Stability through constraints: the impact of fiscal rules on autocratic survival

ABSTRACT

A growing literature has investigated the role that formal and informal economic institutions play for autocratic survival. However, this literature has, so far, ignored a type of formal economic institution that has grown in importance among both democracies and non-democracies in recent decades, namely, national fiscal rules. In this article, we argue that fiscal rules can affect autocratic survival but that the effect is time-dependent. The introduction of a stricter fiscal rules framework causes short-term fiscal retrenchment and might increase the uncertainty among regime supporters about the provision of future patronage and spoils. However, stricter fiscal rules also improve long-term fiscal management and, thus, increase investor confidence and economic performance in the long run. Consequently, fiscal rules stabilize autocratic regimes in the long run but not in the short run. Fixed-effect estimations on a panel of autocracies from 1987 to 2016 provide substantial evidence in favour of this argument.

KEYWORDS:

• Autocratic breakdown
• autocratic survival
• economic institutions
• fiscal policy
• fiscal rules

Acknowledgements

We want to thank the editors and the two anonymous reviewers for good and constructive comments that have helped us improve the article. We also want to thank Jakob Tolstrup and Jørgen Møller for several rounds of helpful feedback and suggestions.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Gandhi, Political Institutions under Dictatorship; 2008; Geddes et al., How Dictatorships Work; Gerschewski, “Three Pillars of Stability”; Svolik, The Politics of Authoritarian Rule; Wintrobe, “The Tinpot and the Totalitarian”.

2 Boix and Svolik, “Foundations of Limited Authoritarian Government”; Gandhi and Przeworski, “Authoritarian Institutions”; Magaloni, “Credible Power-sharing”.

3 Wright, “Do Authoritarian Institutions Constrain? ”; Kennedy, “The Contradiction of Modernization”.

4 Aaskoven, “Oil, Elections and Fiscal Transparency”; Bodea et al., “Monetary Institutions and Autocratic Breakdown”; Hyde and O’Mahoney, “Which Elections Can Be Lost?”.

5 Knutson et al. “Autocratic Elections”; see also Bernhard et al., “Institutionalising Electoral Uncertainty”.

6 Schaechter et al., “Fiscal Rules in Response”.

7 Most famously, Germany, which has had a constitutional balanced budget rule in place since 1969; see Lledó et al., Fiscal Rules at a Glance.

8 Schaechter et al., “Fiscal Rules in Response”.

9 The regime classification of Nigeria is somewhat contested, as some datasets code a democratic transition in the country in 1999. However, the updated Boix et al. dataset codes the country as an autocracy throughout our period of analysis. For this reason, we treat Nigeria as an autocracy in the theoretical discussion as well as in the empirical models using the Boix et al. dataset (Boix et al. “A Complete Data Set”.

10 Baunsgaard, “Fiscal Policy in Nigeria”.

11 [11] This was also backed by the IMF; see (Ushie, “Implementing the Fiscal Responsibility Act”, 18).

12 Lledó et al., Fiscal Rules at a Glance, 57; Ushie, Implementing the Fiscal Responsibility Act, 4.

13 Further evidence against this concern is provided in Appendix A, where we show that autocracies under IMF programs are actually less (rather than more) likely to have fiscal rules in place.

14 Formal models of government spending decisions often assume a central role for the finance minister as a champion of fiscal restraint. See Alexiadou, “Ideologies, Partisans and Loyalists”, 1056–1062; Hallerberg et al., Fiscal Governance in Europe, 24–31. Analyses of the adoption of fiscal rules in democratic states suggest that the finance ministry often plays a pivotal role in initiating fiscal rules reforms (See Suenson, et al., “Why Lash Yourself”; and Wehner, “Budget Reform”). There is little reason to believe that this institutional drive should not exist in non-democratic states.

15 Hanson, “Managing the Economy”, 180–182; Lledó et al., Fiscal Rules at a Glance, 65.

16 Hanson, “Managing the Economy”, 175–175.

17 Lledó et al., Fiscal Rules at a Glance, 65.

18 Additionally, as raised by one of the reviewers, some autocratic regimes might not want to risk the potential short-term upheaval of these fiscal rules. See also the ‘Theory’ section for an elaboration of the potential short-term costs of fiscal rules in autocracies.

19 See also Appendix A for an overview of the distribution of fiscal rules across different types of autocracies.

20 Asatryan et al., “Balanced Budget Rules”; Badinger and Reuter, “The Case for Fiscal Rules”.

21 Asatryan et al. “Balanced Budget Rules”, 115.

22 Gandhi and Lust-Okar, “Elections under Authoritarianism”.

23 Bodea et al., “Monetary Institutions and Autocratic Breakdown”.

24 Badinger and Reuter, “The Case for Fiscal Rules”.

25 Asatryan et al., “Balanced Budget Rules”.

26 Bergman and Hutchison, “Economic Stabilization”; Badinger and Reuter, “The Case for Fiscal Rules”.

27 Although this also seems to be the case in democracies; see Reuter, “National Numerical Fiscal Rules”.

28 Kelemen and Teo, “Law, Focal Points and Fiscal Discipline”.

29 This transparency- and coordination-enhancing mechanism through which fiscal rules discipline dictators’ fiscal policy is similar to the mechanism by which nominally democratic institutions induce dictators to respect power-sharing agreements. See Boix and Svolik, “The Foundations”; Svolik, The Politics of Authoritarian Rule.

30 Aaskoven, “Institutionalizing the Autocratic Penalty Away”.

31 Badinger and Reuter, “The Case for Fiscal Rules”.

32 Aaskoven, “Institutionalizing the Autocratic Penalty Away”.

33 Ibid.

34 Thorton and Vasilakis, “Fiscal Rules and Government Borrowing Costs”.

35 Arias et al., “Cooperative Autocracies”; DiGuiseppe and Shea, “Borrowed Time”; Aaskoven, “Institutionalizing the Autocratic Penalty Away,”

36 DeGuiseppe and Shea, “Sovereign Credit”; DeGuiseppe and Shea, “Borrowed Time”; Lucardi, “Strength in Expectation”; Miller, “Economic Development”, 1010.

37 Clinton et al., “Deficit Reduction”; Papageorgiou and Vourvachaki, “Macroeconomic Effects of Structural Reforms”.

38 On riots see Ponticelli and Voth, “Austerity and Anarchy”. On elite defections see Knutsen and Rasmussen, “The Autocratic Welfare State”; and Reuter and Szakonyi, “Elite Defection under Autocracy”. The effects of fiscal austerity on leader turnover remain contested; see Arias and Stasavage, “How Large Are the Political Costs”.

39 Bodea et al., “Monetary Institutions and Autocratic Breakdown”. At least in democracies, fiscal rules seem to drive down the occurrence of election-year public spending increases (Rose, “Do Fiscal Rules”), a strategy that is also frequently used by autocratic governments (see Blaydes, Elections and Distributive Politics in Mubarak’s Egypt). See also Aaskoven’s “Oil, Elections and Fiscal Transparency” which shows how fiscal institutions can decrease the occurrence of oil-induced election-year spending increases, especially in electoral autocracies.

40 Boix et al., “A Complete Data Set”.

41 Djuve et al. “Patterns of Regime Breakdown”.

42 Geddes et al., How Dictatorships Work.

43 Djuve et al., “Patterns of Regime Breakdown”.

44 In the models using the GWF measure, the dependent variable is lead rather than the independent variables being lagged.

45 Aaskoven, “Signalling to Creditors and Voters”.

46 Some of these rules and institutions are common for all types of fiscal rules, and the final index thus gives relatively high weight to these factors.

47 Since the later empirical analysis relies on country-fixed effects, the variation exposed in this graph is very informative with regards to the cases driving the results.

48 Lledó et al., Fiscal Rules at a Glance.

49 The dataset does not record rule breach. However, if the fiscal rules are de facto suspended, the country can be coded as having no fiscal rules in place, which was the case for Argentina after 2008; see Lledó et al, Fiscal Rules at a Glance.

50 Kelemen and Teo, “Law, Focal Points and Fiscal Discipline”.

51 Although the direction of the effects is the same as in the main analysis.

52 World Bank, World Development Indicators.

53 Djuve et al., “Patterns of Regime Breakdown”; Kennedy, “The Contradiction of Modernization”.

54 Ulfelder, “Natural-ressource Wealth”; Wright et al., “Do Authoritarian Institutions Constrain?”.

55 Djuve et al., “Patterns of Regime Breakdown”; Geddes, “What Do We Know”. One could argue that the inclusion of GDP related variables would be potentially ‘post treatment’ if a stricter fiscal rules framework would improve economic performance. However, the results remain fairly stable if the two GDP related variables are excluded from the estimation; see Appendix E. Additionally, to take into account that it might be economic crises rather than low growth rates that drive regime instability in autocracies, Appendix F contains a version of the main analysis in which a ‘banking crisis’ dummy is added to the estimation (using data from Laeven and Valencia) as well as a version in which all positive growth rates are coded as ‘0,’ and the estimation thereby only takes negative growth rates into account. (See Laeven and Valencia, “Systematic Banking Crises Revisited”.) The results are robust to these alternative specifications.

56 Green et al., “Dirty Pool”.

57 Alt and Lassen, “Fiscal Transparency”, 1418-1419.

58 Hellevik, Linear versus Logistic Regression”.

59 Beck and Katz, “Throwing Out the Baby”.

60 Recent research (Kropko and Kubinec, “Interpretation and Identification”) has cast some doubt about the usefulness of two-way fixed effects estimation. However, as shown in Appendix E, the results are substantially similar if the year-fixed effects are replaced with a linear time trend or if the year-fixed effects are removed altogether.

61 Carter and Signorino, “Back to the Future”.

62 Hanson, “Democracy and State Capacity,” 315-316. Additionally, in Appendix E, we also include an error correction model using Stata’s xtpmg function, which allows us, in an alternative way, to parse out short and long run effects (see Blackburne and Frank, “Estimation of Nonstationary Hetergeneous Panels”).

63 Box-Steffensmeier and Jones, Event History Modeling: A Guide for Social Scientists.

64 The full regression results, both with and without control variables, can be found in Appendix G.

65 Since fiscal rules could potentially have both positive and negative effects on our dependent variable, we rely on a two-tailed test of statistical significance. Due to the limited number of observations, and the fact that there is limited within-country variation, we use 90-percent confidence intervals.

66 Treisman, “Income, Democracy and Leader Turnover”.

67 These additional control variables are all measured in the same year as the fiscal rules index and thus follow the same lag structure.

68 Data for gross debt are from the IMF’s World Economic Outlook Database, whereas the IMF webpage is used as the source for current and previous IMF lending programs.

69 Aaskoven, “Signaling to Creditors and Voters”.

70 Data are from the ‘Autocratic Regimes’ dataset by Geddes et al., (“Autocratic Breakdown and Regime Transitions: A New Data Set).

71 A dummy that takes the value ‘1’ in case of national elections (executive and/or legislative) in a country-year. Data are from version 5 of the NELDA dataset (see Hyde and Marinov, “Which Elections Can Be Lost?”).

72 Knutsen et al., “Autocratic Elections”.

73 Shea and Poast, “Leaders and Default”.

74 Hall and Lawson, “Economic Freedom of the World”.

75 Bernhard et al., “Institutionalising Electoral Uncertainty;”; Knutsen et al., “Autocratic Elections”.

76 Lührmann and Lindberg, “A Third Wave of Autocratization is Here”.

77 Treisman, “Income, Democracy and Leader Turnover”.

78 These effects echo the effects of foreign capital inflows in autocracies (see Ahmed, The Perils of International Capital).

79 See also Meyerrose, “The Unintended Consequences of Democracy Promotion”, for an argument about the role of international organizations in (unintentionally) promoting autocracy.

80 Boix et al. “A Complete Data Set of Political Regimes, 1800-2007”.

81 Boix et al. “A Complete Data Set of Political Regimes”.

82 Djuve et al., “Patterns of Regime Breakdown”.

Additional information

Notes on contributors

Lasse Aaskoven

Lasse Aaskoven is Assistant Professor at the Department of Political Science and Danish Institute for Advanced Study, University of Southern Denmark. His research spans several areas in Comparative Politics and Comparative Political Economy. One of his main research topics is how fiscal institutions, such as fiscal rules, impact national politics in both democratic and non-democratic countries.

Alexander Taaning Grundholm

Alexander Taaning Grundholm has a PhD in Political Science from Aarhus University, Denmark, and is currently working as an independent researcher within the field of Comparative Politics. His main research topics include autocratic regimes and political regime development, with a special focus on dictators' survival strategies.