Call me a policy nerd if you like. I don’t mind. Part of my job involves commissioning, reading, interpreting, and writing about public policy research. As I began researching my 2012 book on North Carolina’s economy, Our Best Foot Forward, I found myself reading or re-reading dozens of scholarly studies, published in academic or professional journals, that examined some aspect of state economic growth. These research findings became the core of my argument in the book, which recommended an empirically based strategy for economic recovery that included structural reforms of North Carolina’s tax code, regulatory process, education system, transportation infrastructure, and other programs.
Even after Our Best Foot Forward was published, I continued to add new findings to my database, which now contains 258 peer-reviewed studies on state economic growth published from 1992 to 2013. Some of the studies examine only a single policy variable, such as corporate tax rates or road mileage. Other studies include several different variables in their models, so there are actually 411 separate findings in my database.
For many common questions about the relationship between public policy and economic growth, the scholarly literature offers clear answers. For example, state and local taxes matter. Of the several dozen studies that spoke to the issue, 63 percent found a negative association between overall tax burdens and some measure of state economic performance such as GDP, personal income, employment, and investment flows of financial or human capital. The proportion was even higher for corporate or business taxes (68 percent) and marginal income tax rates (69 percent).
Education and infrastructure also correlate with state economic growth, but the relationship is complex. Measures of educational outputs, such as student test scores or the share of workers with college degrees, were positively linked to economic growth 57 percent of the time. The amount or quality of highways and other infrastructure were positively linked to economic growth 70 percent of the time. On the other hand, government expenditures on education and infrastructure were not consistently linked to economic benefits. Higher-education spending, for instance, exhibited positive economic effects in 29 percent of studies, negative effects in 24 percent, and mixed or statistically insignificant results in the remaining 47 percent.
How could government spending on education not confer economic benefits? One reason is that there is no consistent relationship between expenditures and outcomes. Some education dollars are clearly not spent in ways that truly improve student learning. Because the spending must be financed by taxing households and businesses, there are many cases in which the economic cost of the taxes required is equal to or greater than the economic benefit of the spending.
If that sounds odd to you, imagine that a politician promised to quadruple North Carolina’s current tax burden in order to fund a proportionately higher level of state spending on public schools, colleges, and universities. Even if you are a hard-core leftist, wouldn’t you honestly have to doubt whether such a policy would be worth it? That means you admit that there has to be a point of diminishing returns. Where we may disagree is the location of that point on the cost-benefit curve, not its existence.
All this having been said, there are some questions in public policy for which my growing research database has yet to produce a clear answer. A highly relevant question for state policymakers is whether it is wiser to fund education, infrastructure, and other programs primarily at the state level, as North Carolina does, or primarily at the local level, as most other states do.
While there is a growing international literature about fiscal centralization, I have been able to locate only 10 modern studies about its relevance to American states and localities. In four of them, centralized fiscal policies such as North Carolina’s had a positive effect on economic growth. In four of them, such policies had a negative effect. The other two studies were inconclusive.
I’ll keep looking for more data. That’s what policy nerds do.