Constraints on the executive and tax revenues in the long run

Antonio Savoia, Kunal Sen, Abrams Tagem

Research output: Contribution to journalArticlepeer-review

Abstract

We argue that tax revenues and political institutions placing constraints on the executive power may reinforce each other over time and so co-evolve in the long run. This may also bring a shift in the composition of revenues, from taxes levied on a narrow base to broadly levied taxes. To test these hypotheses, we use historical cross-country data covering 31 countries for 1800-2012 and panel time series methods allowing for different forms of country-specific heterogeneity and cross-section dependence. The results offer three main findings. First, executive constraints, whether they are judicial or legislative, and tax revenues are cointegrated. While in the short run they can drift apart, this will be temporary because there is a long-run relationship between the two. Second, evidence of cointegration is strongest for revenues from direct taxes, suggesting that the existence and nature of a long-run relationship may be mainly related to the emergence of broad-based taxation. Third, long-run causality runs mostly from executive constraints to taxation. This is most evident for income taxes. Our findings link Sustainable Development Goals 16 and 17, implying that the goal of promoting inclusive and accountable institutions may work in synergy with that of generating internal resources to finance development goals.

Original languageEnglish
Pages (from-to)314-331
Number of pages18
JournalJournal of Institutional Economics
Volume19
Issue numberS3
DOIs
Publication statusPublished - 14 Jun 2023

Keywords

  • Constraints on the executive
  • SDG17
  • fiscal capacity
  • institutions
  • tax revenues

Research Beacons, Institutes and Platforms

  • Global inequalities
  • Policy@Manchester
  • Global Development Institute

Fingerprint

Dive into the research topics of 'Constraints on the executive and tax revenues in the long run'. Together they form a unique fingerprint.

Cite this