Appendix I: Case Discussions

The Georgian state was traditionally highly corrupt, ingrained in patron-client relations hampering central governmental control down through all layers of the administration (Møller and Skaaning 2010: 325; Jones 2013: 9). This merely continued in the 1990s and early 2000s under Shevardnadze (George 2009: 125; Timm 2012: 169; Jones 2013: 134). Tellingly, two studies published in 2005 (Brøgger et al. 2005: 18, 42; Nodia 2005: 66-67) report of a significant yet still insufficient change in government control and effectiveness. The Rose Revolution in 2004 marked the start of a massive state-building effort in which vertical control was sharply improved but, as indicated, meritocracy was sacrificed (Charkviani 2013: 141-142). The literature is thus full of unequivocal praises of the sharp reduction in corruption, the strengthening or government-periphery controls, and presidential capacity to enforce policies (see e.g. George 2009: 169-170; Aphrasidze and Siroky 2010: 122; Timm 2012: 174; Jones 2013: 166, 175). The civil service council set up in 2004 and the law on local self-government in 2005 particularly improved these state capacities (Jones 2013: 166, 175). BTI also pointed to 2004-2005 as the years when government effectiveness decisively increased and that this development continued through 2010 (BTI 2006: 3; BTI 2008: 7; BTI 2010: 6; BTI 2012: 6-7). In sum, responsiveness was achieved from 2005.

GHANA (1997-)

The administration was ineffective throughout the period. Law and order did not extend to the entire territory as traditional law overrode state law (BTI 2003: 3-4). The state was “hardly present” in many rural areas (BTI 2006: 4; see also BTI 2008: 8). With these observations, I code territorial penetration as absent.

Meritocracy was, on balance, in place. The judiciary was reformed through constitutional change in 1992 which reassured its independence from political interference in hiring and firing (Quashigah 2016: 227-237). Whereas the judiciary was firmly autonomous throughout the democratic period (Abdulai and Crawford 2010: 41), the rest of the civil service was more ambiguously protected against political interference. Reports point out that the lower levels were often politicized by party patronage and concerns of ethnicity (Ayee 2001: 4). However, the literature more than not describes how merit was the principle of recruitment in more than half of all civil service positions, and that merit was the dominating principle on an overall level – the default option (Ayee 2001: 24, 31-32; Stevens and Teggemann 2003: 54).

Under the democratic regime, Ghana was more developmentalist than most African countries. Reforms dismantled the oversized and inefficient bureaucracy in the 1980s (Ayee 2001: 2-4). However, in assessments of administrative effectiveness most accounts focus on the fact that the impressive reforms were not continued in the 1990s (Stevens and Teggemann 2003: 48). Thus, the legacy of a lacking public service motivation and excessive corruption among, particularly, lower level and local civil servants continued. Red-tape, continued over-staffing, poor working conditions and poor planning, and a peculiar big man rule diverting attention away from government policies toward local chiefs proliferated – summed up in the notion of weak implementation ability (Amponsah 2007: 120-121; Abdulai and Crawford 2010: 42-43; McCauley 2012). In turn, unresponsiveness was the rule rather than the exception.

GUATEMALA (1986-)

The administration was ineffective throughout the period. During the civil war, the forests and countryside were abandoned places or run by local warlords. By 2007, six of 22 departments were run by criminal gangs (Brands 2010: 10; see also BTI 2003: 3; BTI 2008: 5; BTI 2012: 5). In turn, territorial penetration was never achieved.

The degree of meritocracy was among the lowest in Latin America relying on a clientelist tradition. There was little inclination to reform the judiciary to become more autonomous of political pressures. Judges and servants were constantly removed for political reasons (Sousa 2007: 92-93; Echebarria and Cortazar 2007: 139). The peace accord attempted to install judicial independence but largely failed since the president remained capable of packing the courts (Sieder 2003: 141). Still in the 2000s, interference in administrative hirings caused a severe lack of professionalism among civil servants (BTI 2006: 5).

Responsiveness also suffered as in El Salvador due to societal pressures which put constant pressure on judges and civil servants. The weak autonomy of the state administration enabled state capture by private interests (Sousa 2007: 96-97). This clientelist bureaucracy (Echebarria and Cortazar 2007: 148) was characterized by excessive corruption and infiltration by criminal gangs (Sieder 2003: 146-147; BTI 2006: 5; Brands 2010: v).

GUINEA-BISSAU (1994-1998)

The administration was ineffective throughout the period. The Portuguese only left rudimentary administrative structures (Embaló 2012: 255). Later, the state still lacked infrastructural power (Kovsted and Tarp 1999: 6). In fact, the state was described, as a failed state, close to if not exactly anarchical (Bybee 2011: 264; Havik 2012: 35). Thus, it seems fair to code territorial penetration as absent in the 1990s.

According to Kovsted and Tarp (1999: 9), the party, PAIGC, was synonymous with the state through the democratic years – the state was a “composite of distinct groups owing their primary allegiance to a handful of competing leaders”. Tellingly, the first democratic elections of 1994 opened new opportunities for political elects to reconstruct the state which paralyzed it (Havik 2012: 65).

This distinct pattern of politicization entailed shifting loyalties of bureaucrats which meant different degrees of effectiveness from agency to agency and dependent on the principal approaching them (Kovsted and Tarp 1999: 9). Also, the label of a failed state, which was given since independence, implied limited capacity of implementation and widespread capture by private actors such as narcotics traders (Bybee 2011: 264; Havik 2012: 35). Thus, responsiveness was hampered.

HONDURAS (1982-2009; 2010-)

The administration was ineffective throughout the period. As in other parts of Central America, some areas, albeit rather small, were not covered by any state agencies or control devices (BTI 2008: 7). As I demand full coverage, territorial penetration was not obtained.

Politicization and a traditional spoils system were deeply ingrained in Honduran politics. Politicians were in constant struggle for patronage access to the state apparatus (Boussard 2003: 184-185). This is reflected in Echebarria and Cortazar’s (2007: 139) analysis in which Honduras ranked on a par with El Salvador with a very low degree of meritocracy. Much the same could be said of the judiciary which was routinely packed by presidents or less systematically interfered with in promotions and appointments (Morris 1984: 67; Sousa 2007: 107; Farr 2010: 314-315).

Similarly, patronage resulted in outside interference from societal actors as a routine trait such as when economic actors captured and interfered in judgments by the courts (Boussard 2003: 183; Sousa 2007: 97). The judicial and administrative spheres were thus “largely ineffective” (BTI 2003: 4) and “highly dysfunctional” (Echebarria and Cortazar 2007: 141). Corruption and nepotism were widespread (Boussard 2003: 193). In turn, responsiveness was very weak if meaningfully existing at all.

HUNGARY (1990-)

The administration was effective from 1992 but became ineffective in 2010. Territorial penetration applied to Hungary at independence in 1990 much like in Poland and the remaining Eastern European post-communist countries (for an assessment of the 2000s, see BTI 2008: 5). Among the Eastern European countries, Hungary had a relatively strong legacy of meritocracy at independence in 1990. In the interwar period, separate decrees were adopted to ensure basic job security of civil servants. The relatively professional civil service was interrupted in the 1950s and 1960s by the communist takeover but the establishment of the School of Public Administration in the 1970s marked the return of professionalism in the administration (György 1999: 134-135).

In the literature that I have encountered, the communist legacy does not seem to have been very strong in the 1990s. On the other hand, neither account points to a quick decommissioning of communist patterns of recruitment (for an appreciation of the borderline status of Hungary, see Møller and Skaaning 2010: 332). Whereas judicial independence was secured from 1989 (Fleck 2012: 793-801), the literature agrees that the year 1992 is essential for understanding the pattern of recruitment in the post-communist period (see e.g. György 1999: 143-144; Meyer-Sahling 2001: 962; Gajduschek 2007: 346). The Civil Service Act of May 1992 served to separate the administrative from political positions and ingrain the norm of political neutrality of civil servants. A number of provisions, such as job security for civil servants and the insulation of under-secretaries as politically hired, came to guard against politicization (György 1999: 143-144). According to Meyer-Sahling (2001: 962), this law was only amended slightly in 1997 and 2001 as it functioned well as a regulation against party patronage. Gajduschek (2007: 346) asserted us that until the time of his writing, a meritocratic system had been successfully in place after the 1992 law (see also Ilonszki, Johannsen, and Kas 2002: 22; BTI 2003: 3). In the late 2000s, the degree of politicization was high at the top levels of the administration, among the so-called Administrative and Deputy State Secretaries. However, as with the Politische Beamte in Germany and elsewhere in Eastern Europe these positions were deliberately separated from the rest of the civil service as intermediate positions between the administrative and political levels (see Meyer-Sahling 2008: 9-10).

As in Poland with the government turnover to the Law and Justice Party in 2006, the election Orban as Prime Minister of Hungary in 2010 changed the civil service system. In April 2010, he reformed the 1992 law making it possible to release civil servants without reason (Bozoki 2011: 652; Rupnik 2012: 133-134). He thus removed most of the limits on executive discretion in hiring and firing. The entire senior bureaucracy became politicized, hired as political or personal loyals of Orban from 2010 (Staronova and Gajduschek 2013: 8), and the Constitutional Court and central bank became hampered with politicization (Rupnik 2012: 133-134).

Regarding responsiveness, Hungary was, as most other post-communist Eastern European countries, a borderline case. However, its basic efficiency in implementation resembled more that of Poland and the Baltics than that of Ukraine and the Balkans (Møller and Skaaning 2010: 341). Corruption levels were clearly higher than in most Western European countries but was still basically controlled (Ilonszki, Johannsen, and Kas 2002: 25; Dunay 2005: 17; Bozoki 2011). The administration was described a “well-oiled machine” (Hajnal 2008: 133) and as “effective and credible” (Bozoki 2011). In 2010, the state was certainly ‘captured’ but more in the sense of being politicized than corrupted by interests outside the government.
INDIA (1950-)

The administration was effective throughout the period. Territorial penetration was no problem. The British had built quite an extensive bureaucratic apparatus. After independence, industrial infrastructure helped consolidating state presence (Kohli 1987: 65).

The British installed a meritocracy led by a Public Service Commission during colonialism (BTI 2003: 3; Kenny 2015: 407). Similarly, the British Raj involved independent courts (Beller 1983: 515). At independence, state autonomy was reduced somewhat at the lower levels. Also, quotas were introduced to further representativeness (Kohli 1987: 61; BTI 2003: 3). Yet, this did not fundamentally alter the meritocraticness of the system. The Public Service Commission stayed in power, serving a single and strong civil service with a senior level selected through merits (Dwivedi, Jain, and Dua 1989: 253). Also, state autonomy was never dismantled so much as to remove the bureaucracy’s strong ability to cooperate independently with business interests (Kohli 1987: 61). Lastly, judicial independence was preserved throughout the period (Singh 2016).

The bureaucracy remained a strong organization, autonomous of society but capable of effective cooperation upwards and downwards – as in the notion of embedded autonomy (Kohli 1987: 61, 65). Not least this responsiveness stemmed from the successfully adopted ministerial system with hierarchically managed and specialized secretaries under executive control (Basu 1989: 215; Jain 2001: 1301). The bureaucratic servants remained in office through 1947-1948 which resulted in a continuation of capacity at all levels in the bureaucracy (Dwivedi, Jain, and Dua 1989: 253). The bureaucracy grew immensely after independence but piecemeal reforms ensured continued responsiveness in the judiciary as well as in the rest of the civil service (Basu 1989: 212-213; Dwivedi, Jain, and Dua 1989: 259). Corruption was a major problem but mostly at the political level (Quah 2008: 242).

INDONESIA (1999-)

The administration was ineffective throughout the period. Territorial penetration was established. Although weak, basic infrastructure extended throughout the territory (BTI 2006: 8). Meritocracy in the judiciary was hampered in the 2000s because the old elites managed to maintain privileges via the manipulation of courts, nepotism and favoritism in hiring and firing (Tahyar 2012: 3-4). True, a number of state agencies emerged in these years with ascerted meritocratic principles (Schütte 2011). But recruitment and personnel affairs were still mismanaged and engrained with unsystematic rules and norms of behavior (McLeod 2005: 369, 379). Recruitment rules were highly heterogenous (Kasim 2013: 18-19). In many agencies, public offices were bought and sold in the process of decentralization according to old patrimonial logics (Blunt, Turner, and Lindroth 2012: 64-65).

By the late 2000s, decentralization efforts finally seemed to work out in the sense that civil servants were successfully deployed to local offices. However, no monitoring agencies of the new local administrators were established which provided a high risk of unresponsiveness in a highly heterogenous countryside (BTI 2008: 9). The regional autonomies in fact proved detrimental for implementation (see BTI 2006: 8). Under democracy, the old bonds of loyalty to Suharto disappeared and the loyalty of the servants was completely lost and never reestabslihed (BTI 2003: 5; Webber 2006). Indiscipline and tolerance of corruption was widespread beyond the central-level ministerial departments (McLeod 2005: 379; Blunt, Turner, and Lindroth 2012: 70).
ISRAEL (1948-)

The administration was effective from 1959. There was indeed chaos in many areas of Israel in the first years of its life. However, administrative structures were fundamentally in place as they were adopted from the British (Samuel 1951: 229-230).

The early years of administrative reform downplayed professionalism and valued partisanship instead. Basic loyalty to the party in government and the cause of an Israeli state was required for civil servants and patronage was used as such in hirings and firings already from the time of the British mandate before 1948 and continuing afterwards (Nachmias and Arbel-Ganz 2005: 286). A 1959 reform changed this substantially by installing meritocracy. A civil service commission was established formulating tender regulations which stated that no candidate could be aksed of his/her political preferences in hiring processes (Nachmias and Arbel-Ganz 2005: 287). From this reform, Israel achieved its image of a state with a highly autonomous, meritocratic bureaucracy, highly cohesive, set in a modern ministerial system, and with a strongly independent judiciary (Samuel 1961: 193; Levi-Faur 1998: 70).

All ministries were highly regulated via internal controls to make it a fundamentally responsive and cooperative organ for the parliamentary system (Samuel 1961: 191). A particular tradition of personal relations between ministers and civil servants made coordination highly effective in economic policies (Levi-Faur 1998: 70). The system of cabinet minister appointment of top-level civil servants ensured a strong adherence to government policy down through the bureaucratic ranks (Nachmias and Arbel-Ganz 2005: 286).

JAMAICA (1962-)

The administration was ineffective throughout the period. The relatively small island was early on controlled by the British and from the mid-19^th century, independent Jamaican authorities (Harriott 2000: 30). Basic administrative structures existed throughout the island although their operation was limited in effectiveness some places (BTI 2006: 4; BTI 2012: 4).

Meritocracy was never in place. The British installed a Public Service Commission under colonialism which gave rights of tenure to civil servants and autonomy in hirings and firings. This remained in place formally after colonialism (Perez 2001: 31; Bissessar 2001: 532). Notably, the judiciary was highly autonomous and independent in appointments as it was also based on the British legacies, that of the common law. Judges held their position until the age of 70 and could only be purged on grounds of misconduct. The legacy of the British certainly separates Jamaica from the rest of the Caribbean. Indeed, the Public Service Commission remained powerful under democratic rule managing the relations between politicians and administrators (Bissessar 2001: 532-534; Tindigarukay 2004: 82-84).

Formal powers aside, the period from 1944 to 1962 was detrimental to the development of meritocracy. The independent Jamaican state that emerged was party clientelic, led by a highly politicizing and personalist figure of Bustamante and a two-party system (Dawson 2016: 201-202). Of particular importance, the state expansion begun by the British in the early 1940s was never managed centrally but instead led to a pattern of politicized appointments locally and at lower levels of government (Edie 1991: 19-32). From 1962, the lower classes of Blacks were brought into the state apparatus at lower levels via patronage appointments through the Jamaica Labour Party and the People’s National Party (Edie 1991: 34). In this way, the Public Service Commission only controlled the highest levels of the bureaucracy. The rest was managed by the parties or the executive (St. Hill 1970: 143) – a trait that continued across decades and the wave of NPM reforms in the 1980s, 1990s, and 2000s (Perez 2001: 31).

Unresponsiveness also prevailed. The famous guards against corruption that followed meritocracy in Great Britain did not materialize in Jamaica. Numerous attempts to fight corruption since the early 1980s have been unsuccessful and corruption remained perceivably endemic. 90 % of the population believed corruption to be rampant in 2005 – a figure that can hardly be due to false popular perceptions (Osei 2007: 167). Poor monitoring systems, low morale, low salaries, and a lack of training were at the core of the high levels of administrative corruption (Perez 2001: 31; Bissessar 2001: 534; Osei 2007:169-170; BTI 2012: 6). Furthermore, many accounts point out that patronage and clientelism caused severe problems in implementation and a tendency of lower-level servants to serve private interests at the expense of the public interest or the government policies (see e.g. Edie 1991: 143).
JAPAN (1952-)

The administration was effective throughout the period. Between 1868 and 1900, a bureaucratization of the Japanese state and society occurred as Japan entered the era of modern state-building (Silberman 1993: 159). And as no forces occupied Japanese territory after WWII, the state firmly penetrated it in 1952.

Although the system established under the Meiji Restoration from the late 19^th century was not quite Weberian, meritocracy was a firm trait (Pempel 1992: 20; Silberman 1993: 159). This continued after WWII and under democracy as judicial independence was assured via the recruitment according to abilities rather than political affiliation and with a widely used and systematic recruitment from the public sector schools (Kim 1988: 21, 32-33). The Japanese miracle was largely forged by this recruitment pattern (Johnson 1982: 35-37; see also Haggard 2004: 63-64). This continued in subsequent decades (Furukawa 1999: 444).

The highly developmental nature of the Japanese state, its embedded autonomy, was forged in a strong coordination between the meritocratic bureaucracy and the politicians. Equally so, the role of the ministries in coordinating business interests for the benefit of the economy and the political guidelines was exemplified in the MITI (see Johnson 1982; Haggard 2004). Bureacrats were powerful and participated in political committees on a par with politicians. However, there were no strong conflicts between politicians and bureaucrats and no signs of unresponsiveness (Johnson 1982: 46; Kim 1988: 1; Furukawa 1999: 440). Formally, bureaucracy was better controlled over the years. Fundamentally inspired by the Prussian model but altered by American standards of administration after WWII, the politico-administrative relations were close and effective (Pempel 1992: 19-20).

KENYA (2002-)

The administration was ineffective throughout the period. According to BTI (2003: 4), administrative structures existed throughout the territory despite some deficiencies. Other sources are scarce. I thus assume territorial penetration to be in place.

Under President Moi until 2002, the civil service was highly politicized (Grindle 1996: 97) and the judiciary was similarly described as captured by political and ethinc forces (see Mutua 2001: 98-99). With the new political order of the 2000s, meritocracy was still sacrificed to favor and balance ethnic interests by employment (Kjaer 2004: 406). Under democracy, politicization was widespread and general across ministries (Amadi 2009: 8). The Public Service Commision had been established since independence but never functioned independently and implemented highly ambiguous rules demanding that all civil servants would work to the pleasure of the president. More importantly, the constitution gave access for the president to appoint and dismiss civil servants. In practice, the civil service and judiciary remained mired in patronage appointments (Amadi 2009: 14-15, 29).

An anti-corruption campaign was initiated in 1998 but did not work. Excessive corruption in the administration continued (Hornsby 2012: 633-635). From 2004, the administration and judiciary were overhauled by several purges but this did not fundamentally change administrative efficiency (Hornsby 2012: 704). The elections in 1992 had further marked the deterioration of institutional capacities in the central departments and agencies as well as local administrative capacities. Poor delivery of public services was a firm negative legacy (Grindle 1996: 96-97, 145; Branch and Cheeseman 2008: 11). Still in the late 2000s, loyalty to ethnicities was highly valued and caused state capture, favoritism, and nepotism in government service delivery (Amadi 2009: 7; Kivoi 2010: 45).

LATVIA (1993-)

The administration was effective from 1994. The status of territorial penetration was the same as in Estonia as the time of independence in the early 1990s: Soviet legacies of societal penetration dominated (see BTI 2008: 6). Just as Estonia, Latvia was also a borderline case in terms of meritocracy being the victim of Soviet Nomenklatura but also inheritor of German/Scandinavian norms of administrative staffing (see Vanagunas 1999: 222-230; Møller and Skaaning 2010: 325, 338). A Civil Service Law much like that of Estonia’s in January 1995 was passed in April 1994 serving to legalize meritocracy (Vanagunas 1999: 213). However, judging by the few systematic investigations of the actual nature of hiring and firing deviations from the law of meritocracy were fewer than in Estonia. Despite some issues of non-transparency in hiring processes, meritocracy was deemed firmly in place throughout the period from mid-1990s to 2010 (Palidauskaite, Pevkur, and Reinholde 2010: 47-49). Recruitment by open competition was a strong norm in all ministries (Meyer-Sahling 2009: 26). State secretaries held mediating posts between the political and administrative levels and served to ensure appointment of civil servants on grounds of political neutrality. Problems of politicization pertained to the lack of skilled personnel (Nørgaard and Hansen with Ostrovska and Møller 2000: 25-27).