, in response to Beeri et al. (2019) who claimed that fairness might result in mutual trust and the willingness to pay taxes, we assert that the whole cycle of the local tax system might become a dangerous double-edged sword.Residents who feel that the local authority treats them unfairly, whether that sense is justified or not, make harsh judgments about the performance of the local authority.They may conclude that tax enforcement is weak and ineffective, and that, in their view, governability is poor.From here, it is a short step to having that sense of unfairness deteriorate into delegitimizing the government and evading taxes.
Second, we provide some theoretical insights into political systems that tend to be centralized.Our model controls for democratic elements, managerial techniques and financial performance.Thus, the findings lie in between the public policy and the economics perspectives (Domberger & Jensen, 1997).In our view, the data that were collected in 2015 reflect the modern expectations of national governments on the one hand, and of residents and local communities on the other, in the era of new public governance and local network governance (Osborne, 2010).As was evident in the worldwide outbreak of coronavirus (COVID-19) (Parodi & Liu, 2020), local authorities are expected to provide high quality local services that are relatively independent of national budgets through various local collaborations, not competitions.
We found that most Israeli local authorities were pushed to increase tax collection and indeed increased tax collection rates in the beginning of the 2010s, compared with the end of 2000s, regardless of how they did so.The trend of pushing local authorities to raise revenues independently is thus still alive and especially relevant for Israel where a large percentage of the municipalities' budgets depend on revenues that they raise themselves.In a broader sense, in accordance with Brender (2005) who stressed that local tax systems vary widely in terms of their efficiency and performance, we found that the repeated and continuous outsourcing of tax collection increased local revenues and improved the local economy significantly.
In accordance with Mgonja and Poncian (2019), who concluded that successful outsourcing depends heavily on the managerial qualities of local leader, we stress that successful outsourcing in local authorities also relates to local political and democratic qualities.According to our findings, only in local authorities that outsourced tax collection services were the perceptions about governability and tax enforcement related to better financial performance.Following our assertion that local taxation is a delicate policy action, local leaders and residents seem to wisely differentiate between administrative actions and broader duties that should be reserved to governmental bodies.
Put differently, the findings accord with Haus and Sweeting's (2006b) typology of local democracy.Outsourcing in the era of local network governance has had the unintended consequence of promoting two types of local democracy-market-like local democracy and local network democracy.Side by side, they promote mutual trust between various local stakeholders and the sense that the government is fair, can govern and is legitimate.Local authorities have solved the bureaucratic-democratic paradox, at least to some extent.First, local authorities have entrusted the technical details of tax collection to a private franchisee, creating a modern, efficient, impartial, taxation procedure.Second, they are now freer to focus on strategic policy design that positions them as those responsible for the creation of a community climate that aligns civic duties and tax payments.Thus, the decision to outsource tax collection, combined with actions that improve residents' positive perceptions about strong tax enforcement, trust and governability, have led people to pay local taxes to a greater extent and improve their municipality in financial terms.

T
A B L E 2 Descriptive statistics and Pearson's correlations for the control and research variables T A B L E 4 Conditional full model (mediation-moderation PROCESS model #14) examining (i) two mediation effects of perceived tax enforcement/perceived governability on the relationship between perceived fairness and tax collection rates and (ii) the moderating effect of outsourcing Dependent variable = tax collection rates R 2 = .860;F = 257.7***