Direct Decentralization: Ballot Measures as a Centralization Foil

• In Event: State Politics and Policy & New Directions in Political Economy Virtual Poster Session
  In Virtual Poster Session: Virtual Poster Session: State Politics and Policy

Abstract

Scholars have long documented a march towards centralization in the politics and policy of federal nations. Despite constitutional provisions to ensure strong subnational governments, national courts and legislative bodies find rationales for shifting responsibilities upward. But is this inevitable? In this paper, I advance the discussion of how decentralization can prevail with use of direct democracy. I argue that when citizens are the locus of policymaking, they will be more supportive of proposals which grant them control over policy decisions. I test my theory by coding 4,500 ballot measures on centralization’s impact. I find that citizen-proposed ballot measures are decentralizing in nature and that centralization is negatively associated with probability of passage. To understand why citizens support decentralization, I leverage polling data and coarsened exact matching (CEM) – finding that support for decentralizing ballot measures is conditioned by trust in state-level government. These findings pave the way for advancing research on decentralization.

For over a half-century, scholars have identified an unrelenting march towards centralization in politics and policy of nations. An abundance of research in political science (Dickovick 2011; Kincaid 2019; Malesky, Nguyen, and Tran 2014; Riker 1964; Scott 1909; Veliz 2014), and public policy/administration (Bowman and Krause 2003; Grossman and Lewis 2014; Kincaid 1998; Konisky 2011; Marchildon and Bosset 2018) has found consistent evidence of this centralizing tendency. Termed over-centralization, analysis of this question has been of paramount importance for scholars interested in federalism and intergovernmental relations.

Fundamentally, federalism is endangered when decision making continues to be centralized at the national level (Braun 2011). This is problematic as: (1) citizens are broadly supportive of a federated – or decentralized – system of governance (Dinan and Heckelman 2020); and (2) federalism is associated with numerous institutional benefits including economic gains (Qian and Weingast 1997), improved representation (Brzinski et al. 1999, but see Wiezien and Soroka 2011), innovation (Rabe 2004; Nivola 2017), policy outputs (Jenkins and Henry 2016), and enhanced citizen participation in government (Derthick 2001, 2008; Eichenberger 1994). Federalism in the United States is in a period of centralization due in part to increased political polarization – with an increasing reliance on unilateral executive action, congressional inaction, and federal-state arbitration through the federal courts (Goelzhauser and Konisky 2019). Similar centralization is evident at the state-local area where states are increasingly preempting local decisions, taking on more responsibilities of local governments, and limiting local revenue options (DuPuis et al. 2014; Fowler and Witt 2019; Hicks et al. 2018; Phillips 2018).

How can over-centralization be curtailed? Federalism presupposes a shared relationship between national and subnational governments, not one where one government dominates the other. Elazar (1987) suggested that a robust federal culture was necessary to preserve a balanced federation. Bednar (2008) provides the most satisfying answer to date, arguing that there are constitutional and judicial safeguards that – under certain conditions – can preserve governmental balance. While previous scholarship no doubt aids in expanding scholars’ understanding of the push and pull of centralization within nations, it remains hamstrung by relying almost exclusively on federalism theory.

I address the question of centralization by looking at another institution — direct democracy. I posit a theory of direct decentralization – arguing that when citizens can determine which level of government resolves political and policy issues, they will favor localized government over broader, centralized levels of government. I evaluate this theory by coding the centralizing impact of 4,500 ballot measures across twenty-five states and 120 years in the Unites States. I find evidence for direct decentralization: citizen-proposed ballot measures are consistently decentralizing while legislative-proposed ballot measures are mildly centralizing. Additionally, regardless of who proposes a ballot measure, decentralizing ballot measures are more likely to succeed than centralizing ballot measures. Finally, leveraging polling data and CEM, I find that political trust determines when a citizen is supportive of centralizing ballot measures.

The results presented in this paper provide several implications for scholars. First, they provide a partial institutional answer to the puzzle of over-centralization. Second, they illustrate the importance of integrating federalism scholarship with other political and policy fields. Third, this paper raises several new questions about how decentralization can influence – and be influenced by – the political and policy process.

Author

• Matthew Joseph Uttermark, Binghamton University -- SUNY