Centralization

Top Takeaways

Federal Outlays

During the 20th century, the public sector grew dramatically in many or most developed countries (Di Matteo 2013). The average level of public spending expressed as a share of GDP (for 13 select industrialized countries) rose from 12.3 percent in 1913 to 27.9 percent by 1960, reaching a high point of 43 percent in 1990 (Tanzi and Schuknecht, 1997a, 165; 1997b)

Government expenditure to GDP ratios have grown across most developed countries. This ratio was usually below 10 percent in the 19th century, and grew to well over 40 percent of GDP by the late twentieth century. After growing for much of the twentieth century, public sectors began to decline in size after 1980 (Di Matteo 85). According to Di Matteo, the first decade of the twentieth century saw a "resumption of growth in public sector size" (the average government expenditure to GDP ratio for the world had climbed back to 33 percent"). Di Matteo concludes that "there has been a reversal of the international trend towards smaller government that marked the 1980s and 1990s"

Despite major advances in research techniques and data collection, however, these and other such measures remain problematic, especially when it comes to understanding American federalism. Despite the impressive volume of economic studies, it remains difficult to measure the "size of government" in the American federal system (see Higgs 1991, p. 21-34 for a very helpful introduction to measurement problems).

Employment Share

Still, it can be useful to use economic indicators as a rough cut. In American federalism studies, the most common measure of government size are:
- Government spending as a share of national output (GDP/GNP)
- Government spending per capita
- The number of public sector employees
- Public sector employment as a share of total employment

As shown above, these conventional measures can be useful, if used selectively. For example, using government spending as a share of national output, Higgs (1991) estimated that "[US] government has grown about five to six times larger relative to the economy during the twentieth century" (p. 25)

Other measures of government size can be misleading, however. Federal employment data is especially problematic, as it does not count the millions of individuals who are employed indirectly (e.g., through federal grants to states, contracts to for-profit businesses, and other aid programs)

IMF Data

When it comes to measuring federalism, most quantitative indexes of the size of government share a similar defect: they (generally) do not account for transfer payments, and other means of subsidizing state or local governments - thus obscuring the amount of involvement of the federal government

This dashboard does include analyses of the impact of the level of public spending on economic performance or social indicators. Readers will have to consult other sources for analyses of proposed "ideal" levels (e.e.g Tanzi and Schuknecht 1997a). We also omit normative analyses, which answer questions like "Are taxpayers getting value for their money"? "Is a larger public sector correlated wit more and better public services?" etcetera. Generally, economists seem to agree that while some public sectors are more efficient than others, more and larger government is not always associated with improved outcomes

As for federalism, or decentralization, readers will need to consult other studies. Some studies show that decentralization may improve accessibility and local responsibility, thereby making government more effective as it moves "closer to the people" (see Bird, 1993: 209). However, the empirical literature on the effectiveness of decentralization on restraining public sector size is mixed (see Di Matteo 95). States and provinces are often marked by substantial expenditure growth.

Power Shift

When it comes to measuring federalism, most quantitative indexes of the size of government share a similar defect: they (generally) do not account for transfer payments, and other means of subsidizing state or local governments - thus obscuring the amount of involvement of the federal government

This dashboard does include analyses of the impact of the level of public spending on economic performance or social indicators. Readers will have to consult other sources for analyses of proposed "ideal" levels (e.e.g Tanzi and Schuknecht 1997a). We also omit normative analyses, which answer questions like "Are taxpayers getting value for their money"? "Is a larger public sector correlated wit more and better public services?" etcetera. Generally, economists seem to agree that while some public sectors are more efficient than others, more and larger government is not always associated with improved outcomes

As for federalism, or decentralization, readers will need to consult other studies. Some studies show that decentralization may improve accessibility and local responsibility, thereby making government more effective as it moves "closer to the people" (see Bird, 1993: 209). However, the empirical literature on the effectiveness of decentralization on restraining public sector size is mixed (see Di Matteo 95). States and provinces are often marked by substantial expenditure growth.

There is a lack of systematic empirical inquiry on the question of de/centralization, despite a rich qualitative literature

"Power shifts" are difficult to measure, and display (as above) considerable variation year to year

Peak policy centralization occurs in 1968 ("the highest sum score of the half-century time period")

Measurement for a drop in centralizing activity appears to coincide with Nixon's "New Federalism" agenda. The 1960s is a decade of both high centralization and also "reaction" against centralizing forces

Decade by decade, there are more scores above the neutral point of zero, with negative scores only occasionally punctuating an otherwise overall trend toward centralization (see, for careful analysis, Bowman and Krause p. 310)

According to scholars Bowman and Krause, "the greatest amount of decentralization (...1953, 1955, 1972, and 1981) occurred under Republican presidents" (p. 310)

For context, however, average centralization occurs at a steady increase, with centrifugal shifts that "tend to be low in magnitude and infrequent in occurrence" (p. 310)

American federalism, overall, during the postwar era, is marked by "a general upward movement by a factor of 0.44 and 0.05 units for each subsequent year for the summed and averaged measures" (p. 310)

The evidence supports a standard view that federal balance of power has shifted "toward the federal government and away from subnational levels" (p. 310)

Greater decentralization, often associated with the Nixon or Reagan era, is not supported by the data

Evidence is highly suggestive that the pattern of power "decentralization" does not support philosophical or rhetorical signals of devolution, bolstering other accounts that suggest a disjunction between the language of devolution and reality of power centralization. See, for example, Donahue, J.D. (1997) Disunited States. New York : Basic Books. Greve, M.S. (1999). Real Federalism: Why it matters, how it could happen. Washington, D.C.: AEI Press.

Aggregate data, while not perfect, are useful in providing a "big picture view" of changes to the American federal system over time (p. 312). More research is needed (a "more nuanced examination") of the ways in which centralization is/is not constant "across different policy areas."

Conclusion: readers interested in centralization or "power shift" are encouraged to explore centralization in or among agencies and policy areas. Our measures support and confirm bowman and Krause's conclusion, that an overall dominance of centralizing actions by the federal government, even or despite policy area*******************

Centralization

Top Takeaways

Federal Outlays

Employment Share

IMF Data

Power Shift

Recommended Readings

The Federalism Index Policy Tracker is live

Please help us improve this version of the Federalism Index by emailing support@federalismindex.org

Centralization

​

Top Takeaways

Federal Outlays ​

Employment Share

IMF Data

Power Shift

​

Recommended Readings

Federal Outlays