¹ A number of scholars have pointed to the surprising resilience of the state after communism, including Grzymala-Busse, Anna, “Political Competition and the Politicization of the State in East Central Europe,” Comparative Political Studies 36 (December 2003Google Scholar); Gimpelson, Vladimir and Triesman, Daniel, “Fiscal Games and Public Employment,” World Politics 54 (January 2002Google Scholar); Bartkowski, Jerzy, “Pahstwo i wielka przemiana,” in Jasinska-Kania, A. and Raciborski, J., eds., Naro'd, Wladza, Spoleczeristwo (Nation, power, society) (Warsaw:Scholar, 1996Google Scholar); Kiezun, Witold, “Czterej jezdzcy apokalipsy polskiej biurokracji,” Kultura 3 (March 2000Google Scholar); and the Bank, World, Corruption in Poland: Review of Priority Areas and Proposalsfor Action, Report by the Warsaw Office of the World Bank, October 11,1999Google Scholar.

² This term departs from the conventional usage of “state building.” Traditionally, state building has referred to the process by which the state gains greater power over or autonomy from society-by war making and/or by bureaucratizing; see Tilly, Charles, “War-Making and State-Making as Organized Crime,” in Evans, Peter, Rueschemeyer, Dietrich, and Skocpol, Theda, eds., Bringing the State Back In (Cambridge:Cambridge University Press, 1985Google Scholar); and Silberman, Bernard, Cages of Reason: The Rise of the Rational State in France, fapan, the United States, and Great Britain (Chicago:University of Chicago Press, 1993), 1–4Google Scholar. In either case, expansion implies increasing state capacity. In contrast, runaway state building is driven not by interstate competition or bureaucrats seeking legal-rational legitimation but by elected politicians seeking patronage resources for the task of party building. Thus, in runaway state building, a bigger state is a sign of both patronage and state underdevelopment; see Kitschelt, Herbert, “Linkages between Citizens and Politicians in Democratic Polities,” Comparative Political Studies (August-September 2000CrossRefGoogle Scholar); Piattoni, Simona, “Clientelism, Interests, and Democratic Representation,” in Piattoni, , ed., Clientelism, Interests, and Democratic Representation: The European Experience in ical Perspective (Cambridge:Cambridge University Press, 2001), 200–206CrossRefGoogle Scholar; and Scott, James, “Corrup tion, Machine Politics, and Political Change,” American Political Science Review 63 (December 1969CrossRefGoogle Scholar)

³ Following Martin Shefter, I define patronage as “a divisible benefit that politicians distribute to individual voters, campaign workers, or contributors in exchange for political support”; see Shefter, , Political Parties and the State: The American Historical Experience (Princeton:Princeton University 1994), 283Google Scholar fn. 3. These benefits may take many forms. I focus here on positions within the state administration.

⁴ See Mainwaring, Scott, Rethinking Party Systems in the Third Wave of Democratization: The Case Brazil (Stanford, Calif.:Stanford University Press, 1999), 3–4Google Scholar; Mair, Peter, Party System Change: Ap proaches and Interpretations (Oxford:Clarendon Press, 1997Google Scholar); Toole, James, “Government Formation and Party System Stabilization in East Central Europe,” Party Politics 6, no. 4 (2000), 458CrossRefGoogle Scholar; Shabad, Goldie and Slomczynski, Kazimierz, “Interparty Mobility among Political Elites in Post-Communist East Central Europe,” Party Politics 10, no. 2 (2004CrossRefGoogle Scholar); and Kreuzer, Marcus and Pettai, Velio, “Patterns of Political Instability: Affiliation Patterns of Politicians and Voters in Postcommunist Estonia, Latvia and Lithuania,” Studies in Comparative International Development 38 (Summer 2003CrossRefGoogle Scholar).

⁵ Shefter (fn. 3), 14–60.

⁶ I will address a more nuanced political-cultural hypothesis below.

⁷ Specifically, this includes the central ministries and offices, their branch offices, the territorial administration, state inspectorates, and tax offices.

⁸ In each of these countries, it is considered an abuse of power to place political appointees in any office below that of deputy minister or advisor to the minister. Department heads, deputy department heads, and other staff below these levels are supposed to be professional appointments.

⁹ Other areas of the state have proved less open to patronage than the administration. Welfare agencies, for example, were less directly linked to the nomenklatura system. Moreover, even lower-level positions in the welfare system-such as nurses and teachers-require more specialized knowledge and professional expertise than those in the administration, which serves as a barrier to patronage. Author's interviews with Jaromfr Vepfek, head ofTym DG Plus, a Czech health care policy consultancy, Prague, July 27,2001; Zuzana Srankova, Orava Project for Democracy in Education, Bratislava, July 11,2001; and Jozefina Hrynkiewicz, Instytut Spraw Publicznych, Warsaw, June 21, 2001.

¹⁰ In her analysis of patronage in these countries, Grzymala-Busse (fn. 1) uses a wider operational definition of the state: the public administration. However, the public administration includes personnel whom the national government does not appoint, most notably the local-level administration; therefore, local administration growth cannot be taken as evidence of national-party patronage. The second problem with the public administration is that it includes different categories of personnel across countries.

¹¹ Unfortunately, Czech and Slovak data are unavailable before 1993.

¹² The post-1998 numbers for Poland are adjusted to reflect personnel reassigned to regional governments after that country's 1998 decentralization. Even comparing all three countries from 1993 to 1998 to avoid this complicating factor shows the same ranking: 82 percent Slovakia, 48 percent Poland, 16 percent Czech Republic.

¹³ Noting the latter argument are Bartkowski (fn. 1); and Gimpelson andTriesman (fn. 1).

¹⁴ These additional data on administrative expansion capture general trends but may not to be directly comparable with the Polish, Czech, and Slovak data, since they were collected by other scholars and may use different classification schemes. The data for Bulgaria, covering the 1990–95 period, come from Verheijen, Tony, “The Civil Service System of Bulgaria: Hope on the Horizon,” in Verheijen, , ed., Civil Service Systems in Central and Eastern Europe (Cheltenham, U.K.:Edward Elgar, 1999), 126Google Scholar. Russian data, covering the 1993–99 period, come from Gimpelson and Triesman (fn. 1), 158. The rate of economic expansion is calculated using real per capita GDP in 1995 U.S.S for the first and last years for each country; see World Bank, World Development Indicators (2002 CD ROM).

¹⁵ Poland's presidency is stronger, but it is not a presidential system. As tempting as it is to attribute the Slovak state's expansion to its winning independence in 1993, this explanation has two important shortcomings. First, Czechoslovakia was a federal state, and the Slovak Republic had had its own state apparatus-republican governments, ministries, and branch offices-since 1968; see Hendrych, Dusan, “Transforming Czechoslovakian Public Administration: Traditions and New Challenges,” in Hesse, Joachim, ed., Administrative Transformation in Central and Eastern Europe (Cambridge, Mass.:Blackwell, 1993), 45–47Google Scholar. Second, the timing of Slovak expansion suggests that party political factors were paramount. As Figure 1 shows, the lion's share of expansion took place not in the first years after independence but later, in a concentrated spike from 1996 to 1997. This spike coincided with a blatantly self-serving “reform of the public administration” undertaken by the governing political machine, which I describe below.

¹⁶ The first serious attempt at regulating party financing was in Poland but was not until 1997, well after the greatest expansion of the administration; see Walecki, Marcin, ed., Finansowaniepolityki: Wybory, pieniadze, partie polityczne (Financing politics:Elections, money, political parties) (Warsaw: Wydawnictwo Sejmowe, 2000Google Scholar).

¹⁷ Because Poland lacked a civil service law until 1997, there were no standard guidelines for hiring and firing state personnel. After 1997 it passed two civil service acts within approximately a year. The first, legislated by the postcommunists, was perceived as a political gambit by the following post-Solidarity government and replaced with a new version. The irony, of course, is that this is precisely the kind of maneuvering that civil service legislation is supposed to prevent. Slovakia and the Czech Republic lacked this legislation until after 2001.

¹⁸ Nunberg, Barbara, Readyfor Europe: Public Administration Reform and European Union Accession Central and Eastern Europe, World Bank Technical Paper no. 466 (Washington, D.C.:World Bank, 2000Google Scholar).

¹⁹ Kitschelt, “Accounting for Outcomes of Post-Communist Regime Change: Causal Depth or Shallowness in Rival Explanations” (Paper presented at the annual meeting of the American Political Science Association, Atlanta, September 1–5,1999).

²⁰ See Rose, Richard, Mobilizing Demobilized Voters in Post Communist Societies, Studies in Public Policy, no. 246 (Glasgow:University of Strathclyde, 1995Google Scholar); and Howard, Marc, The Weakness of Civil Society in Post-Communist Europe (Cambridge:Cambridge University Press, 2003CrossRefGoogle Scholar). Ekiert and Kubik offer a caveat in their survey of protest events, suggesting that postcommunist society is more mobilized than is generally recognized. As they also note, however, “the magnitude of protest is by and large lower than in more established democracies”; see Ekiert, and Kubik, , Contentious Politics in New Democracies: Hun gary, Poland, Slovakia, and the Former East Germany since 1989, Central and Eastern Europe Working Paper Series, no. 41 (Cambridge:Harvard University, Center for European Studies, 1997), 31Google Scholar. Thus, their “contentiousness” measure does not contradict the widely noted low identification with political parties.

²¹ Quoted in Rose, Richard, Mishler, William, and Haerpfer, Christian, Democracy and Its Alternatives: Understanding Post-Communist Societies (Baltimore:Johns Hopkins University Press, 1998), 155Google Scholar. See Linz, Juan and Stepan, Alfred, Problems ofDemocratic Transition and Consolidation: Southern Europe, America, and Post-Communist Europe (Baltimore:Johns Hopkins University Press, 1996), 245Google Scholar–47.

²² Rose, Mishler, and Haerpfer (fn. 21), 85–88,153–57.

²³ Rose (fn. 20), 20–24.

²⁴ Rose, Mishler, and Haerpfer (fn. 21), 157.

²⁵ Party membership in Poland has been estimated at no higher than 1.5 percent of the electorate; the comparable figures for the Czech Republic and Slovakia are 6.4 and 3.1 percent, respectively. Party membership figures in Western Europe are around 9 percent; see Aleks Szczerbiak, “Party Structure and Organizational Development in Post-Communist Poland,” Journal ofCommunist Studies and Transition Politics 17 (June 2001), 111–12. The West European figure is calculated as a simple average of the party membership figures for the fourteen countries as reported in Szczerbiak (p. 112).

²⁶ The conventional wisdom is that public sector salaries in this region are too low to be attractive. While many are-notably nurses' and teachers' salaries-those of state administrative officials are quite attractive in relative terms. In Poland, for example, the average central-level official's salary was 44 percent higher than the general average in 1998, and in Slovakia it was twice as high; see Statistical Yearbook of the Republic of Poland (Warsaw:GUS, 1999Google Scholar); and Plenipotentiary for the Reform of the Public Administration, Strategy of Public Administration Reform in the Slovak Republic, http://www.mesalO.sk/vs/ (accessed June 19, 2001). This is not to mention the opportunities for rent seeking as a state official.

²⁷ See Geddes, Barbara, Politician's Dilemma: Building State Capacity in Latin America (Berkeley:University of California Press, 1994), 40–41Google Scholar. The availability of public funding and interest groups may also have an impact on parties' demand for patronage. Though small in size, public funding has been available in all three countries since the early 1990s and so cannot account for differences in patronage. One might also hypothesize that organized interest groups were stronger in the Czech Republic, allowing its parties to do without patronage or membership. The scholarship emphasizes these groups' weakness, however; see Mitchell Orenstein and Raj Desai, “State Power and Interest Group Formation,” Problems of Post-Communism 44 (November-December 1997).

²⁸ Again, I am not speaking here of the welfare state, which was generous under communism, providing free health care, education, and housing-though the necessity for informal payments was notorious. Many still expect the state to provide these services. My characterization of the administration as delegitimized is not, therefore, intended to describe the welfare state, and I exclude it from the personnel data. See Rose (fn. 20), 19; Kapstein, E. and Mandelbaum, M., eds., Sustaining the Transition: The Social Safety Net in Postcommunist Europe (New York:Council on Foreign Relations, 1997Google Scholar); and Cook, Linda, Orenstein, Mitchell, and Rueschemeyer, Marilyn, eds., Left Parties and Social Policy in Postcommunist Europe (Boulder Colo.:Westview Press, 1999Google Scholar.

²⁹ Letowski, Janusz, “Polish Public Administration between Crisis and Renewal,”Google Scholar and Olga Vidlakova, “Administrative Reform in the Czech Republic,” in Hesse (fn. 15), 5, 9, 70–71.

³⁰ Rose, Mishler, and Haepfer (fn. 21), 124–25.

³¹ Ibid., 154. The NDB III survey includes fifteen public institutions. See also World Bank (fn. 1), 7.

³² Linz and Stepan (fn. 21), 48–49.

³³ Guillermo O'Donnell calls this mechanism “vertical accountability.” Party-system institutionalization also enhances “horizontal accountability,” parties holding each other accountable, especially between elections; see O'Donnell “Horizontal Accountability in New Democracies,” in Schedler, A., Diamond, L., and Plattner, M., eds., The Self-Restraining State (Boulder, Colo.:Lynne Rienner, 1999), 29–30Google Scholar, 42–44.

³⁴ Given voter disengagement from politics in Eastern Europe, it may seem that these are uninstitu-tionalized systems. However, institutionalization is a continuous, not dichotomous, variable; see Main-waring (fn. 4), 22–26. Moreover, institutionalization varies considerably in this region, both across countries and over time; see Shabad and Slomczynski (fn. 4); Kreuzer and Pettai (fn. 4); and Took (fn. 4).

³⁵ Shabad and Slomczynski (fn. 4) have measured institutionalization in terms of interparty switching. Consonant with the analysis here, they find that the rate of party switching has decreased significantly in the Czech Republic though not in Poland. Kreuzer and Pettai (fn. 4) measure institutionalization in terms of the electoral success of nonestablished parties: start-ups, splinters, and mergers. On one dimension of institutionalization, the emergence of career politicians, Shabad and Slomczynski offer a caveat on the Polish-Czech comparison, finding that reelection rates for MPs have been increasing in both countries; see Shabad, and Slomczynski, , “The Emergence of Career Politicians in Post-Communist Democracies: Poland and the Czech Republic,” Legislative Studies Quarterly 17 (August 2002Google Scholar).

³⁶ Mainwaring (fn. 4), 89,128–31.

³⁷ Ibid.; O'Donnell (fn. 33), 30.

³⁸ Toole (fn. 4), 458; Mainwaring (fn. 4), 123.

³⁹ Mair (fn. 4), 199–223; Took (fn. 4).

⁴⁰ I have in mind what Giovanni Sartori describes as “predominance": one party is significantly stronger than the rest and is supported by a winning majority. Elections matter in such systems since the strongest party can be turned out of government if it cannot find a supporting coalition; see Sartori, , Parties and Party Systems: A Frameworkfor Analysis (Cambridge:Cambridge University Press, 1976), 1Google Scholar: 131–201.

⁴¹ This is a key difference between my conception of robust competition and Grzymala-Busse's, in whose formulation fractionalization is an unqualified good because it disperses power; Grzymala-Busse (fn. 1), 1131. While overconcentration of power (such as in Meciar's Slovakia) is harmful, too much dispersion creates its own incentives for patronage, as described above. As long as no party is dominant, some degree of concentration is beneficial because it increases vertical accountability and creates parties whose survival does not depend on winning the next elections.

⁴² Laakso, Markku and Taagepera, Rein, “‘Effective’ Number of Parties: A Measure with Application to West Europe,” Comparative Political Studies 12 (1979), 3–27CrossRefGoogle Scholar, 4.

⁴³ Sartori(fn. 40), 193.

⁴⁴ Laakso and Taagepera (fn. 42).

⁴⁵ The volatility index measures the net change in the vote shares of all parties across elections; see Mainwaring (fn. 4), 28. Because frequent splits and mergers represent lack of institutionalization, I count splits and mergers as fully new parties. This maximizes volatility, but does so consistently while avoiding difficult judgment calls about party continuity.

⁴⁶ Toole(fn.4),458.

⁴⁷ Mair (fn. 4).

⁴⁸ I define alternation as any recomposition of the government coalition or major restructuring of the cabinet, such as the replacement of the prime minister.

⁴⁹ Szczerbiak, Aleks, “Interests and Values: Polish Parties and Their Electorates,” Europe-Asia Studies 58 (1999), 1432Google Scholar.

⁵⁰ Ibid., 1431.

⁵¹ I do not count the aborted government of 1992; had it gained approval, however, it would have been a partial alternation.

⁵² The coalition restructurings of the early 1990s defied full enumeration in Figure 3. After 1993 the governing formula appeared to simplify: the field of coalitions reduced to the SLD together with the Peasant Party (PSL) and AWS together with Freedom Union (uw). Appearances were misleading, however, because the coalition of AWS-UW was more in the nature of an umbrella group than an alliance of two parties. As in Slovakia, new parties were easily generated because access to power was open.

⁵³ Quoted in Bogusz, A., Macieja, D., and Wojtkowska, Z., “Jak urzadza sie SLD,” Wprost (June 25, 2000Google Scholar).

⁵⁴ Quoted in Letowski (fn. 29), 2. On patronage in the Polish administration, the World Bank (fn. 1) reported: “Other forms of high level corruption are manifested in nepotism in public sector appointments. ... This tendency is exacerbated by the practice of making political appointments down to medium levels in the administration” (p. 9). See also Michat Matys, “Nie mozesz bye mczyj,” Gazeta Wyborcza (June 2–3,2001); Kiezun (fn. 1), 8–11; and Ewa Jakubkowska, Corruption in Procurement in Poland: Analysis and Recommendations, www.batory.org.pl/ftp/program/przeciw-korupcji/publikacje/ corruption_procurement.rtf (accessed August 30, 2004).

⁵⁵ Michal Kulesza, “Szanowny Panie Premierze,” Polityka (May 28,1994). See also Louisa Vinton, “Power Shifts in Poland's Ruling Coalition,” RFE/RL Research Report (March 18,1994), 7–10.

⁵⁶ See Wojciech Taras, “Changes in Polish Public Administration, 1989–1992,” in Hesse (fn. 16), 14,20–21.

⁵⁷ Biitora, Martin et al., The 1998 Parliamentary Elections and Democratic Rebirth in Slovakia (Bratislava:Institute for Public Affairs, 1999Google Scholar).

⁵⁸ Kusy, Miroslav, “Slovakia ‘97,” Perspectives 9 (1998), 45Google Scholar.

⁵⁹ HZDS's popularity was based on nationalism and Meciar's charisma. After Slovakia's independence in 1993, HZDS's appeal began to decline, and it used its control over the machinery of government to maintain its position through patronage. Meanwhile, the Slovak opposition parties underwent continuing fragmentation, their popular appeal tarnished by the memory of their ambivalence toward Slovak independence. In 1994 the opposition parties succeeded in bringing down the HZDS government with a vote of no-confidence. Their success was short-lived, however, as HZDS swept back into power in elections five months later. HZDS then ruled in coalition with two very junior and compliant parties, who shared in the spoils of patronage. See Haughton, Tim, “HZDS: The Ideology, Organization, and Support Base of Slovakia's Most Successful Party,” Europe-Asia Studies 53 (July 2001CrossRefGoogle Scholar).

⁶⁰ Ironically, HZDS was undone by its own dominance. Hoping to disqualify a number of opposition parties, it changed the electoral law just months before the 1998 election to require all parties, even those in electoral alliances, to win 5 percent of the vote for representation; see Biitora et al. (fn. 57). Unintentionally, it forced the opposition to cooperate at last and form a single party, the Party of the Democratic Coalition (SDK). Although technically one party, SDK was a confederation of five parties that broke apart after the election. In computing fractionalization and turnover, I treat it asfiveparties.

⁶¹ In the early 1990s the opposition coalition consisted of Public against Violence, the Christian Democratic Movement, and the Democratic Party. In 1994 the composition shifted, now including the postcommunists, the Democratic Union, and the Christian Democratic Movement. The year 1998 saw another change with the inclusion of the SDK, the Movement for Civic Understanding, and the Hungarian party.

⁶² Matys(fn. 54).

⁶³ Fisher, Sharon, “Slovak Government's Personnel Changes Cause Controversy,” RFE/RL Research Report (May 27, 1994), 10Google Scholar.

⁶⁴ A leaked internal party document entitled, “The Main Tasks of HZDS,” stated that HZDS “should work to strengthen its position within Slovak society by continuing to reshuffle personnel within the state administration and diplomatic corps”; see Knvy, Vladimir, “Slovakia and Its Regions,” in Biitora, Martin and Skladony, Thomas, eds., Slovakia 1996–1997: A Global Report on the State of Society (Bratislava:Institute for Public Affairs, 1998), 59Google Scholar.

⁶⁵ Fisher (fn. 63), 11.

⁶⁶ Ibid., 11.

⁶⁷ Kunder, Peter, “Vyhrali politici: porazili reformu,” Sme (July 6, 2001Google Scholar); Vagovic, Marek, “Reformne K.O.: Skutocmi reformu verejenej spravy parlament odmietol,” Domino-Forum (July 12–18, 2001Google Scholar).

⁶⁸ In another contrast, the Czech parties located themselves on a clearly distinguishable left-right, socioeconomic issue spectrum; see Kitschelt, Herbert et al., Post-Communist Party Systems: Competition, Representation, and Inter-Party Cooperation (Cambridge:Cambridge University Press, 1999CrossRefGoogle Scholar).

⁶⁹ In opinion polls between the 1992 and 1996 elections, CSSD support ranged between 12 and 21 percent; see Fitzmaurice, John, Politics and Government in the Visegmd Countries (New York:St. Martin's Press, 1998), 130CrossRefGoogle Scholar.

⁷⁰ There was one partial exception here: a technocratic caretaker government oversaw the transition after ODS's fall in November 1997 until the next elections in June 1998.

⁷¹ Until 1998 governments were led by ODS in coalition with the Christian Democrats (KDU-CSL) and Civic Democratic Alliance (ODA) as junior members. During that time, the opposition consisted primarily of the Social Democrats and the outsider Communist and Republican Parties. Access to government was closed, as the latter two were excluded.

⁷² One does not find in accounts of Czech administrative development the same emphasis on politicization as in Poland and Slovakia. See, for example, the country studies by Hendrych, Pomahac, and Vidlakova, in Hesse (fn. 15); and Illner, Michal, The Territorial Dimension of Public Administration Reforms in East Central Europe, Institute of Sociology Working Papers (Prague:Czech Academy of Sciences, 1997Google Scholar). Even in more critical accounts, such as Abby Innes's, the claim that ODS sought patronage in the administration is tempered with the qualifier that it did so regarding the “top-flight” and “senior” positions; see Innes, , Czechoslovakia: The Long Goodbye (New Haven:Yale University Press, 2001), 232Google Scholar. By contrast, accounts of Slovakia and Poland emphasize that patronage occurred extensively at the middle and lower levels.

⁷³ Krause, Kevin, “Accountability and Party Competition in Slovakia and the Czech Republic” (Ph.D. diss., University of Notre Dame, 2000), 72Google Scholar.

⁷⁴ Each country studied here has had party financing scandals, with Poland's ongoing Rywin Affair the latest example. Moreover, as Krause writes of the ODS scandal, “The scale of the acknowledged donations is relatively small, even by the standards of Czech politics, and only a few such donations were uncovered”; Krause (fn. 73), 78. The most damaging charges-that ODS manipulated privatization in exchange for campaign contributions-were never substantiated, even after an extensive audit by the American firm Deloitte and Touche; see Stroehlein, Andrew, “The Czech Republic, 1992 to 1999,” Central Europe Review (September 13, 1999Google Scholar), http://www.ce-review.org/_archives99.html (accessed November 14,2004).

⁷⁵ See Roberts, Andrew, “Demythologising the Czech Opposition Agreement,” Europe-Asia Studies 55 (December 2003), 1275CrossRefGoogle Scholar–76.

⁷⁶ In Slovakia, Meciar's machine saw local governments as a threat and curtailed their financial resources from 1992 to 1998; see Bryson, Phillip and Cornia, Gary, “Fiscal Decentralization in Economic Transformation: The Czech and Slovak Cases,” Europe-Asia Studies 52 (May 2000CrossRefGoogle Scholar). Consequently, local governments had no capacity for administrative expansion, and personnel actually decreased in number by 17.7 percent between 1993 and 1998.

⁷⁷ Kroupa, Ales and Kostelecky, Tomas, “Party Organization at the National and Local Level in the Czech Republic since 1989,” in Lewis, Paul, ed., Party Structure and Organization in East-Central Europe (Cheltenham, U.K.:Edward Elgar 1996Google Scholar).

⁷⁸ Kitscheltetal. (fn. 68).

⁷⁹ Toole (fn. 4).

⁸⁰ Nunberg (fn. 18), 280.

⁸¹ Verheijen(fn. 14), 126