IDEAS AND INSTITUTIONAL CONVERSION THROUGH LAYERING: THE CASE OF ISRAELI IMMIGRATION POLICY

Transcription

1 doi: /padm IDEAS AND INSTITUTIONAL CONVERSION THROUGH LAYERING: THE CASE OF ISRAELI IMMIGRATION POLICY ILANA SHPAIZMAN Gradual transformative policy change has attracted increasing attention in recent times. However, existing explanations for the evolvement of the various modes of such change do not account for the direction and content of change and also have difficulty with more complex policy change processes in which several modes are employed. Ideational analysis can fill this gap, albeit most ideational research does not address gradual policy and ideational changes. Based on a process tracing of two cases from Israeli immigration policy, this article argues that a low level of discretion and a strong status quo bias are conditions that enable institutional conversion through layering. Ideational change in policy solution and problem definition ideas, and the interaction between the two, can explain this pattern and its direction. The article not only expands our knowledge of gradual transformative change, but also moves ideational research a step beyond punctuated equilibrium. INTRODUCTION How and why does policy change take place? These questions have been occupying policy researchers for decades. Lately, attention has been drawn to more hidden facets of policy change: changes which evolve gradually, and eventually lead to a significant shift in the status quo. These are called gradual transformative changes and may evolve through five distinct modes: displacement, drift, exhaustion, layering, and conversion (Streeck and Thelen 2005). There are two factors explaining why change proceeds in one of those modes rather than another: the status quo bias of the political environment and the policy s level of discretion (Hacker 2004; Mahoney and Thelen 2010; Hacker et al. 2013). A typology based on these factors has two weaknesses, however. First, only one mode of change is associated with each set of political and institutional characteristics. Thus it has difficulties explaining patterns and sequences of several modes of change as found in a number of empirical works (Van der Heijden 2010). Second, it does not explain the direction of change nor its content (Béland 2007; Schmidt 2010). Consequently, although gradual transformative policy changes have characterized most institutional changes in socioeconomic policies over the past 50 years (Streeck and Thelen 2005), our understanding of these changes remains partial. Ideas can fill these gaps because they shape actors policy preferences and institutional development (Hay 2006). However, although the literature notes the importance of ideas in explaining the modes of change, it does not provide theorization for their role. As for ideational scholarship, more often than not it explains abrupt policy changes (e.g. Hall 1993; Blyth 2002) at the expense of gradual ideational change that can lead to transformative results (Carstensen 2011). In addition, existing theories on gradual ideational change (e.g. Carstensen 2011; Mehta 2011) are still underdeveloped both theoretically and empirically. Contributing to the theoretical debate on gradual transformative change, this article examines the conversion through layering pattern and the role of ideas behind it; based on process tracing of two cases in Israeli immigration policy which did not Ilana Shpaizman is at the Federmann School of Public Policy and Government, Hebrew University of Jerusalem, Jerusalem, Israel.

2 IDEAS AND INSTITUTIONAL CONVERSION 1039 experience any authoritative change from the 1990s, it argues the following. First, existing explanations are ill-equipped to explain the Israeli case because under strong status quo bias and low level of discretion, Israeli policy underwent conversion through layering, in which new policies were added to existing policies, thereby converting the Ministry of Immigrant Absorption to new purposes. Second, conversion through layering expands the existing institutional purpose rather than replacing it, and includes change in both policy practices and institutional aims. Third, the direction and content of ideational change correspond to those of policy change. Fourth, this pattern is the fruit, among other things, of ideational change in policy solution and problem definition ideas, and the interaction between the two. Specifically, new policy solution ideas turn actors attention to new policies which gradually expand the existing problem definition, an expansion made possible when policies can reconcile multiple goods. The article proceeds as follows. The next section presents the analytical framework, focusing on modes of gradual transformative policy change, conditions for their evolution, and ideas and ideational change. The following section presents the Israeli setting. Next, the methodology and data are presented followed by empirical analysis of policy and ideational changes in two cases: immigration encouragement and returning residents. The final section discusses the article s conclusions. ANALYTICAL FRAMEWORK Gradual transformative policy change Incremental change that leads to a significant shift from the status quo over time is called gradual transformative change (Streeck and Thelen 2005). This type of change has characterized most socioeconomic institutional changes in advanced Western political economies over the past 50 years. Streeck and Thelen (2005) identify five major modes of gradual transformative change: displacement, drift, exhaustion, layering, and conversion. This article addresses the last two. The evolution of each mode is a function of the political context: strong/weak status quo bias and institutional characteristics; high/low level of discretion in implementation and enforcement (Hacker 2004; Mahoney and Thelen 2010; Hacker et al. 2013). Status quo bias occurs as a result of multiple veto points and strong veto players (Tsebelis 2002), or when there are vested interests or feedback effects that make it difficult to change past decisions (Pierson 2004). Level of discretion is, among other things, a function of the ambiguity of formal rules (Mahoney and Thelen 2010; Hacker et al. 2013). Layering occurs when new institutions, mechanisms, or practices are attached to existing institutions without replacing them, and so gradually change their structure and status. The new practices are first introduced at the system s margins in order to complement or reform it, but gradually, because of differential growth, these practices extend further (Schickler 2001; Streeck and Thelen 2005). Layering is more likely when the level of discretion is low since the actors cannot change the existing institutions and are thus forced to add institutions to the existing ones (Hacker 2004; Mahoney and Thelen 2010). As for the status quo bias, Hacker (2004) argues that low bias enables adding new practices to the existing ones, and Mahoney and Thelen (2010) argue that layering is more likely when there are strong veto possibilities that block any formal change, since layering does not require direct changes in existing institutions. Conversion is a transformation of an existing institution or policy through redirection to new ends and purposes. Thus, formal institutions or policies remain structurally

3 1040 ILANA SHPAIZMAN unchanged but their impact changes. Conversion is required, among other things, when the institution has outlived the conditions prevalent at the time of its foundation, creating a gap between it and the environment. In order to adapt the institution to the changing environment, active redirection and reinterpretation of the policy/institution are needed (Streeck and Thelen 2005; Hacker et al. 2013). Conversion will be more likely when the rules are more ambiguous and there is high level of discretion. This is because it gives change agents an opportunity to take advantage of the ambiguity to redirect policies. The second condition for conversion is high status quo bias that makes it harder to formally change existing policies (Hacker 2004; Hacker et al. 2013). There is also a conflicting view that sees conversion as the result of weak veto possibilities that enable institutional redirection (Mahoney and Thelen 2010). Despite making an important contribution, existing explanations for the evolvement of various change modes are partial and somewhat simplistic. First, it seems that each combination of political and institutional factors is linked to only one change mode. Thus they have difficulty explaining more complex policy change processes, in which several strategies are employed (Van der Heijden 2010). For example, Béland (2007) found that the US social security policy had first changed through layering and then through drift, and Falleti (2010) demonstrated healthcare reform evolution in Brazil through layering followed by conversion. This critique characterizes many institutional works based on typologies, because in a typology each case must fall into a certain cell, while in reality variables are more continuous in space-time (Radaelli et al. 2012). Moreover, change resulting from the combination of several modes might be distinct from one resulting from each mode separately. Second, existing explanations are still mostly descriptive rather than explanatory as they do not account for the content of change or its direction. For example, why do agents of change choose to layer some policies rather than others? Or why is specific change more or less conservative in nature (Béland 2007; Schmidt 2010)? Among other things, this gap is due to historical institutionalism s tendency to examine only institutional and material constraints on policy change, and when examining the motivations of agents of change its focus is on their strategic interests. Yet institutional development and actors perceptions of which polices are more or less feasible are shaped also by the ideas behind the institutions and by actors wider motivations and beliefs (Hay 2006; Schmidt 2010). Interestingly, while existing explanations ignore the role of ideas and discourse in their theorization of gradual transformative change, some of the authors contributing to the edited volumes of Streeck and Thelen (2005) and Mahoney and Thelen (2010) (e.g. Palier 2005; Falleti 2010) provide evidence for the importance of ideas in explaining the various modes of gradual institutional change. Furthermore, Rocco and Thurston (2014), who examine empirical indicators of conversion, layering, and drift, also note that agents of change can use intellectual resources in order to promote or prevent change. However, all these authors only note the importance of ideas and do not theorize how they affect policy (Schmidt 2010). Thus, a more comprehensive account is required, one that will take into fuller account the significant role played by ideas in institutional change, and close the gap between theory and empirical evidence. Punctuated evolution theory could fill this gap, since it views ideas and discourse as one of the main determinants of policy dynamics, limiting the range of responses to policy problems. Specifically it views subjective narrations of crises by agents of change as leading to policy evolution and shaping the direction of change. During periods of crisis, policy change escalates and afterwards consolidates (Kerr 2002; Van der Heijden

4 IDEAS AND INSTITUTIONAL CONVERSION ). However, since this theory examines policy evolution resulting from crisis, it is difficult to apply it to gradual transformative changes which take place during periods of relative stability. Thus, a different ideational explanation is required. Ideas, ideational dynamics, and policy Ideas are causal beliefs (Béland and Cox 2011). Ideational scholars believe that actors have non-material interests and goals that affect the way they act. In addition, there are times when ideas constrain actors behaviour even if the actors do not hold these ideas. Thus, ideas can influence political behaviour and outcomes both by helping determine actors motivations, interests, and goals and by shaping institutional constraints and opportunities (Béland and Cox 2011; Berman 2013). Ideas are represented through discourse. There are two discursive interactions: coordinative discourse among policy actors and communicative discourse between political actors and the public (Schmidt 2008). Policy makers constantly work within a framework of ideas and standards that specifies not only the goals of policy and the kind of instruments that can be used to attain them but also the very nature of the problem they are meant to be addressing (Hall 1993, p. 279). Thus, ideas provide cognitive and normative causal propositions that help define the problem, exclude other problem definitions, direct attention, and act as roadmaps for decision makers, helping them choose between different alternatives (Berman 1998; Béland 2005, 2007; Béland and Cox 2011). Most ideational research examines abrupt rather than gradual policy change (e.g. Hall 1993; Blyth 2002), and argues that ideas affect policy in a punctuated equilibrium manner. In periods of stability, they remain stable and help maintain the status quo; in periods of crisis, however, they help to promote abrupt change by providing actors with new objectives, justifications, and legitimacy for a new policy (Seabrooke 2009; Wincott 2011). Because of this perspective, research often ignores gradual ideational change. For example, Wincott (2013) examined the idea of the golden age of the welfare state and found that both scholars and policymakers took it for granted, ignoring gradual ideational changes taking place during the so-called golden age. As a result, despite its inaccuracy the idea continued to influence research as well as policymaking. Ideas may be classified into three levels of generality (Schmidt 2008; Mehta 2011). Policy solutions are the means for solving the problem and accomplishing the objectives. These ideas can change relatively quickly when a window of opportunity opens for a new agenda. Here, change is usually the result of a discrepancy between ideas and reality (Schmidt 2010). When actors understand that the existing policy solution ideas fail to properly address the problem, existing ideas are eroded and the actors start looking for new ones, more consistent with the new reality. In turn, these ideas focus attention on fresh policy alternatives (Jones and Baumgartner 2005; Jacobs 2009). Problem definitions frame the policy goal and scope of possible solutions, and they are at the same analytical level as programmatic beliefs (Berman 1998) or paradigms (Hall 1993). In any issue there is usually more than one problem definition, each addressing different aspects of a given situation. Policymakers often use symbolic representation such as metaphors to define the problem (Stone 1988). These ideas usually change during periods of crisis when one paradigm is being replaced by another (Schmidt 2008). This paradigmatic shift brings about change not only in policy tools but also in policy goals (Hall 1993). Finally, public philosophies are beliefs about the purpose of government or public policy in light of a certain set of assumptions about society and the market. These ideas are also

5 1042 ILANA SHPAIZMAN called core values (Cox 2004) or worldviews (Campbell 1998). They are often independent of agency since they include societal values (Kuisma 2013). While ideas at the first two levels are foreground ideas because they tend to be constantly discussed and debated, public philosophies are usually in the background, acting as underlying assumptions that are widely shared, rarely contested or changed, and not open to criticism (Campbell 1998; Schmidt 2008). Public philosophies are the slowest to change, since they include changes in the core values of the society (Schmidt, 2008). Moreover, they constrain significant policy change, because each change is evaluated according to a logic of appropriateness (March and Olsen 2004) to these ideas. Hence, policymakers will not pursue a change that contradicts these ideas either because they think the public and political system will consider it illegitimate, or because they themselves consider it so. Furthermore, when the policy is rooted in public philosophies there is a tendency to expand these ideas so that policy changes will appear consistent with them (Cox 2004). Accordingly, these ideas are path-dependent: once created they are constantly reinforced. As a result, public philosophies behind a policy make for strong status quo bias (Pierson 2004). The interaction between the levels can flow downwards: public philosophies shape how specific problems are defined, which in turn affects which policy solution ideas are available. It can also flow upwards: when success or a failure resulting from policy solution ideas is extended beyond the original policy it can be used to question or expand the problem definition. Expansion may occur if they [the new policies] find a way to reconcile multiple goods that previously seemed incommensurable (Mehta 2011, p. 43). In addition, several problem definitions can create a public philosophy, while constant discrepancy between the problem definition and reality might lead to gradual change in public philosophy. This more dynamic conception of ideas corresponds to Carstensen s (2011) theory of gradual ideational change, which views ideas not as stable entities, but as a web of related elements with varying importance. Hence, gradual ideational change may occur when there is change in the relative importance of the different elements within an idea, or in the composition of elements that constitute it. Despite Mehta s and Carstensen s important contributions in moving ideational research beyond punctuated equilibrium, gradual ideational change is still relatively underdeveloped. Carstensen only provides illustrative evidence for his theory, while Mehta himself does not claim to have a theory that would explain when these outcomes are likely (Mehta 2011, p. 43), and only provides examples of how this might work. This article aims to address these gaps. SETTING Israel s immigration policy is an open door policy to all Jews. Moreover, Israeli integration policy is characterized as categorical universalism every member of that category (Jewish immigrant) is entitled to some benefits regardless of economic status (Gal 2008). This policy is rooted in Zionist ideology, which viewed the State of Israel as an asylum for all Jews, and considered immigration as the main nation-building instrument. Even before its establishment in 1948, decision makers viewed the gathering of exiles as one of the main missions of the emerging Jewish state. Consequently, it was believed that Jewish immigration should not be restricted. Through the years, and despite the enormous costs of large-scale immigration, any suggestion of restricting it was dismissed and seen as taboo (Hacohen 2003). Moreover, government responsibility for immigrant integration

6 IDEAS AND INSTITUTIONAL CONVERSION 1043 was part of the national ethos. For example, during the 1990s large numbers of elderly immigrants arrived from the former Soviet Union (FSU). The government not only maintained responsibility for their integration, but also built public hostels for those who could not afford housing (Borukhov 1998). Thus, Zionist public philosophies created a strong status quo bias. These ideas are formally manifested in two laws. The Law of Return ( ) entitles any Jew, a descendant or a spouse of a Jew to immigrate to and become a citizen of Israel. This law is seen as the cornerstone of the State of Israel and thus, despite the fact that a simple majority can change it, all attempts made hitherto have failed (Plenum Protocols available at main.knesset.gov.il/activity/plenum). Second, the Absorption Basket Law ( ) entitles Jewish immigrants from most countries to in-cash benefits over the first year after arrival. Not only has no attempt ever been made to revoke this law, in 2002 the government decided to expand its entitlement to include immigrants from Western countries as well (PMO 2002). These laws and the detailed regulations implementing them offer almost no room for discretion. Two institutions are in charge of immigration policy: the Jewish Agency for Israel (JAFI) (an international, semi-governmental philanthropic organization) and the Ministry of Immigrant Absorption (MOIA). DATA AND METHOD The analysis is based on process tracing (George and Bennett 2005; Jacobs 2011; Beach and Pedersen 2013) of ideas and policy in the Israeli immigration policy. Specifically, it examines two cases: immigration encouragement and returning residents. The analysis starts in 1989, when the large immigration wave from the FSU began, and ends in The rationale is that since 1989 there have been no other large immigration waves or other exogenous shocks in this field, so that the period can be seen as relatively stable. Policy changes were identified by qualitative analysis of available policy documents on immigration policy, such as government decisions, MOIA regulations, archival documents, national budgets, and protocols of the permanent parliamentary Committee of Immigration, Absorption and Diaspora Affairs (CIAD). Change was considered gradual transformative when there was a stepwise change in policy tools, objectives, or allocated resources during the research period, resulting in policy discontinuity. New policies were programmes which had a different logic of action and had either never been proposed before or had been introduced and ignored. Changes in policy aims were examined in light of MOIA objectives as presented in the annual national budget, the ministry s work plan, and the periodic reports to CIAD. Policy solution and problem definition ideas were inferred by qualitative analysis of archival documents on immigration policy, MOIA press releases, explanations of the annual budget, MOIA work plans, explanations of government decisions, and CIAD minutes. In addition, the author conducted 15 in-depth interviews with past and present ministers, general directors, and heads of departments in MOIA. The interviewees were selected for their policymaking roles, covering the entire period. Ideas were uncovered by focusing on the reasoning behind the policy as provided in the interviews and documents, and as indicated in the metaphors used. Since metaphors are mostly used subconsciously, their analysis provides another insight into actors true beliefs (Lakoff and Johnson 1981). Ideational change was a change in the composition and/or relative importance of the different parts of the idea (Carstensen 2011) compared to the beginning of the period examined. Change was evidenced by use of metaphors, causal reasoning, problem definition,

7 1044 ILANA SHPAIZMAN or advocacy of policy tools with a different logic of action from that suggested earlier by various actors and dismissed, or originating in other fields. In addition, in order to make sure that ideational change took place before policy change, the actors familiarity with the ideas preceding the policy was examined. To ensure that the reasoning provided in the interviews was reliable, the interviews, CIAD minutes, and archival documents were triangulated. Finally, close attention was paid to the context in which the actors were expressing their opinions, in order to consider who is speaking to whom, for what purpose and under what circumstances (George and Bennett 2005, p. 100). EMPIRICAL ANALYSIS Immigration encouragement Until 2002, JAFI was responsible for immigration recruitment or promotion overseas, and upon arrival of the immigrants the responsibility for their integration was transferred to MOIA. For the most part, JAFI s policy was universal and treated all eligible immigrants equally. Its policy was mostly reactive providing information to immigrants wishing to immigrate. The policy focus was almost entirely on rescue immigration for Jews in distress, rather than choice immigration, mostly from Western countries. In addition, most resources were directed to the FSU and only a small percentage was allocated to Western countries (Meridor 1999; Edelshtain 2001). During the period examined the annual number of immigrants decreased gradually from a peak of 200,000 in 1990 to 18,000 in 2010 (MOIA 2010a). Moreover, from 2002, most potential immigrants were immigrants by choice, mostly from the USA (Della Pergola 2010). Although these changes began at the end of the 1990s and were apparent to decision makers, as seen in their reports to parliament (Meridor 1999; Edelshtain 2001), until 2005 JAFI and MOIA pursued the same policy. Thus a gap was created between the reality and the policy impact. The gradual change started in Until 2004 immigration encouragement was never explicitly stated as one of the ministry s objectives, and the main focus was on integration issues, such as housing, employment, and social assimilation (MOF ). In 2004, the minister presented a four-year immigration encouragement project whose main programme was group immigration. Under this programme, MOIA would proactively contact potential immigrants from Western countries. In addition to the regular assistance, these immigrants would be entitled to additional Hebrew teaching hours, employment assistance, social and communal services, and a group coach (MOIA 2009b). As a result of the four-year project, in immigration encouragement was presented in the budget as one of the ministry s additional objectives (MOF ). At first, there was only one community programme for a small number of French immigrants. From 2005, the community programme was expanded to include additional countries of origin (MOIA 2009b). In that same year, another programme was initiated that offered financial support to non-governmental organizations promoting immigration from France and North America. This programme grew from 10 m New Israeli Shekels (NIS) in 2005 to 30 m NIS (more than 5 m) in 2010 (PMO 2005, 2007, 2010b). From 2010, the target audience of the community immigration programme changed from immigrants from specific Western countries to immigrants with specific occupations, like doctors and engineers (MOIA 2010c). This made Israeli immigration policy more similar to that of other immigration countries which recruit skilled immigrants by offering various economic incentives (Shachar 2006).

8 IDEAS AND INSTITUTIONAL CONVERSION 1045 These changes were accompanied by a significant increase in the total budget. In 2004 the immigration encouragement budget was 17.6 m NIS, less than 1 per cent of the total MOIA budget; in 2005 it was 72 m and in 2010 it was 128 m, more than 14 per cent of the total (MOF ). This increase was not a result of an increase in the number of immigrants, since they decreased from 22,806 in 2005 to 18,755 in 2010 (MOIA 2010a). The selective proactive programmes were added to the existing universal reactive policy, rather than replacing it. JAFI continued to provide information and assist those who wanted to immigrate as it did before (Sheranski 2010), the laws did not change, and MOIA s universal categorical integration policy remained in place (Landver 2009). From 2007 onwards, immigration encouragement became the ministry s primary objective as expressed both in the budget and in the annual work plans (MOF ; MOIA 2008, 2009a, 2010b). Thus, by the end of the research period MOIA was gradually converted to a new purpose from integration per se to the encouragement of skilled immigration. This is also evident in the change made in the ministry s name, from The Ministry of Immigrant Absorption to The Ministry of Immigration and Absorption ( This conversion was enabled through layering, such that with each additional policy with a proactive and selective logic of action, MOIA moved further towards immigration encouragement. Two questions remain unanswered. First, what can explain this policy change and its direction? Although environmental changes were a necessary precondition, they alone don t provide an explanation because most changes occurred in , at a time when the number of immigrants and their socioeconomic profile did not alter significantly (MOIA 2010a). Neither do political interests suffice as an explanation: in there were five ministers from three different parties, but the policy continued in the same direction. Moreover, the last minister, Sofa Landver, was a member of a party representing veteran immigrants, mostly from the FSU, such that the policy changes were contrary to her direct political interests. The second question is why MOIA changed not only its function but also its purpose. Why was layering accompanied by conversion? An analysis of the ideational dynamics can provide an answer to these two questions. At the beginning of the period examined (1989), the problem definition was that only push factors (such as anti-semitism or economic crisis) increased Jewish immigration to Israel. The metaphor used was competition between countries, and decision makers believed that Israel was bound to lose. Consequently their attention was focused on prospective immigrants in distress. The policy solution ideas were to make sure that the push factors pushed them to Israel rather than elsewhere. Although the decision makers believed it was important to improve integration so that those already in Israel would not emigrate, this element was marginal in the ideas behind immigration policy, evidenced by the fact that immigration and integration were discussed as two distinct processes (MOF ; Denitz 1991; Tzaban 1993; Burg 1995). These ideas did not change for most of the period under examination. From 2002, as a result of the decrease in immigration, decision makers began questioning the existing policy solution and problem definition ideas. JAFI executives began saying that they had to re-examine their policy tools. Several meetings were held in CIAD, in which members accused JAFI of failure given the falling number of immigrants, and requested that it change its approach (Meridor 2002, 2003). MOIA officials also started to publicly assert that the world in general, and Israel and the Jewish world in particular, were changing. Hence, if Israel wanted to continue attracting immigration, it should invest more in proactive immigration encouragement (Livni 2004; MOIA director). These

9 1046 ILANA SHPAIZMAN suggestions were accompanied by repeated statements by immigration ministers to the CIAD, that perhaps Israel could win the competition after all (Edelshtain 2002; Livni 2004). Since JAFI s acknowledgement of failure was public, it was more likely driven by its belief that the existing solutions did not work rather than by other interests. As a result, the decision makers attention was gradually diverted to immigration by choice. They started to believe that the number of immigrants was determined by pull factors such as better conditions in the receiving country, and the problem definition changed to Israel s attractiveness. For example: Israel today has the potential and the ability to compete with other welfare states. It is attractive today and can be even more attractive, depending on what we would do to encourage the immigrants to come here. (Minister Zeev Boim, CIAD, 31 January 2007) However, attractiveness was not conceived in terms of economic conditions but rather in terms of making the immigrants feel at home. Consequently the policy solution ideas focused on providing personal guidance before and after immigration and creating a peer group that can provide support during the integration process (Livni 2004, 2005; Boim 2006, 2007; Landver 2009). These ideas were not new and decision makers were familiar with them since they had been implemented in previous ad hoc programmes to promote immigration from France and South Africa (CIAD 2003). When these ideas became widely accepted by decision makers, their attention was turned to fresh policy alternatives based on pull rather than push factors. From 2007 the problem definition was narrowed from attractiveness for immigrants in general to attractiveness for economic immigrants. The decision makers emphasized the economic benefits from immigrants from Western countries. The immigrants were presented as an economic resource, bonds and assets : When I went to the Minister of Finance I told him: You should invest in these immigrants [from North America] the same way you invest in a bond. You invested 300 million NIS in the absorption basket and you gained 600 million. (MOIA director) These metaphors implied that the solution should be further investment in skilled immigrants. This discourse intensified even more when the minister began to present MOIA as an economic ministry (Boim 2007; Landver 2009). These ideas were imported from global immigration policies that considered skilled immigrants to be an economic resource (Shachar 2006). The decision makers were familiar with these ideas as indicated by their reference to global competition, both at CIAD meetings and in the interviews. Another set of metaphors was borrowed from the fashion world, emphasizing the need for selective as well as universal assistance: In order to attract skilled immigrants MOIA should be like a department store which specializes in tailor-made suits along with mass-produced clothes. (MOIA director) These ideas led to the expansion of the selective programmes aimed at professional immigrants, in addition to the universal programmes. As a result of the new immigration encouragement policies, the decision makers started to question MOIA s entire approach of universal assistance and separation between immigration and integration. I believe in providing public services at different levels. I think that skilled immigrants require more focused and intense assistance than unskilled immigrants If I have for example an engineer coming from Argentina and a homemaker, I will not use the same tools for both. I will give the homemaker a voucher for vocational

10 IDEAS AND INSTITUTIONAL CONVERSION 1047 training and he/she will manage. The engineer, on the other hand, needs special treatment because I need engineers in Israel. (MOIA director) At the end of the period under examination, immigration and integration were discussed as a single process, the metaphor used was continuum, and the MOIA s problem definition was expanded. For example, when asked about the future of MOIA, one director said: I don t see the future of MOIA in adding social workers to work with immigrants, but I do see the future in connecting immigration and integration and thus seeing the big picture. I believe it s more important to both the ministry and country, and if it were up to me, that s where I d go. (MOIA director) This expansion was possible because the new policy solution ideas did not question the public philosophies on which immigration and integration policy was based. The decision makers continued to justify the policy using Zionistic notions of nation-building and the gathering of exiles (Boim 2007; Landver 2009, 2010). Thus it was possible to reconcile integration with attracting skilled immigrants. Returning residents MOIA had been responsible for returning residents since However, this issue was one of the least significant on its agenda, as evidenced by its small annual budget (2 m NIS) and the small unit in charge. The policy was reactive and its activities included assistance to children of returning residents, some minor assistance in employment (subject to socioeconomic status), and keeping in touch with Israeli émigré communities (Prigat 1993, 2004). Since 2001 there has been a constant decrease in the number of emigrants and an increase in the number of returning residents (CBS 2007). Furthermore, the constant decrease in immigration turned the decision makers attention to returning residents as another source of potential immigration (interviews with two MOIA directors). The policy change started marginally when in 2005 MOIA increased subsidies to youth programmes for former residents. In 2006, for the first time in the examined period, returning residents were included in the ministry s budget statement (MOF ). In 2007, MOIA decided to launch a large campaign abroad, calling on emigrants to return, and also held employment fairs, mostly in the USA. In the same year the small returning residents unit was replaced by a new, large department. Subsequently, returning residents became MOIA s first priority, as indicated in its work plan (MOIA 2008). In 2008, a special programme was launched that included tax subsidies and health insurance payments for returning residents. In 2010, MOIA decided to equalize the benefits received by returning residents to those of new immigrants, including tax exemptions on various products and educational assistance for children (PMO 2010a). In addition, the budget allocated to returning residents grew from 3 m NIS in 2004 to 14 m in 2010 (MOF ) an increase much higher than the increase in the actual number of returnees during the same period (3,565 to 7,112; MOIA 2010a). Although not stated explicitly, the new policies targeted emigrants of higher socioeconomic status (Cohen 2009), in line with policies in other countries (Shachar 2006). As in immigration encouragement, all the new programmes and funds were added to instead of replacing existing integration services (MOF ). However, the change represented more than just adding policies, since the new policies gradually converted MOIA to the new purpose of assisting returning residents. At the end of the period examined, not only did returning residents become one of MOIA s main objectives but its official goal was changed from assistance to immigrants

11 1048 ILANA SHPAIZMAN to assistance to immigrants and returning residents, as evidenced by its work plans and website (MOF ; MOIA 2008, 2009a, 2010b; Environmental changes cannot by themselves explain this policy change. There were periods in the past when the number of immigrants was even lower (for example, in the mid-1980s) and MOIA did not engage in an active return policy. In addition, there was no significant change in the number of immigrants or emigrants from 2006 to Political interests also represent only a partial explanation, first because the minister in 2010, Sofa Landver, was from a party representing FSU immigrants, while most returning residents are from the USA, and second due to significant public opposition to the new policy in The fact that the minister at that time did not withdraw the programme demonstrates that her motives were more than just political. Moreover, as in immigration encouragement it is not clear why the change in function was accompanied by change in purpose in other words, why layering was accompanied by conversion. Here, too, ideational analysis can provide an answer. The positive attitude towards immigration and its important place in the national ethos have made emigration from Israel a negatively perceived phenomenon. This is evidenced by the respective Hebrew terms aliyah (going up) and yerida (going down). The problem definition at the beginning of the period examined was, to use the prevalent metaphors, waging war on the emigration epidemic, and most policy solutions ideas were aimed at preventing emigration. Incentives for returning residents were dismissed because it was believed that providing incentives to this population would only encourage further emigration. In addition, the decision makers thought that return depended on macro factors, such as the economic situation in Israel and abroad, so incentives would not be effective (CIAD 1989; Cohen 2009). In 2004 decision makers started to question these ideas. That year, the first official MOIA delegation visited emigrants in North America, making decision makers realize the potential and not only the negative aspects of this population. Moreover, one of the major decision makers in this field was a returning resident himself, and he believed that: This is a global world and people are moving from place to place. This is no reason to boycott them. (MOIA director) Consequently, from 2004 onwards, MOIA leaders and politicians from several parties began to point out that the situation had changed and that the government should do more than it presently did to bring back the emigrants and that the traditional discourse did not fit the new reality (Livni 2005; Boim 2006, 2007). Moreover, during CIAD held only one meeting on emigration prevention, compared to three or four in the previous year. All that time negative public attitudes towards emigrants persisted, as indicated by the objection to the 2008 programme expressed in the letters of complaint sent to the ministry (MOIA director). Hence it is more likely that decision makers arguments were earnest, mostly because they were made in CIAD meetings open to the public. From 2005 the problem definition changed from emigration prevention to promoting return. However, You must strike the balance between an immigrant who s never lived here and an Israeli citizen who s never left. You don t want to create an incentive for Israelis to leave Israel in order to return and get the benefits they wouldn t have gotten if they hadn t left. (MOIA director) In addition, the decision makers started to believe that return was a result of government policy and not only of macro factors. Policy solution ideas which focused on active

12 IDEAS AND INSTITUTIONAL CONVERSION 1049 encouragement were put forward. These ideas were not new, since the former head of the returning residents unit had expressed most of them publicly at CIAD meetings from the mid-1990s (Prigat 1993). However, at the beginning of the period examined her ideas were dismissed since assistance to returning residents was seen as a prize for those who had gone astray. Only when the ideational change began were new policies introduced. From 2006 decision makers started talking about the economic potential of returning residents, especially given their high skills and potentially rapid integration. The discourse on economic benefits changed the policy solution ideas from incentives as increasing emigration, to incentives as increasing return (Halfon 2007). In addition emigration was put into a global context of people moving from place to place, rather than in the ideological context of Zionism (Boim 2006, 2007). This corresponded to global ideas on immigration and returning residents (Shachar 2006). As a result decision makers shifted their attention from the negative aspects of emigration to the positive benefits of return migration. By 2007 this trend intensified, and the director-general said that MOIA is the only social ministry that produces money for the State of Israel: 18,000 returning emigrants added 2bn NIS to the GNP. According to our research, a returning emigrant creates about 100,000 NIS in taxes. (Erez Halfon, MOIA Director-General, CIAD, 19 December 2007) However, the status of returning residents was still somewhat inferior to that of new immigrants: We try to maintain some gap between what a returning resident gets and what a new immigrant gets, for normative reasons, but still we can do more to encourage them to come back. (Minister Zeev Boim, CIAD, 31 January 2007) Alongside the economic reasoning, this policy was also rationalized by expanding the public philosophy of immigration as a tool for nation-building. The returning residents policy was described as a national cause, with returnees presented as contributing to Israel s national strength (Landver 2010). In addition, decision makers constantly associated immigration encouragement and returning residents to give the impression that returnees were immigrants as well, and used positive and welcoming metaphors such as open doors and open arms (Halfon 2007; Apartzev 2010; Landver 2010). At the end of the period examined, when the rights of immigrants and returning emigrants were almost equalized, the latter s lower normative status was dismissed: For the first time we have equalized some of the benefits for immigrants and for returning residents I would like to say to each family: Come home, we want you here and you will get personal treatment just like every immigrant. (Minister Sofa Landver, CIAD, 10 November 2010) Thus, the change in problem definition from emigration prevention to encouraging return made decision makers seek different policy solution ideas of economic incentives and proactive initiatives. Their gradual implementation narrowed the actual as well as the normative gap between immigrants and returning residents. Once both populations were seen as equal it was only logical to expand MOIA goals from assisting immigrants to assisting returning residents as well. This expansion was possible not only because the new return programmes did not replace the old integration policies, but also because they both came to be legitimized by the same public philosophies. CONCLUSION This article has tried to gain a better understanding of gradual transformative change by examining conversion through layering in Israeli immigration policy in ,

13 1050 ILANA SHPAIZMAN and the role played by ideas and ideational change in that change. It is evident from the Israeli case that contrary to Mahoney and Thelen (2010), each combination of political and institutional factors is linked to more than one change mode. Specifically, the case study demonstrates that given a low level of discretion and strong status quo bias, institutions can be converted through layering. In addition, the case also expands our concept of gradual transformative change, since the combination of conversion and layering is qualitatively distinct from the manner in which Streeck and Thelen (2005) account for each of them separately. This is because, first, in conversion through layering the institutional aim is expanded rather than replaced. Second, in this pattern, the change is in both institutional practices and aims. In the Israeli setting, the layering of selective immigration programmes and active return policies not only transformed the Ministry of Immigrant Absorption s structure and function, but also gradually converted it to a new purpose. Consequently, change resulting from a combination of two modes can be seen as more extensive than change resulting from only one mode. Existing explanations for the evolvement of the various modes of change do not account for the evolution of this pattern, nor for its content or direction. Among other reasons, this is because they do not address ideational components important for understanding actors wider motivations and constraints. Thus, an analysis of ideas at different levels of generality and the interaction between those levels has been attempted here. The analysis reveals, first, that the content and direction of policy change correspond to those of ideational change, even when running counter to actors strategic interests or changes in electoral politics. Second, conversion thorough layering is a result of a change in problem definition and policy solution ideas, as well as of the top-down and bottom-up interactions between the two. Specifically, changes in problem definition ideas turn the decision makers attention to different policy solution ideas. The policies resulting from these ideas gradually expand the existing problem definition and so expand the policy goal. This expansion is possible when the new policies do not question the existing public philosophies in which the old policies are grounded, enabling them to reconcile multiple goods. Since gradual ideational change is quite extensive, affecting both the problem definition and policy solution ideas, policy change is extensive as well. Accordingly, this article also contributes to ideational research by providing empirical evidence to support theories of gradual ideational change and its role in gradual transformative changes. Thus, it helps move ideational research one more step beyond punctuated equilibrium. To conclude, this study has not only narrowed the gap between theory and empirical evidence by going beyond the existing typologies of the evolvement of gradual transformative change, but has also underlined the important role ideas play in gradual transformative change. Future studies should consider additional patterns of gradual transformative change and examine the conditions for their evolvement, paying particular attention to ideational dynamics. ACKNOWLEDGEMENTS The research leading to these results was supported by the Harry and Sylvia Hoffman Leadership and Responsibility Program. I wish to thank Sharon Gilad, Roni Holler, Sarah Anderson, and the three anonymous reviewers for Public Administration for their useful comments.

14 IDEAS AND INSTITUTIONAL CONVERSION 1051 REFERENCES Apartzev, D Report of General Director to the Committee of Immigration, Absorption and Diaspora Affairs Beach, D. and R.B. Pedersen Process-Tracing Methods: Foundations and Guidelines. Ann Arbor, MI: University of Michigan Press. Béland, D Ideas and Social Policy: An Institutionalist Perspective, Social Policy and Administration, 39, 1, Béland, D Ideas and Institutional Change in Social Security: Conversion, Layering, and Policy Drift, Social Science Quarterly, 88, 1, Béland, D. and R.H. Cox (eds) Ideas and Politics in Social Science Research. New York: Oxford University Press. Berman, S The Social Democratic Moment: Ideas and Politics in the Making of Interwar Europe. Cambridge, MA: Harvard University Press. Berman, S Ideational Theorizing in the Social Sciences Since Policy Paradigms, Social Learning, and the State, Governance, 26, 2, Blyth, M Great Transformations: Economic Ideas and Institutional Change in the Twentieth Century. Cambridge: Cambridge University Press. Boim, Z Report of the Minister of Absorption to the Committee of Immigration, Absorption and Diaspora Affairs Boim, Z Report of the Minister of Absorption, to the Committee of Immigration, Absorption and Diaspora Affairs Borukhov, E Immigrant Housing and its Impact on the Construction Industry, in E. Leshem and M. Sicron (eds), Profile of an Immigration Wave: The Absorption Process of Immigrants from the Former Soviet Union, Jerusalem: The Magnes Press, The Hebrew University, pp [Hebrew]. Burg, A Report of Jewish Agency Executive to the Committee of Immigration, Absorption and Diaspora Affairs. Knesset Archive, Jerusalem [Hebrew]. Campbell, J.L Institutional Analysis and the Role of Ideas in Political Economy, Theory and Society, 27, 3, Carstensen, M.B Ideas Are Not as Stable as Political Scientists Want Them to Be: A Theory of Incremental Ideational Change, Political Studies, 59, 3, CBS (Central Bureau of Statistics) Emigration of Israelis 2005 ( CIAD (Committee of Immigration, Absorption and Diaspora Affairs) Emigration Prevention. Knesset Archive, Jerusalem [Hebrew]. CIAD (Committee of Immigration Absorption and Diaspora Affairs) Immigration Encouragement ( protocols/heb/protocol_search.aspx [Hebrew]). Cohen, N Come Home, Be Professional: Ethno-Nationalism and Economic Rationalism in Israel s Return Migration Strategy, Immigrants & Minorities, 27, 1, Cox, R.H The Path-Dependency of an Idea: Why Scandinavian Welfare States Remain Distinct, Social Policy and Administration, 38, 2, Della Pergola, S World Jewish Population, Mandell L. Berman Institute North American Jewish Data Bank, Connecticut ( Denitz, S Report of Jewish Agency Executive to the Committee of Immigration, Absorption and Diaspora Affairs. Knesset Archive, Jerusalem [Hebrew]. Edelshtain, Y Report of the Minister of Absorption to the Committee of Immigration, Absorption and Diaspora Affairs Edelshtain, Y Report of the Minister of Absorption to the Committee of Immigration, Absorption and Diaspora Affairs Falleti, T Infiltrating the State: The Evolution of Health Care Reforms in Brazil, , in J. Mahoney and K.A. Thelen (eds), Explaining Institutional Change: Ambiguity, Agency, and Power. New York: Cambridge University Press, pp Gal, J Immigration and the Categorical Welfare State in Israel, Social Service Review, 82, 4, George, A.L. and A. Bennett Case Studies and Theory Development in the Social Sciences. Cambridge, MA: MIT Press. Hacker, J.S Privatizing Risk without Privatizing the Welfare State: The Hidden Politics of Social Policy Retrenchment in the United States, American Political Science Review, 98, 2, Hacker, J.S., K.A. Thelen and P. Pierson Drift and Conversion: Hidden Faces of Institutional Change, in APSA 2013 Annual Meeting Paper, Chicago ( Hacohen, D Immigrants in Turmoil: The Great Wave of Immigration to Israel and its Absorption, New York: Syracuse University Press. Halfon, E Report of Ministry of Immigrant Absorption Director-General to the Committee of Immigration, Absorption and Diaspora Affairs