Until today, I had not understood the trends shown in the graph above (from Aherne, Labonte, & Lynch 2024).

As a proportion of the economy, total federal spending has been fairly constant since 1962. Entitlements (Social Security, Medicare, Medicaid) and defense keep the whole cost pretty stable. The cost has risen during recessions because bad times increase eligibility for entitlements. This means that the early Reagan years saw a temporary peak in total federal spending (notwithstanding Reagan’s anti-government rhetoric), and the Great Recession and COVID caused big temporary increases.

Meanwhile, federal discretionary spending quite steadily declined from 1965 and 2000. It has fluctuated since then from a lower baseline.

That means that the basket that includes highways and air traffic control, prisons and border control, diplomacy and foreign aid, agricultural subsidies, Food Stamps, etc. represents a smaller percentage of the economy than it did in the 1960s.

Looking more closely at components, we can often find anomalous patterns. For example, total federal spending on education (k12 and college, including financial aid and research) was 1 percent of GDP in 1975 and 1.1 percent in 2024, with spikes during recessions.

Since the economy has grown each decade, a shrinking proportion of GDP could still purchase more goods and services. But that has not really happened during the 21st century. Another telling graph from the same report (below) shows discretionary spending in billions of dollars, adjusted for inflation. It separates defense from non-defense spending. Until COVID hit, neither component had risen (or fallen) in real terms compared to 2005. The Obama stimulus did cause a temporary boost, but that went away. Then COVID spending and the Biden stimulus boosted non-defense spending, which has come down but remains about 25 percent higher than it was in 2019.

These graphs explain why the kinds of public goods that we expect from the national government in the United States often seem to have shrunk or deteriorated, even while the total cost and size of the federal government has remained at least constant.

These data challenge certain assumptions popular among conservatives–that federal spending has risen and that Republican presidents have cut government while in office. (By the way, Elon Musk’s recent rampage will hardly be visible on these graphs when the lines are extended into 2025. Total federal spending rose during the first quarter of 2025.)

These graphs also challenge progressives’ assumptions that government has been shrinking in the era of neoliberalism. Indeed, even discretionary domestic spending is quite a bit higher than it was in 2005 or 2012-19, when adjusted for inflation. What progressives observe is not a shrinking government but a decline in non-defense discretionary spending (as a proportion of the economy) between 1965 and 2000, which has left many national government functions weaker than they were in the mid-1900s.

---

Source: Aherne, Drew C., Labonte, Marc & Lynch, Megan S., “Discretionary Spending in 10 Graphs” (2024), Congressional Research Service https://www.congress.gov/crs-product/R48164. See also taxing and spending are more compatible with democratic values than regulation is; how public opinion on social spending has changed: a generational approach