Power Shift

Political scientists have focused, for understandable reasons, on changes or threats to the "separation of powers" among the executive, legislative, and judicial branches. The enormous volume of scholarship on separation of powers tends to obscure and overshadow research on the "vertical" separation of powers, or federalism. As many scholars have shown in recent years, the vertical distribution of powers is of vital importance to the understanding of American government (Bowman and Krause 2003, p. 302). It is not just the early founding generation who argued over the "line to be drawn between the authority of the general government and that of the several states" (McDonald, 2000, p. vii, quoted in Bowman and Krause, 2003, p. 302).

What are the causes of centralization, and how do we measure a system "out of balance"? These questions continues to provoke debate, both in the theoretical literature and- increasingly- in the empirical literature. Below, we offer a selection of the most important efforts by scholars in recent decades to measure or track shifts in power to the national government away from states and localities. The following studies and measures provide an important first step in understanding how the American federal system works and the myriad ways in which power has shifted between levels of government over time. 

  • moneyup
  • Deezer
  • worldglobe
  • moneyshift
  • percentage

Federal outlays

Employment share

IMF data

Power Shift


Policy Centralization

Scholars Ann O'M. Bowman and George A. Krause have analyzed patterns of policy centralization by the federal government over time. The figure below represents a sophisticated "score" for combined public laws and executive orders over time. The authors show, that while there have been decentralizing "pushes" at times, the "pulls have been far more powerful than the pushes." The analysis below challenges the assertion of scholars who claim that power has shifted toward states and localities since the postwar era.  

The CCS Federalism Index would like to acknowledge the work of scholars Ann O'M. Bowman and George A. Krause, for access to the original data. The following represents our attempt to summarize one measure or one section of this rich, sophisticated analysis by the researchers. There is no substitute for access and study of the original published research: Bowman, Ann O'M., and George A. Krause. "Power Shift: Measuring Policy Centralization in U.S. Intergovernmental Relations, 1947-1998."American Politics Research 31, no. 3 (May 2003): 301-25. doi:10.1177/1532673X03251381.

Measuring De / Centralization

Americans have debated centralization versus state autonomy since the founding. "No question of government has been more vigorously debated than...centralization versus states rights," once wrote (future) Senator Paul H. Douglas (quoted in Kincaid 2019). Measuring de/centralization, however, has proved to be difficult. The below study by federalism scholar John Kincaid is a rigorous, thought-provoking analysis of the centralization of policy functions from 1790 to 2010. Part of a larger project to measure centralization in six federations, Kincaid's research provides a powerful overview not only of the cumulative effects of centralization but also a major contribution to the study of how centralization has occurred across policy dimensions. Kincaid finds evidence across twenty-two policy fields and five fiscal indicators for increasing centralization. Kincaid finds that centralization is evidence "across the twenty-two policy fields," with the largest shift occurring on legislative dimensions. 

For analysis see John Kincaid, Dynamic De/Centralization in the United States, 1790-2010, Publius: The Journal of Federalism, Volume 49, Issue 1, Winter 2019, Pages 166-193, https://doi.org/10.1093/publius/pjy032.

Screen Shot 2021-05-06 at 9.55.12 AM.png

  • In a review of previous literature, Kincaid notes that all previous studies have reported centralization (albeit using different methods and with limitations on the validity of measures)

  • In his own research, Kincaid finds evidence for both legislative and administrative centralization across twenty-two policy fields. Trends move from "almost exclusively state in 1790" to predominantly federal" over time

  • However, consistent with other research (see above, "Power shift"), there are variations in the timing, sequence, and pace of centralization across fields

  • Notably, Kincaid finds that fiscal centralization is less evident over time. Four of Kincaid's five fiscal measures, for example, experienced "mild centralization." This supports the view that the United States is comparatively non-centralized, fiscally (citing Kim Junghun, ed., Institutions of intergovernmental fiscal relations. Paris: OECD 2015)

  • The magnitude of legislative centralization is nearly twice as large as administrative change (consistent with the idea of states "serving as administrative agents of the federal government")

  • The pace of centralization was slow, overall, "picking up speed in the twentieth century." Centralization spiked during the New Deal. Yet, more "substantial centralization occurred from 1960-1980 as Congress and the Supreme Court acted increasingly on social matters and expanded federal power into the remaining policy fields, including areas previously having few federal constraints, such as culture, education, environmental protection, health, and language." 

  • Measurement indicates that states have "lost" more legislative than administrative autonomy, suggesting that we are moving toward "indirect federal administration" (quoting Macmahon 1972, 22), a system in which states are increasingly administrative agents for the national government. 


We believe the above analyses merit extra attention, not only for researchers, but teachers and policymakers and members of the general public. These two projects on "Power Shift" and "De/Centralization" provide an essential starting point for any inquiry in to the measurement of American federalism, but also, interpretation and analysis of centralizing trends. 

In particular, readers interested in changes to the federal system should take note of the primary empirical conclusion, namely, that the federal system has moved almost exclusively from state authority in the founding (1790) to the modern era (2010). Four caveats, however, from this same research suggest a need for extra nuance, and perhaps more research:


  1. states still wield significant administrative autonomy, despite the centralizing forces;

  2. there is less evidence for radical fiscal centralization, suggesting a need for nuance and perhaps more data in debates over fiscal federalism;

  3. triggers for centralization are multifaceted and complex and should be included in discussions or attempts to "redraw" the boundaries;

  4. the Constitutional basis for centralizing fields is ambiguous, and likely to provoke subjective analysis. Perhaps these debates are healthy and are to be welcomed as one of the many side-effects of centralization over time.