The ‘living wage’ fallacy
The April jobs report was bad news for the post-COVID-19 economic recovery. Across the United States, employers added just 266,000 jobs and the unemployment rate ticked up to 6.1%. This…
Supreme Court Justice Louis Brandeis famously described the states of the union as “laboratories of democracy” where a “state may, if its citizens choose…try novel social and economic experiments without risk to the rest of the country.”
State governments have done just that. Comparisons between Minnesota and its neighbors—Wisconsin, Iowa, South Dakota, and North Dakota—offer rich examples of how tax rates and regulatory burdens differ significantly among states. As of January 2020, the top rate of state income tax ranged from 0 percent in South Dakota—it levies no state income tax—to 9.85 percent on income over $164,400 in Minnesota, the fifth-highest top rate in the country. Like$10.00 wise, state corporate income $9.75 tax rates range from 0 percent $9.50 in South Dakota—it levies $9.25 no state corporate income tax—to 12 percent on income over $250,000 in Iowa, the highest top rate in the United States. Finally, state sales tax rates range from 4.5 percent in South Dakota to 6.875 percent in Minnesota, the sixth-highest rate in the United States, albeit with many exemptions. These five states are spread widely a cross rankings of state tax regimes. The Tax Foundation’s 2020 State Business Tax Climate Index ranks Minnesota 45th, Iowa 42nd, Wisconsin 26th, North Dakota 16th, and South Dakota 2nd.
Table 1 shows the variation in major tax rates among Minnesota and its neighbors in 2010 and 2018. Over this period, corporate tax rates and brackets in the states remained unchanged except for in North Dakota, where rates fell. The sales tax rate changed only in South Dakota, with a 0.5 percentage point increase. There is more variation when it comes to state income taxes. South Dakota’s remained unchanged at 0 percent, and lawmakers increased the thresholds for Iowa’s brackets largely to guard against inflation. North Dakota retained five brackets but sharply reduced the rates, and Wisconsin both lowered rates, albeit by less than North Dakota, and reduced the number of brackets. At the other end, Minnesota added a new top tax rate above its former top rate. Overall, while South Dakota and Iowa saw little or no change in income taxes, Wisconsin and North Dakota both cut them and Minnesota increased them on the highest earners. It is also worth noting that Minnesota taxes its lower earning residents at a comparatively high level. It levies a 5.35 percent tax on the first dollar of taxable income, a rate that, in 2018, Wisconsin didn’t levy until the resident earned $11,230, Iowa didn’t levy until the resident earned $14,382, and North Dakota levied on nobody at all.
Analysts find it more challenging to quantify the overall burdens of government regulations. First, merely counting the number of regulations yields an imprecise analysis at best. Businesses, for example, might find it easier to comply with 10 precisely worded regulations than one with vague wording. Second, the burden of a regulation can also be tied to how zealously it is enforced.
We can, though, quantify some specific regulations, such as the minimum wage. A state can impose its own, or use the federal rate. As Figure 1 shows, three of Minnesota’s neighbors—Wisconsin, Iowa, and North Dakota—use the federal minimum wage as the effective in-state minimum wage. This was the case in Minnesota until 2014 and in South Dakota until 2015, when the two states raised their minimum wage rates above the federal level.
The “natural experiments” that such variations create allow us to assess the impacts of particular economic policies on economic performance.
Analysts frequently attribute the economic outcomes in Minnesota and Wisconsin to the differing policies of the two states. After the November 2010 gubernatorial elections of Democrat Mark Dayton in Minnesota and Republican Scott Walker in Wisconsin, the Economic Policy Institute wrote: “The two states’ geographic proximity—as well as their similarities in population, demographics, culture, and industry composition—make comparing outcomes in Wisconsin versus Minnesota a useful natural experiment for assessing how state policy is affecting economic outcomes and residents’ welfare.”
But policy analysts should not push these arguments too far. Minnesota and Wisconsin differ in ways beyond economic policy that could account for at least some of the disparate outcomes.
One difference can be the positive contributions of big cities as engines of economic growth. Urbanization—as labor moves from agriculture to industry and services—reduces the per capita costs of providing infrastructure and government services; knowledge spillovers and specialization enhance worker productivity. So, Minnesota’s economic advantage over Wisconsin might be due to its large urban areas more than the state’s economic policies.
The Twin Cities, for example, comprise the 15th largest Metropolitan Statistical Area (MSA) in the United States by Gross Domestic Product (GDP). The next largest MSA by GDP in Minnesota and its four neighbors is Milwaukee, which ranks 37th nationally. While Minnesota derives 80 percent of its GDP from five MSAs, Wisconsin gets 80 percent of its GDP from 12 MSAs. The Twin Cities MSA accounts for 71 percent of Minnesota’s GDP; Milwaukee provides 31 percent of Wisconsin’s GDP.
True, state policy can contribute to the rise or fall of cities, but this tends to happen over a term too long to be driven by a particular administration—the Twin Cities have been among America’s most populous metropolitan areas since 1880.
Economists can most accurately assess the impact of state economic policy by excluding all possible non-policy factors that help determine economic outcomes. In Why Nations Fail, economists Daron Acemoglu and James A. Robinson illustrated the importance of institutions in driving economic outcomes by comparing living standards in Nogales, Arizona in the United States with those in Nogales, Sonora in Mexico. The town is divided by an arbitrary line with economic outcomes, such as wages, household incomes, or per capita GDP, vastly different on either side. Variables such as “geography, climate, or the types of diseases prevalent in the area” can be assumed to be the same on both sides of the line and, so, can be excluded as causes of observed differences. What remains is simply what side of the line you are on, which determines the institutions you live under.
This situation illustrates the value of comparing border counties. As with Nogales, we can assume that factors such as geography or demographics are fairly uniform between, say, Washington County in Minnesota and St. Croix County on the other side of the I-94 bridge in Wisconsin.
Analyzing border counties enables us to filter out some of the distorting effects like the presence of the Twin Cities. North Dakota illustrates this. Its economy has recently generated the highest real GDP growth in America, increasing by 40.3 percent between 2010 and 2018. But 34.3 percent of that growth comes from just two counties, McKenzie and Williams, in the heart of the Bakken oil fields. True, choosing to develop natural resources is an aspect of state policy—as Minnesota’s hesitance to develop its non-ferrous mining demonstrates—but geology, not policy, drives the allocation of such resources. Therefore, it would be inaccurate to attribute North Dakota’s superior GDP growth—Minnesota grew by 17.2 percent between 2010 and 2018—to superior state policy.
Because McKenzie and Williams counties are on the opposite side of North Dakota from its border with Minnesota, economists won’t attribute differences in economic outcomes between Cass County in North Dakota and neighboring Clay County in Minnesota to the geological windfall of the Bakken oil fields. The economies of Minnesota and North Dakota might be too different to compare too closely on a statewide basis, the same cannot be said about the economies of Moorhead and Fargo.
Comparing the economies of neighboring border counties provides an excellent way to assess the impact of state economic policy on economic outcomes. My analysis compares the relative performance of border communities from 2010 to 2018.
The outcomes fall into three broad categories: where people want to be, where businesses want to be, and how people’s living standards are impacted.
To look at where people want to be, we use data on resident population from the Census Bureau. To get some idea of whether people are moving to or remaining in these areas because they anticipate opportunities, we look at Census Bureau data on median age of the population. To understand how these movements are likely to impact the productivity of the labor force, we use data on the share of the population with a bachelor’s degree or higher, again from the Census Bureau.
To see where businesses want to be, we look at data on the number of private establishments from the Bureau of Labor Statistics (BLS). We look at how successful these businesses have been at expanding using data on total employment and the composition of employment, again from the BLS.
Finally, to measure people’s living standards, we look at per capita Personal Income data from the Bureau of Economic Analysis. Personal Income contains income from three categories: labor income, capital income, and transfer income, and we also look to see which of these categories has driven changes in total per capita Personal Income. We also look at Census Bureau data on the number of people in poverty.
We look primarily at growth rates and changes over the period 2010 to 2018. Levels of something like Personal Income can, as the size of the Twin Cities or the Bakken oil fields show, reflect policy choices or natural resource endowments of long ago. Rates of change, on the other hand, are more likely to reflect the impact of the particular policies we wish to assess.
No single one of these measures on its own explains how state policy affects the economy. What we need to do, having pulled the data together, is to step back and see the big picture.
We see that Minnesota performs best against Wisconsin. The picture is somewhat different when we look at the other borders, where the distorting effects of a large MSA are mostly absent and where the borders are more porous. Minnesota loses to Iowa and North Dakota and fairly comprehensively to South Dakota.
What do we see when we compare these observed results to state economic policies, particularly the tax policies in Table 1? Minnesota fares best against Wisconsin, which ranked as a lower tax state and which, over this period, lowered its income tax rates and reduced the number of brackets. This statistic might seem to contradict the consistent research finding that high taxes have significant negative effects on economic growth. But bear in mind that the Twin Cities MSA—a legacy of past development, not of current policy—straddles this border and drives much of Minnesota’s strong performance here. Also, a lesser factor, the rivers Mississippi and St. Croix run along much of it, expanding the “policy space.”
The other comparisons support that consistent research finding. When we compare the performance of Minnesota—where taxes are high on all and the top rate was increased—to Iowa, where a lighter tax burden was largely held, Minnesota loses. Minnesota’s performance loses handily to North Dakota, where taxes were lower and were cut. Minnesota loses very badly when compared to South Dakota, which maintained its light tax burden throughout this period. Taken together, these results suggest a reasonably strong effect of state economic policy on economic outcomes, and in particular, they support the consistent research finding that high taxes have a significant adverse impact on economic growth.