A new paper (pdf) from Jessica Perez, Gabe Horwitz, and David Kendall of Third Way alleges that increased spending in Social Security, Medicare, and Medicaid is crowding out spending on investments like education, research and development, and infrastructure. The evidence presented amounts to the fact that investment spending as a percentage of GDP has decreased over the past 50 years, while spending on social insurance programs has gone up.

There are two reasons why this is misleading. For one thing, the authors group Social Security, Medicare, and Medicaid together as "entitlements." But Social Security has been growing much more slowly than Medicare and Medicaid, according to the OMB budget tables:

Social Security spending has been basically stagnant since the 1980s, while health care spending has increased sharply. So it isn't "entitlements" that are crowding out investment, it's health care. That suggests if policymakers want to reduce "entitlement" spending, they should be thinking about how to control health care cost generally.

The second problem with the Third Way paper's approach is that it presumes that the increase in Social Security/health spending caused the drop in investment spending. But just because less is spent on one thing and more on another does not mean that one is crowding the other out. Indeed, if you break down the decline in investment spending into defense and non-defense categories, it becomes obvious that the patterns in investment spending have a lot more to do with the arc of the Cold War:

Non-defense investments in education shot up under Lyndon Johnson and then stagnated under Nixon, while the détente policy of Nixon, Ford, and (to a lesser extent) Carter, as well as winning the race to the moon, led to a decline in defense investment throughout the 1970s. Then Reagan cut education and other social investments while increasing defense spending. Once Gorbachev took the helm of the Soviet Union in the mid-1980s, defense spending started to fall.

Why does this story make more sense than Third Way's "crowding out" narrative? Well, for one thing, the geopolitical account can make sense of why both defense and non-defense investments have been stagnant or grown for the past 20 years even as spending on Medicare and Medicaid has grown. If entitlement programs are pushing down investment spending, why didn't they do so during that period?

One last thing. Third Way argues that investment spending is highly correlated with economic growth, using the following graph:

Source: Third Way

Though the overall story here is correct, there are two problems. One, as you can see even in Third Way's graph, the correlation was much stronger before the 1980s than since then. Second, this doesn't differentiate between defense and non-defense spending. From 1962 to 1985, defense investment spending explained 30.4 percent of the change in growth rates, and 54.8 percent of the change in technological development, as measured by the San Francisco Fed's total factor productivity figures. From 1985 to 2011, it only explained 3.1 percent of the change in growth rates, and 2.5 percent of the change in technology:

Non-defense investments are a slightly different story. Between 1962 and 1985, non-defense investments explained 18.6 percent of variation in growth, and 25.5 percent of variation in technological development. Between 1985 and 2011, they only explained 11.2 percent of the variation in technological development, but 35 percent of the variation in growth.

All of which is to say, not all investments are created equal. The defense investments we made in the 1960s and 1970s were associated with growth and technological development, and the ones since then haven't been. The non-defense investments we make have gotten worse at helping technological development and better at helping growth. So if you want to increase growth, the key isn't just increasing investment full stop, it's identifying the types of investment that yield growth and the types that don't.