Appendix I: Case Discussions

COLOMBIA (1958-)

The administration was ineffective throughout the period. Territorial penetration was never reached. Colombia was traditionally divided in four regions not well integrated by infrastructure: Caribbean Coast, Antiquoia, Cauca, and the Eastern Zone. This remained so even after La Violencia (Safford and Palacios 2002: 297; Manwaring 2002: 77). Not least, this is because of the country’s mountainous geography and the historical inability to integrate peasants into any state-building project and the preserved powers of landed elites administering in the absence of the state (Martz 1997: 79, 146, 165-166; Kline 2003: 162-163; McDougall 2009: 328). Uribe’s strong anti-insurgency campaign in the 2000s greatly extended state presence beyond the coastal regions and highland region around Bogota. However, the state is still virtually absent in large areas and a severe disjuncture between de jure state territories and de facto rule by guerillas or caciques paid to be loyal (Bejarano and Pizarro 2004: 105; BTI 2006: 4; BTI 2010: 7; BTI 2012: 6-7).

Meritocracy was never established. The National Front that came to dominate politics in 1958 marked a renewal of the clientelist pattern of recruitment to the public sector to a more systematic spoils system between the two major parties of the Conservatives and the Liberals (Peeler 1985: 58-59). While the judiciary was traditionally fairly independent (Taylor 2009a: 31), the state administration was basically patrimonial: public services and jobs were distributed according to political powers and the need for cooptation of regional and economic elites (Martz 1997: 40). Through the 1970s and 1980s, administrative reform efforts waxed and waned with the willingness of presidents to circumvent regional elites but no reforms were quite successful enough to institutionalize meritocracy beyond a few agencies (Martz 1997: 120, 127, 146, 165-166). This pattern of politicization and clientelism continued through the 1990s and the Uribe presidency (Hernandez 2015: 35, 90) while the judiciary experienced renewed problems of politicization as guerillas pushed the state towards greater levels of repression (Alesina 2000: 6).

Responsiveness was also weak throughout the period. Capture of the administration by regional elites and caciques was widespread and the rule of law varied considerably between regions on this account (Martz 1997: 80). This dependence on societal forces for implementation also nurtured high levels of corruption while low levels of professionalism caused weak public service delivery (Martz 1997: 80; Hernandez 2015: 90-94).

COSTA RICA (1949-)

The administration was effective throughout the period. Territorial penetration was obtained in the late 19^th century as resource supremacy. The period entailed heavy investments in infrastructure and education which broadened state presence to the countryside (Rankin 2012: 65-66). Despite accusations of inefficiency in the public sector of post-war Costa Rica, these emerged because of overload of public servants and rules rather than their absence (see BTI 2003: 3).

The Costa Rican bureaucracy is an exception in the patrimonial Latin America. The constitution of 1948 clearly separated the judiciary from politics. This was respected in subsequent years by freeing the judiciary from executive and legislative control (Biesanz, Biesanz, and Biesanz 1999: 65-67; Wilson and Fernandez 2015). The Spanish colonial history surprisingly gave way to a tradition of rule of law in Costa Rican politics. The Spanish had to work the land themselves since the indigenous population fled to the forest. In this way, individualism and protection from the state became ingrained from the early years of independence (Knox 2001: 137-150). Beyond the judiciary, meritocracy was firmly installed by a civil service law in 1949 (Peeler 1985: 115-116; Ancochea 2005: 695). This law was only changed in 1998 by opening up the upper level director offices for non-civil servants but this did not fundamentally alter the meritocratic traits of the bureaucracy (Ancochea 2005: 713). The law achieved practical significance by a series of purges in 1948 by the new political elites, thus dismantling the old oligarchic elites, and the creation of a number of autonomous state agencies (Peeler 1985: 63-64; Biesanz, Biesanz, and Biesanz 1999: 68).

The bureaucracy was accused of being corrupt and inefficient at times due to its oversize and bewildering mass of agencies (Biesanz, Biesanz, and Biesanz 1999: 65; BTI 2003: 3). This stemmed back to the state expansion in 1948 (Biesanz, Biesanz, and Biesanz 1999: 34). However, the vast majority of studies adhere to the dominant version of Costa Rica as an island of efficiency and effectiveness in Latin America. Notably, it is said that the founding junta of 1948 managed to streamline the bureaucracy and get rid of the most autocratic elements in the purges of that year (Rankin 2012: 114-115). Beyond this, the bureaucracy was remarkably effective in delivering public goods to the people; this efficiency and willingness to cooperate with the executive was only boasted by the New Public Management reforms of the 1990s (Wilson and Fernandez 2015). In a couple of revealing analyses, Denton (1969: 420-422) and Taylor (1995: 1072) further point out that the bureaucracy’s budgets and decisions follow the executive and the party that won the last election. Civil servants consider themselves as apolitical instruments of the government.

CROATIA (2000-)

The administration was ineffective throughout the period. The first modernized administrative structure in Croatia was established by the Austro-Hungarian Empire from 1850 to 1865; the feudal state was dismantled (Goldstein 1999: 72; Magas 2007: 255). But in 1991, “the earlier well-regulated and sophisticated Communist administration began to be replaced by a confused and chaotic new order” (Magas 2007: 657). However, with the reclaiming of territories from the mid-1990s the government of 2000 was in charge of a fully penetrating administration (BTI 2003: 4).

Croatia belongs to the Eastern European administrative tradition affected by the Hungarian traditions, albeit with notable elements of the Austrian bureaucratic tradition (Goldstein 1999: 72; Møller and Skaaning 2010: 334, 340). In the judicial sphere, inspirations from the Ottoman Empire were prevalent until comparatively late (the beginning of the 19^th century) when Roman law was gradually introduced (Dallara 2014: 32). Under Tudjman in the 1990s, government officials often depended on personal loyalty to Tudjman for job security – politicization of the state administration and judiciary was almost complete (Goldstein 1999: 259; Ramet 2010a: 259; Dallara 2014: 44). The first administrative reforms were adopted in 2001 which marked the legislative end of the personalist politicization era. In 2004, a judicial academy was installed to ensure the required amount of qualified candidates (Dallara 2014: 48). Upon failed implementation, a renewed State Administration Reform Strategy of 2008 was adopted but the system of recruiting civil servants remained decentralized with low thresholds for passing the exam opening up for politicization in later phases of the screening process, and turnover rates of civil servants between elections remained high (15-20 %) (Meyer-Sahling 2012: 34, 37, 43). This also applies to the judicial sphere (Dallara 2014: 31).

As in many other Balkan states, the part Ottoman, part Hungarian administrative legacy included a pattern of corruption and patrimonialism at local and central levels of the administration (Møller and Skaaning 2010: 334, 340). Whereas politicization from above was most problematic during the era of Tudjman, the gravest problem after 2000 was politicization from below in which judges and civil servants could be bribed or sacked implementation on their own (Blitz 2008: 132-133). From 2000, institutional strength was improved to combat corruption and comply with EU demands but a lack of political will and instruments prevailed through to 2006 at least. By 2010, still 57 % of all enterprises were exposed to corruption in public contracting via civil servant bribes. Only when Prime Minister Sanader resigned on accusations of corruption in 2009, real judicial reform took off (Dallara 2014: 48). But corruption and cronyism is still prevalent (Ramet 2010a: 260). As Meyer-Sahling (2012: 31) notes, inspection, control, and enforcement capacity was extremely weak in all administrative layers throughout the 2000s.

CZECH REPUBLIC (1993-)

Administrative effectiveness prevailed throughout the period. Territorial penetration, through a so-called ‘differentiated administrative structure’, was deemed stable in all BTI-reports of the 2000s (see e.g. BTI 2006: 5). The Czech Republic merely continued administrating affairs as usual from the Czechoslovakian system (see Simon 2004: 23).

In much the same way, meritocracy and responsiveness were reestablished as the ‘Slovakian laggard’ in this regard was ‘cut loose’ in 1993 (Pehe 1993; for contrasts between Slovakia and Czech Republic, see e.g. Adam 2004: 8; O’Dwyer 2006: 82-83). During communism, the Czechoslovak system was ‘bureaucratic-authoritarian’ owing to the bureaucratic qualities of the Czech half. This continued after 1993 (Møller and Skaaning 2010: 325, 338). While judicial independence was secured legally and in practice from 1993 (Pehe 1993), the relations and separation between politics and administration in the ordinary ministries were surprisingly much less codified. In fact, as Meyer-Sahling (2009. 21) puts it, there was no formal distinction between the political and administrative levels leaving much room for politicization. However, he also asserts us that in practice, such a distinction could be drawn from the deputy minister level down in which job vacancies were mostly advertised and open competition institutionalized (Meyer-Sahling 2009: 21, 25). O’Dwyer (2006: 68) similarly finds that the level of actual politicization was in fact much lower than even in Poland, putting the Czech Republic in the top rank of meritocracy among post-communist countries.

Much the same story can be told about responsiveness. Some administrative corruption remained. Yet, the Czech administration is consistently referred to as one of the least corrupt countries in Eastern Europe (see e.g. Møller and Skaaning 2010: 341). As pointed out by Adam (2004: 8), the Czech state achieved a relatively high degree of autonomy from society.

CZECHOSLOVAKIA (1990-1992)

The administration was effective from 1990 but became ineffective in 1992. Territorial penetration was no problem in 1989 and beyond as Havel could rely on pre-WWII federal lines of command between the Czech and Slovak entities and communist administrative structures (see e.g. Simon 2004: 9; Heimann 2009: 309). As in the interwar Czechoslovak republic, political power resided in the Czech part. This was reestablished by the constitution of 1960 which deprived Slovaks of political power and installed federal control from Prague over Slovak ministries and agencies (Simon 2004: 9). Thus, the Czech traditions of Austro-German meritocratic administration could be reinstalled in 1990 without much interference from Slovakia. The judiciary was guaranteed independent and ministries abandoned Nomenklatura and purged communist high-level officials already in 1989 (Pehe 1993; Meyer-Sahling 2004: 81). The President had the power to appoint Supreme Court judges but this practice only affected the Supreme Court level (Kosar 2013: 16-17). The purging of communists from the administration also served to continue its responsiveness to the new governments under Havel. However, from the elections in June 1992 the federal government system diminished sharply in authority. From July, administrative authority de facto functioned from two contradictory poles in Bratislava and Prague despite the formal continuance of the federation until January 1993 (Simon 2004: 17-19). Also, this meant that political power moved away from Prague to the more patrimonial administration of Slovakia. Therefore, I consider responsiveness and meritocracy as weak in 1992.

DOMINICAN REPUBLIC (1966-)

The administration was ineffective throughout the period. In the late 19^th century, a strong bureaucratic apparatus and a modern infrastructure were built. As state power extended to the small villages during the Trujillo regime, territorial penetration was obtained (Betances 1995: 19; Fearon and Laitin 2005: 3). Important infrastructural investments were made in the 1960s but these only expanded the already established state control (see Conaghan and Espinal 1990: 560). This continued under democracy despite the heightened tensions along the border with Haiti and the empowerment of drug cartels (BTI 2003: 4; BTI 2012: 5).

Meritocracy was never installed. The Trujillo regime was personalist-patrimonial and Balaguer basically continued this line of recruitment after 1966 under a more competitive but still oligarchic rule (Kearney 1986: 145; Betances 1995: 100). Central agencies were and remained heavily politicized and characterized by hiring of family members (Ancochea 2005: 696). A new constitution in 1997 gave independent powers to the judiciary but the rest of the bureaucracy remained outside any civil service law (Ancochea 2005: 711). Reform attempts in the 1980s and late 1990s sought to change this but the resulting civil service statutes were never implemented (Kearney 1986: 144; BTI 2003: 4; Ancochea 2005: 715; BTI 2012: 6).

Responsiveness also suffered under the legacy of clientelism, the unfulfilled but instead competitive form of politicization, and the power of oligarchic families to circumvent state rule (Kearney 1986: 146; Sanchez and Lozano 2012). The state was described as highly corrupt, inefficient, and incapable of implementing rules consistently across the territory (Kearney 1986: 146; Ancochea 2005: 696). Laws were sometimes only spottily enforced on the ground (BTI 2003: 4).

EAST TIMOR (2002-)

The administration was ineffective throughout the period. Territorial penetration suffered a the state-building project never succeeded in mediating between central and local interests. After 2000, people thus gradually started withdrawing to the localities, sukus, which undermined any attempt at widening the existence of the state to the inhospitable mountainous forests. There was no working bureaucracy when Indonesia withdrew in 1999 and the peacekeeping force could not keep its momentum of the development in 2000 (Croissant 2008: 662-663). By 2010, water and road networks remained in very poor condition and the infrastructure was generally under prioritized until 2010 (Nixon 2012: 153).

Meritocracy was one of the few state traits that were successfully installed from 2000. From 2001, most recruits to the new bureaucracy were educated from the academies set up by the UN (Federer 2005: 102). This also involved the judiciary, albeit educational quality was low (Nixon 2012: 161). Despite the risk of the return of traditional patronage relations, a civil service commission remained firmly in place and in function throughout the 2000s (Nixon 2012: 158).

As there was no stable connection between the local and the central institutions of governance, public service delivery locally was greatly hampered (Beauvais 2010: 1106). The weak education led to a pool of badly traumatized and committed to public service and with poor work ethics. In the judiciary, the lack of qualified personnel was almost complete and led to a very poorly functioning judiciary (Federer 2005: 102-103). Rampant corruption and problems of hierarchical management in the ministries also characterized the administration (Nixon 2012: 157-159).

ECUADOR (1979-2000; 2003-)

The administration was ineffective throughout the period. Fundamental infrastructures built in the early 20^th century lasted throughout the period (BTI 2003: 4; Bowen 2015: 91-92). There was a strong tradition for presidential powers in appointments to the state apparatus dominating in the 1980s and 1990s (Mangelsdorf and Reeves 1989: 195). Still in the early 2000s, the president appointed some mid-level servants and all ministers and regional governors (who subsequently appointed regional officials). Thus, extensive politicization existed (Araujo et al. 2004: 21; Ruiz-Giraldo 2004; BTI 2008: 7). The state reached its highest levels of autonomy under Correa from 2007 as the president withdrew from control of key agencies (Bowen 2015: 100). However, judges remained high insecure in job appointments and tenure. This included highly unstable Supreme Court teams as evidenced in an analysis stretching from 1979 to 2009 (see Basabe-Serrano 2012). Thus, despite Correa’s moves toward meritocracy, it remained ineffective at an overall level due to the politicization of the judiciary.

Responsiveness suffered from legendary problems of poor performance and high levels of corruption in the Andean states, including Ecuador. In the early 2000s, corruption increased further, mostly at central levels, and the quality of regulations, the rule of law, and effectiveness in implementation deteriorated (Seligson and Recanatini 2003: 411; BTI 2008: 7). This continued being a severe problem under Correa despite his attempts to build a more effective state. Instead, the state came close to breaking down in terms of delivering even the most key public goods, and business elites achieved ever more privileged access to state resources despite Correa’s socialist stance (Bowen 2015: 85, 93-94).
EL SALVADOR (1984-)

The administration was ineffective throughout the period. State-building in the 19^th century expanded the territory controlled by the oligarchic coffee elite to large areas of the rain forest and countryside (Dunkerley 1982: 9). However, according to BTI (2003: 3; 2006: 4) the state’s presence was highly limited in many rural areas and wholly absent some places in the deep forest. Thus, I assume territorial penetration to be weak throughout the period.

Meritocracy was never close to being established. According to Echebarria and Cortazar (2007: 139), El Salvador ranked second-lowest among Latin American countries on the degree of meritocracy in the public administration. In classifying the administration qualitatively, they further coded it as unambiguously clientelist (Echebarria and Cortazar 2007: 148-151). Also traditionally, the oligarchic elite assured political influence by placing family and friends in the bureaucracy and the judiciary. Even though the peace accord formally established judicial independence, the judiciary itself rejected this and remained under the control of the major parties (Negroponte 2012: 18, 158-159; Wade 2016: 16-18). Changes in the Supreme Court in 1994 pluralized the hiring process in the judiciary but did not depoliticize it (Call 2003: 853; Stanley 2006: 111).

Unresponsiveness was similarly dominant in the implementation of policies. Echebarria and Cortazar (2007: 141) characterize the functional capacity of the bureaucracy as low and the technical capacity as weak. More specifically, it was the routine circumvention and traditional powers of the oligarchy interfering in banking affairs, economic management, and judicial reforms that caused unresponsiveness (see Negroponte 2012: 17-21; Wade 2016: 16-18). Corruption among judges was pervasive (Call 2003: 855-856; Stanley 2006: 111).

ESTONIA (1991-)

The administration was effective from 1996. Given the Soviet legacy and the small territory, penetration by administrative institutions was strong at independence. This continued throughout the spell (BTI 2003: 4).

In terms of meritocracy, Estonia is a borderline case. Mired in the peculiar Soviet legacy of absolute politicization of the bureaucracy, Nomenklatura, the pre-communist legacies of Scandinavian and German administrative influences nevertheless took a hold early on in the 1990s (Vanagunas 1999: 222-230). Kitschelt qualified communist Estonia as ‘national-accommodative’ – the relatively most law-governed type of communist administration with continuance into the post-communist era (see Møller and Skaaning 2010: 325, 338). A civil service law was passed in 1995, and a public service act was enacted from 1 January 1996 which abandoned the patronage system and introduced meritocracy by defining civil servants as regulated employees with high job security, recruited in competitive and open exams (Vanagunas 1999: 213; Randma 2001: 43). Afterwards, the practice of hiring and firing has been criticized by scholars (e.g. Drengsgaard with Hansen 2003: 17-19, 43; Palidauskaite, Pevkur, and Reinholde 2010: 47-49) pointing to 15 % yearly turnover rate in ministries, a lack of transparency in recruitment and promotion, and legislative openness to politicization. However, these scholars also ensure that meritocracy is generally strongly implemented (see also Meyer-Sahling 2009: 26) with politicization being the exception when the small country lacks qualified personnel. Judicial independence was secured from the moment of independence (Kasemets 2012).

The responsiveness of the administration to government decrees, policies, and orders was generally strong from the moment of independence. Being a small country with a small but dedicated staff the administration functioned relatively efficiently (Drengsgaard with Hansen 2003: 17-18). Despite some persistent problems of corruption, the state was basically autonomous of society and was able to prevent the capture of interest groups in the transition phase (Drengsgaard with Hansen 2003: 32, 45-49; Møller and Skaaning 2010: 340). It was considered one of the success stories of governance in the post-communist area (BTI 2012: 5; Kasemets 2012).

GAMBIA (1972-1994)

The administration was ineffective throughout the period. Penetration of the small territory is indicated by Davis, Hulme, and Woodhouse’s (1994) assertion that basic services did not reach some rural areas because of lack of funding – that is, not because of the mere absence of administrative structures.

There is not much literature on the existence or not of meritocracy. The few indications point toward the existence of a British-inspired meritocracy. As Sallah (1990) asserts us, Western educated elites of Gambian descent came to rule the country under colonialism and beyond. In 1997, the constitution reaffirmed the British system with a Public Service Commission advising the president on appointment and firing of judges and civil servants (Senghore 2010: 222-223).

The high-level civil service was as autonomous as it was corrupt and self-serving, colluding with business interests (Sallah 1990: 643; Wiseman 1996). Corruption was so frequent that the civil service was referred to as the ‘Banjul mafia’ (Sallah 1990: 643). As Yeebo (1995: 8, 22-23) notes, the governments of Gambia in fact originally had noble motives of development for the people but this never succeeded because of a predatory state apparatus working by bribery, nepotism, and favoritism.

GEORGIA (2004-)

The administration was ineffective throughout the period. Although Saakashvili increased state presence in Ajara and the Armenian populated regions, South Ossetia and Abkhazia remained outside state jurisdiction (Nodia 2005: 44-45; BTI 2006: 4). These regions ran their own administrations or rather a gathering of patron-client ties set up where, frequently, central authorities failed (George 2009: 132; Aphrasidze and Siroky 2010: 131). But even inside the nominally controlled area of the Georgian state, penetration remained weak (George 2009: 96; Jones 2013: 11).

The personnel management system in Georgia was indebted with a strong Soviet legacy being a colonial periphery of the union (Møller and Skaaning 2010: 325). The state apparatus had not been modernized to the same degree as many other Soviet Republics being managed by a ‘divide-and-rule’ logic (Jones 2013: 9). Under Shevardnadze, who was President from 1995 to 2003, the state was highly personalist, an extension of Shevardnadze’s personal and political powers into the judiciary and ministerial bureaucracy by means of hiring personal and political loyals (Timm 2012: 169; Jones 2013: 170). Political changes aside, the system of patrimonial politicization continued under Saakashvili. His party, the United National Movement, was the single strong party and thus could penetrate the entire bureaucratic apparatus demanding complete loyalty to the party (Timm 2012: 175). But several elements hindered meritocracy: family connections (nepotistic hirings), oligarchic rule (hiring due to class interests), and cronyism (personalist-rational hirings) (Charkviani 2013: 140). In turn, judicial independence and oversight was also weak (Jones 2013: 110-111). Under the banner of extending and strengthening vertical control and state capacities, meritocracy was never seen as advantageous or a political necessity and was thus breached by reforms in 2003 and 2004 (O’Shea 2014: 223-227). In 2013, meritocratic recruitment was still rare (Parrado 2014: 60).