Monday, August 6, 2018

The Emergence and Erosion of the Retail Sales Tax

About 160 countries around the world, including all the other high-income countries of the world, use a value-added tax. The US has no value added tax, but 45 states and several thousand cities, use a sales tax as an alternative method of taxing consumption.  John L. Mikesell and Sharon N. Kioko provide a useful overview of the issues in "The Retail Sales Tax in a New Economy," written for the 7th Annual Municipal Finance Conference, which was held on July 16-17, 2018, at the Brookings Institution.  Video of the conference presentation of the paper, with comments and discussion, is available here.

Here's a short summary of the emergence and erosion of the retail sales tax (footnotes omitted):
"The American retail sales tax emerged from a desperation experiment in Mississippi in the midst of the Great Depression. Revenue from the property tax, the largest single source of state tax revenue at the time, collapsed, falling by 11.4 percent from 1927 to 1932 and by another 16.8 percent from 1932 to 1934. State revenue could not cover their service obligations or provide expected assistance to local governments. Mississippi (followed by West Virginia) showed that retail sales taxes could produce immediate cash collections, even in low-income jurisdictions. Other states paid attention. In 1933, eleven other states adopted the tax (two let the tax expire almost immediately). By 1938, twenty-two states (plus Hawaii, not yet a state) were collecting the tax; six others had also imposed the tax for a short time but had let them expire. ...
"The national total retail sales tax collections exceeded the collections from every other state tax from 1947 through 2001. It was also the largest tax producer in 2003 and 2004 also (years in which individual income tax revenue was still impacted by the 2001 recession), but it was surpassed by state individual income tax revenues in other years since 2001. ,,, By fiscal 2016, total state individual income tax collections exceeded $345 billion, compared to over $288 billion for state retail sales taxes. However, those national totals conceal the continuing dominance of the retail sales taxes in a number of states ...
A major and ongoing US sales taxes is that, from the start, they mostly did not cover services. Thus, as the US has shifted to a service-based economy, the amount of consumer spending doing to goods covered by the sales tax has diminished. As the base of the sales tax diminished, then states have gradually raised the rate of the sale tax so that it would bring in a similar proportion of overall state revenue. This dynamic of higher sales tax rates imposed on a shrinking base is not sensible.
looking only at the 45 states with sales taxes.
"[Here is] the history of mean retail sales tax breadth (implicit tax base / state personal income) across the states from 1970 to 2016. The record is one of almost constant decline, from 49.0 percent in 1970 to 37.3 percent in 2016. ... The typical state retail sales tax base has narrowed as a share of the economy of the state over the years and this has meant that, in order for states to maintain the place of their sales tax in their revenue systems, they have been required to gradually increase the statutory tax rate they apply to that base. ... [L]ittle good can be said about a narrow base / high statutory rate revenue policy. ...
"Unfortunately, many states got off to a bad start when they initially adopted their sales taxes and excluded all or almost all household service purchases from the tax base and it has proven to be difficult to correct that initial error. Extending the retail sales tax to include at least some services is a perennial topic whenever states are seeking additional revenue or considering reforms in their tax systems. ... While the current typical sales tax base is around 20 percent narrower in 2016 compared with 1970, the base with all services added is actually about 11 percent broader, and the base without health care and education services is only 8 percent below its 1970 level. ..."
Another perennial sales tax issue is that legislatures like to list items that will be exempt from the sales tax, or tax "holidays" where sales tax doesn't need to be paid during certain time periods on items or like back-to-school items, energy-saving appliances, emergency preparedness supplies, and other items. These policies are often justified as helping those with low incomes, but any policy which cuts taxes for 100% of the population in the name of helping the 15-20% of the population that is poor has a mismatch between its stated intentions and its reality. Several states have taken a much more sensible course: if the goal is to help poor people, then give poor people a tax credit, based on their income, so that sales taxes they pay can be rebated to them. Mikesell and Kioko write:
"The problems with [a sales tax] exemption are well-known – absence of targeting, high revenue loss, additional cost of compliance and administration, distortion of consumer behavior, reward for political support, etc. – and it is particularly distressing in light of the fact that the credit/rebate system normally operated through the state income tax provides an alternative relief approach that eliminates virtually all these difficulties. Currently five states (Maine, Kansas, Oklahoma, Idaho, and Hawaii – operate some form of sales tax credit that returns to families some or all of sales tax paid on purchases, giving greatest relative relief to lowest income families and lesser (or no) relief to more affluent families. ... The credit / rebate system promises efficiency, equity, and less revenue loss. Its apparent unpopularity is somewhat surprising, particularly in light of the spread of the earned income tax credit program, a program with some similar characteristics."
Yet another perennial sales tax issue is that the logic of the tax is that it should apply to goods and services purchased by households, not to business purchases. If a sales tax is applied to business purchases, it raises a risk of "pyramiding," where one business pays sales tax on equipment and supplies from another business, and the consumers also pay sales tax on the finished product. If there are layers of businesses buying from each other along the supply chain, the sales tax can be imposed on a given product multiple times. Again, Mikesell and Kioko write:
"American retail sales taxes have not entirely gotten over the confusion that the tax is not on finished goods but rather should be on goods (and services) purchased for household consumption. The reality of sales taxation is that a considerable share of the overall sales tax base, roughly 40 percent on average, consists of input purchases by businesses. The tax on those purchases embeds in prices charged by those businesses, meaning that this share of the tax is effectively hidden from households, allowing legislatures to claim a statutory rate that is considerably less than the true rate borne by individuals. ... The pattern does show a considerable movement toward removal of these business input purchases from the tax base, thus reducing the prospects for pyramiding, hidden tax burdens, distortions, and discrimination. However, states continue to tax purchases made by other business activities. Lawmakers are inclined to try to pick favorites for tax relief and appear to like glitz. Targeted preferences for motion pictures, certain sorts of research and development, or bids for the Super Bowl are attractive to politicians because they provide identifiable credit and possibly ribbon cutting not available with general exemption. Super Bowl bids are particularly egregious."
A more recent issue is how jurisdictions with a sales tax can react to the rise of online sales from other jurisdictions. There seem to be several models developing. First, there is a "South Dakota" model of collecting sales tax from companies physically located in other states if they have total sales above a certain minimum level to South Dakota residents. The US Supreme Court upheld this law as constitutional this summer.  Other states that have adopted this model include Indiana, Iowa, North Dakota, Massachusetts, Maine, Mississippi, Wyoming, and Alabama.

An alternative "Colorado" model require sellers in a different state to notify both Colorado buyers and the Colorado tax authorities that state sales tax was due--but did not seek to collect the sales tax from those out-of-state sellers.  Other states that have enacted this approach are Louisiana, Pennsylvania, Vermont, and Washington.

Yet another approach addresses the question of when the producer is in one state, the buyer is in another state, and the "market facilitator" through which an online transaction is carried out is in still another state. This approach seeks to make the market facilitator based in one state responsible for collecting sales taxes on behalf of other states. Alabama, Arizona, Oklahoma, Pennsylvania, Rhode Island, Washington, and Minnesota have taken this approach.

The lurking difficulty with the lower base and higher rates for the retail sales tax is that, at some point, the tax rate gets  high enough that it becomes lucrative to find ways to avoid paying it.
"The problem is that there has been a consensus, heavily based on pre-value-added tax experience in Scandinavian countries with high-rate retail sales taxes, that retail sales tax rates much above 10 percent are likely to produce compliance issues so difficult that the tax becomes almost impossible to administer. American retail sales tax rates are drifting ever closer to that danger level, particularly when local governments add their own rates to the rate levied by the state. ... [S]tate statutory rates have drifted upward since 1970. Rates of 6 and 7 percent are no longer rare and a narrowing base will require more rate increases if the position of the sales tax is to be preserved (or expanded) in state revenue systems. Rates are moving toward the danger zone in which significant non-compliance begins to become more attractive and, unless states can manage the narrowing base problem, a compliance gap may become a significant challenge for state tax administrators in the first part of the 21st century."
Outside the US, where value-added taxes are high, there has been a spread of what is called "sales suppression software," under names like "phantomware" and "zappers." Basically, this software cooks the accounting books to make sales look lower, either by substituting lower prices for the higher price that was actually charged, or by reducing the number of transactions. This software takes care of other issues too, by producing fake inventory records if needed, or by running certain transactions through international cloud-based services that will be more difficult to track. Tax administration has become increasingly based on electronic records, so sales suppression software may turn into a real problem.