Tag Archives: growth

The use of locally-imposed selective taxes to fund public pension liabilities

Many eyes are on Kentucky policymakers as they grapple with finding a solution to their $40 billion state-reported unfunded public pension liability. As talks of a potential pension bill surface, various proposals have been made by legislators, but very few have gained traction. One such proposal stands out from the rest. A proposal that has since been shut down suggested imposing selective taxes on tobacco, prescription opiates, and outsourced labor to generate revenue to direct towards paying down the state’s pension debt. Despite its short-lived tenure, this selective tax proposal reflects a recent trend in pension funding reform; a trend that policymakers should be wary of. Implementing new taxes on select goods or services may seem like a good idea as it could, in theory, potentially raise additional revenues, but experience at the local level suggests otherwise.

In chapter 12 of a new Mercatus book on sin taxes, NYU professor Thad Calabrese examines the practice of locally-imposed selective taxes that are used to fund public pension liabilities and doesn’t find much evidence to support their continued usage.

Selective taxes are sales taxes that target specific goods and are also known as ‘sin taxes’ because of their popular usage in taxing less healthy goods such as cigarettes, junk food, or alcohol. In the examples that Calabrese examines, selective taxes are used to target insurance premiums as revenue sources for pensions.

Only a select few states have begun this practice – including Illinois, Pennsylvania, as well as municipalities in West Virginia and Missouri – but it may become more popular if courts begin to restrict the way in which current pension benefits can be modified. Once benefits are taken off the table as an avenue for reform, like in Illinois, policymakers will feel more pressure to find new revenue sources.

The proposal in Kentucky may seem appealing to policymakers, especially because of its potential to raise $600 million a year, but this estimate overlooks the unintended effects that such new taxes could facilitate. Thankfully, the proposal did not go through, but I think some time should be spent looking at what similar proposals have looked like at the local level, so that other states do not get tempted pick up where Kentucky left off.

Calabrese draws on the experiences in Pennsylvania and Illinois to examine how these taxes have operated, how the decoupling of setting and financing employee benefits tends to lead to these taxes, and how the use of these taxes is associated with significantly underfunded pension systems. Below I highlight Pennsylvania’s experience and caution against further usage of this mechanism for pension funding.

How it works (or doesn’t)

In 1895, Pennsylvania implemented a 2 percent tax on out-of-state fire and casualty insurance companies’ premiums on in-state property and then earmarked this for distribution to local governments to pay for pensions. Act 205 of 1984 replaced the original act in which the state of Pennsylvania allocated pension aid based on where the insured property was located and instead the new allocation was based on the number of public employees in a locality.

Calabrese explains how the funds were distributed:

“Each public employee was considered a ‘unit,’ and uniformed employees (such as police and fire) each represented two units. The pool of insurance tax revenue collected by the state was then divided by the sum of municipal units to arrive at a unit value. This distribution could subsidize local governments’ pension expenditures up to 100 percent of the annual cost. In 1985, this tax generated $62.3 million in revenues; as a result, each unit value was worth $1,146 – meaning that local governments received $1,146 for pension funding for each public employee and an additional $1,146 for pension funding for each uniformed public employee. Importantly, 75 percent of municipalities received enough funding from this revenue in 1985 to fully offset their pension costs.”

The new mechanism raised more funds, but it also unexpectedly raised costs. If a municipality had to contribute less than the $1,146 annually for a regular employee or $2,292 for a uniformed employee, for example, the municipality was essentially incentivized to increase benefits to public employees up to this limit, because local public employees would receive increased benefits at no direct budgetary cost to the municipality.

“…the tax likely increased insurance costs for residents and businesses (and then only a small fraction of the cost), but not directly for the government employer. Further, this system privileged benefits relative to other compensation, because these payments (borne at least statutorily by out-of-state companies) could only be used for financing pensions and not other forms of compensation.”

A tax originally implemented to fund pension costs statewide resulted in a system that encouraged more generous benefits.

Despite increased subsidies from the state, only 38 percent of municipalities received sufficient allocated funds from the pool to fully offset the costs of pensions. This was because annual pension contributions were growing at a faster rate than the rate at which the subsidy from the state insurance tax was growing.

To highlight a city with severely distressed pension plans, Philadelphia continued to struggle even following the implementation of the state insurance tax. The police pension plan, nonuniformed plan, and firefighter pension plan were all only 49, 47, and 45 percent funded, respectively. In 2009, the City Council passed a temporary 1 percentage point increase in their sales tax and when the temporary rate was renewed in 2014, any revenue in excess of $120 million was dedicated to the city’s pension plans. Additionally, the state permitted the city to pass a $2 per pack cigarette tax to fund a planned budget deficit for the school system; likely because its income tax capacity was largely exhausted.

Philadelphia’s new taxes technically generated new revenues, but they did little to improve the funding of the city’s pension plans.

The selective taxes implemented to fund pension liabilities in Pennsylvania were effectively a Band-Aid that was two small for the state’s pension funding problem, which in turn required the addition of more, insufficient pension Band-Aids. It merely created a public financing system that encouraged pension benefit growth which led to the passage of additional laws requiring certain pension funding levels. And when these funding levels were not met, even more laws were passed that provided temporary pension funding relief, which further grew liabilities for distressed municipalities.

Act 44 became law in 1993 and provided plan sponsors pension funding relief, but primarily by allowing sponsors to alter actuarial assumptions and thereby reduce required pension contributions. Another law delayed funding by manipulating how the required contribution was calculated, rather than providing any permanent fix.

Moving forward

Selective taxes for the purpose of funding pensions are still a relatively rare practice, but as pension liabilities grow and the landscape of reform options changes, it may become increasingly attractive to policymakers. As Calabrese has demonstrated in his book chapter, however, we should be wary of this avenue as it may only encourage the growth of pension liabilities without addressing the problem in any meaningful way. Reforming the structure of the pension plan or the level of benefits provided to current or future employees would provide the most long-term solution.

A solution with the long-term in mind and that doesn’t involve touching current beneficiaries includes moving future workers to defined contribution plans; plans that are better suited to keeping costs contained. The ballooning costs aren’t stemming solely from overly generous plan benefits, but more seriously are the result of their poor management and incentives for funding, only exacerbated by poor accounting practices. The problem is certainly complicated and moving towards the use of defined contribution plans wouldn’t eliminate all issues, but it would at least set governments on a more sustainable path.

At the very least, policymakers interested in long-term solutions should be cautioned against using selective taxes to fund pensions.

State tax refunds and limiting spending growth

This fall eligible Alaskans will be receiving a check of $1,100 from their state government. Although the amount of the check can vary, Alaskans receive one every fall – no strings attached. Other state residents are probably more familiar with IRS tax refunds that come every spring, but this “tax refund” that Alaskans receive is unique. It’s a feature that residents have benefited from for decades, even in times when the government has experienced fiscal stress. Considering the state’s unique and distressed budget situation that I’ve described in an earlier post, I think it warrants a discussion of the fiscal viability of their refunds.

A narrow tax base reliant on volatile revenue sources, restricted funds, and growing spending are all factors that made closing Alaska’s budget gap this year very difficult. It even contributed to pulling down Alaska from 1st in our 2016 ranking of states by fiscal condition to 17th in our 2017 edition. Given this deterioration, it will be helpful to look into how and why Alaska residents receive dividend payments each year. There is no public finance rule that says giving refunds to residents is fiscally irresponsible, but there definitely are better ways to do it, and Alaska certainly hasn’t proven to display best practices.

Another state that we can look at for comparison is Colorado, which has a similar “tax refund” for residents but is structured very differently. Colorado’s Taxpayer Bill of Rights (TABOR) requires that higher than expected tax revenues each year be refunded to taxpayers and acts as a restraint on government spending growth. In contrast, Alaska’s check comes from the state’s Permanent Fund’s earnings that are generated from oil severance taxes each year, and acts more like a dividend from oil investment earnings.

Are distributing these refunds to taxpayers fiscally responsible? I am going to take a deeper look at these mechanisms to find out.

First, Alaska’s refund.

The figure below displays Alaska’s Permanent Fund checks since 2002 overlaid with the state’s revenue and expenditure trends, all adjusted for inflation. The highest check (in 2015 dollars) was $2,279 in 2008 and the lowest was $906 in 2012, with the average over this time period being about $1,497 per person. Although the check amounts do vary, Alaska has kept on top of delivering them, even in times of steep budget gaps like in 2002, 2009, and 2015. The Permanent Fund dividend formula is based on net income from the current plus the previous four fiscal years, so it makes sense that the check sizes are also cyclical in nature, albeit in a slightly delayed fashion behind oil revenue fluctuations.

Alaska’s dividend payments often end up on the chopping block during yearly budget debates, and there is growing pressure to at least have them reduced. Despite this, Alaska’s dividends are very popular with residents (who can blame them?) and probably won’t be going away for a long time; bringing a new meaning to the Permanent Fund’s name.

The Alaska Permanent Fund was established in 1976 by constitutional amendment and was seen as an investment in future generations, who might no longer have access to oil as a resource. Although this may have been decent forward-thinking, which is rare in state budgets, it does illustrate an interesting public finance story.

Alaska is a great example of a somewhat backwards situation. They generate high amounts of cash each year, but because of the way many of their funds are restricted they are forced to hoard much of it, and give the rest to citizens in the form of dividends. If a different state were to consider a similar dividend before dealing with serious structural budget flaws would be akin to putting the cart before the horse.

Luckily for Alaskan dividend recipients, there are many other areas that the state could reform first in order to improve their budget situation while avoiding cutting payments. As my colleague Adam Millsap has recommended, a fruitful area is tax reform. Alaska doesn’t have an income or sales tax; two of the most common sources of revenue for state governments. These are two potentially more stable sources of income than what the state currently has.

How does Colorado’s “tax refund” compare?

Colorado’s Taxpayer Bill of Rights (TABOR) has a feature that requires any tax revenue growth beyond inflation and population growth be refunded to taxpayers. It was adopted by Colorado voters in 1992 and it essentially restricts revenues by prohibiting any tax or spending increases without voter approval.

A recent example of this playing out was in 2014 when the state realized higher than expected tax revenues as a result of marijuana legalization. At the point of legalization, the plan was to direct tax revenues generated from the sale of marijuana towards schools or substance abuse program funding. But because of the higher than expected revenues, TABOR was triggered and it would require voter approval to decide if the excess revenues would be sent back to taxpayers or directed to other state programs.

In November of 2015, Colorado voters approved a statewide ballot measure that gave state lawmakers permission to spend $66.1 million in taxes collected from the sale of marijuana. The first $40 million was sent to school construction, the next $12 million to youth and substance abuse programs, and the remainder $14.2 billion to discretionary spending programs. A great example that although TABOR does generally restrain spending, citizens still have power to decline refunds in the name of program spending they are passionate about.

 

The second figure here displays TABOR refunds compared with state revenues and expenditures over time. Adjusted for inflation, checks have varied from $18 in 2005 to $351 in 1999, much smaller than the Alaska dividend checks. TABOR checks have only tended to be distributed when revenues have exceeded expenses. The main reason why checks weren’t distributed between 2006 and 2009, despite a revenue surplus, was because of Referendum C which removed TABOR’s revenue limit for five years, allowing the state to keep collections exceeding the rule. The revenue limit has since been reinstated, but some question the effectiveness of TABOR given an earlier amendment in 2000 which exempts much of education spending from TABOR restrictions.

The main distinguishing factor between Colorado’s refund and Alaska’s Permanent Fund dividend is that the former also acts as a constraint on spending growth. By requiring the legislature to get voter approval before any tax increase or spending of new money, it implements automatic checks on these activities. Many states attempt to do this through what are called “Tax and Expenditure Limits” or TELs.

The worry is that left unchecked, state spending can grow to unsustainable levels.

Tax and Expenditure Limits

A review of the literature up to 2012 found that although the earliest studies were largely skeptical of the effectiveness of TELs, as time has passed more research points to the contrary. TELs can restrain spending, but only in certain circumstances.

My colleague Matt Mitchell found in 2010 that TELs are more effective when they (1) bind spending rather than revenue, (2) require a super-majority rather than a simple majority vote to be overridden, (3) immediately refund revenue collected in excess of the limit, and (4) prohibit unfunded mandates on local government.

Applying these criteria to Colorado’s TABOR we see that it does well in some areas and could improve in others. TABOR’s biggest strength is that it immediately refunds revenue collected in excess of the limit in its formula, pending voter approval to do otherwise. Automatically refunding surpluses makes it difficult for governments to use excess funds irresponsibly and also gives taxpayers an incentive to support TABOR.

Colorado’s TABOR does well to limit revenue growth according to a formula, rather than to a fixed number or no limitation at all. The formula partially meets Mitchell’s standards. It stands up well with the most stringent TELs by limiting government growth that exceeds inflation and population growth, but could actually be improved if it limited actual spending growth rather than focusing on tax revenue. When a TEL or similar law limits revenues, policymakers can respond by resorting to implementing more fees or borrowing. There’s some evidence of this occurring in Colorado, with fees becoming more popular as a way to raise revenue since TABOR’s passing. A spending-based TEL is more difficult to evade.

Despite its faults, Colorado’s TABOR structure appears to be doing better than attempts to constrain spending growth in other states. The National Conference of State Legislatures still considers it one of the strictest TELs in the nation. Other states, like Arkansas, could learn a lot from Colorado. A recent Mercatus study analyzes Arkansas’ Revenue Stabilization Law and suggests that it is missing a component similar to Colorado’s TABOR formula to refund excess revenues.

How much a state spends is ultimately up to its residents and legislature. Some states may have a preference for more spending than others, but given the tendency for government spending to grow towards an unsustainable direction, having a conversation about how to slow this is key. Implementing TEL-like checks allows for spending to be monitored and that tax dollars be spent more strategically.

Alaska’s Permanent Fund dividend is not structured as well as Colorado’s, but perhaps the state’s saving grace is that it has a relatively well structured TEL. Similarly to Colorado’s TABOR, Alaska’s TEL limits budget growth to the sum of inflation and population growth and is codified in the constitution. Alaska’s TEL doesn’t immediately refund revenue that is collected in excess of the limit to taxpayers as Colorado’s TABOR does, but it does target spending rather than revenues.

Colorado’s and Alaska’s TELs can compete when it comes to restraining spending, but Colorado’s is certainly more strict. Colorado’s expenditures have grown by about 55 percent over the last decade, while Alaska’s has grown approximately 120 percent.

The Lesson

Comparing Colorado and Alaska’s situations reveals two different ways of giving tax refunds to residents. Doing so doesn’t necessarily have to be fiscally irresponsible. Colorado has provided refunds to residents when state revenues have exceeded expenses and as a result this has acted as a restraint on over-spending higher than expected revenues. Although Colorado’s TABOR has been amended over time, its general structure illustrates the effectiveness of institutional restrains on spending. The unintended effects of TABOR, such as the increase in fees, could be well addressed by specifically targeting spending rather revenue, like in the case of Alaska’s TEL. Alaska may have had their future residents’ best intent in mind when they designed their Permanent Fund Dividend, but perhaps this goal of passing forward oil investment earnings should have been paired with preparing for the potential of cyclical budget woes.

Government Spending and Economic Growth in Nebraska since 1997

Mercatus recently released a study that examines Nebraska’s budget, budgetary rules and economy. As the study points out, Nebraska, like many other states, consistently faces budgeting problems. State officials are confronted by a variety of competing interests looking for more state funding—schools, health services and public pensions to name a few—and attempts to placate each of them often leave officials scrambling to avoid budget shortfalls in the short term.

Money spent by state and local governments is collected from taxpayers who earn money in the labor market and through investments. The money earned by taxpayers is the result of producing goods and services that people want and the total is essentially captured in a state’s Gross Domestic Product (GSP).

State GSP is a good measure of the amount of money available for a state to tax, and if state and local government spending is growing faster than GSP, state and local governments will be controlling a larger and larger portion of their state’s output over time. This is unsustainable in the long run, and in the short run more state and local government spending can reduce the dynamism of a state’s economy as resources are taken from risk-taking entrepreneurs in the private sector and given to government bureaucrats.

The charts below use data from the BEA to depict the growth of state and local government spending and private industry GSP in Nebraska (click on charts to enlarge). The first shows the annual growth rates in private industry GSP and state and local government GSP from 1997 to 2014. The data is adjusted for inflation (2009 dollars) and the year depicted is the ending year (e.g. 1998 is growth from 1997 – 1998).

NE GSP annual growth rates 1997-14

In Nebraska, real private industry GSP growth has been positive every year except for 2012. There is some volatility consistent with the business cycles over this time period, but Nebraska’s economy has regularly grown over this period.

On the other hand, state and local GSP growth was negative 10 of the 17 years depicted. It grew rapidly during recession periods (2000 – 2002 and 2009 – 2010), but it appears that state and local officials were somewhat successful in reducing spending once economic conditions improved.

The next chart shows how much private industry and state and local GSP grew over the entire period for both Nebraska and the U.S. as a whole. The 1997 value of each category is used as the base year and the yearly ratio is plotted in the figure. The data is adjusted for inflation (2009 dollars).

NE, US GSP growth since 1997

In 2014, Nebraska’s private industry GSP (red line) was nearly 1.6 times larger than its value in 1997. On the other hand, state and local spending (light red line) was only about 1.1 times larger. Nebraska’s private industry GSP grew more than the country’s as a whole over this period (57% vs 46%) while its state and local government spending grew less (11% vs. 15%).

State and local government spending in Nebraska spiked from 2009 to 2010 but has come down slightly since then. Meanwhile, the state’s private sector has experienced relatively strong growth since 2009 compared to the country as a whole, though it was lagging the country prior to the recession.

Compared to the country overall, Nebraska’s private sector economy has been doing well since 2008 and state and local spending, while growing, appears to be largely under control. If you would like to learn more about Nebraska’s economy and the policies responsible for the information presented here, I encourage you to read Governing Nebraska’s Fiscal Commons: Addressing the Budgetary Squeeze, by Creighton University Professor Michael Thomas.

Why the lack of labor mobility in the U.S. is a problem and how we can fix it

Many researchers have found evidence that mobility in the U.S. is declining. More specifically, it doesn’t appear that people move from places with weaker economies to places with stronger economies as consistently as they did in the past. Two sets of figures from a paper by Peter Ganong and Daniel Shoag succinctly show this decline over time.

The first, shown below, has log income per capita by state on the x-axis for two different years, 1940 (left) and 1990 (right). On the vertical axis of each graph is the annual population growth rate by state for two periods, 1940 – 1960 (left) and 1990 – 2010 (right).

directed migration ganong, shoag

In the 1940 – 1960 period, the graph depicts a strong positive relationship: States with higher per capita incomes in 1940 experienced more population growth over the next 20 years than states with lower per capita incomes. This relationship disappears and actually reverses in the 1990 – 2010 period: States with higher per capita incomes actually grew slower on average. So in general people became less likely to move to states with higher incomes between the middle and end of the 20th century. Other researchers have also found that people are not moving to areas with better economies.

This had an effect on income convergence, as shown in the next set of figures. In the 1940 – 1960 period (left), states with higher per capita incomes experienced less income growth than states with lower per capita incomes, as shown by the negative relationship. This negative relationship existed in the 1990 – 2010 period as well, but it was much weaker.

income convergence ganong, shoag

We would expect income convergence when workers leave low income states for high income states, since that increases the labor supply in high-income states and pushes down wages. Meanwhile, the labor supply decreases in low-income states which increases wages. Overall, this leads to per capita incomes converging across states.

Why labor mobility matters

As law professor David Schleicher points out in a recent paper, the current lack of labor mobility can reduce the ability of the federal government to manage the U.S. economy. In the U.S. we have a common currency—every state uses the U.S. dollar. This means that if a state is hit by an economic shock, e.g. low energy prices harm Texas, Alaska and North Dakota but help other states, that state’s currency cannot adjust to cushion the blow.

For example, if the UK goes into a recession, the Bank of England can print more money so that the pound will depreciate relative to other currencies, making goods produced in the UK relatively cheap. This will decrease the UK’s imports and increase economic activity and exports, which will help it emerge from the recession. If the U.S. as a whole suffered a negative economic shock, a similar process would take place.

However, within a country this adjustment mechanism is unavailable: Texas can’t devalue its dollar relative to Ohio’s dollar. There is no within-country monetary policy that can help particular states or regions. Instead, the movement of capital and labor from weak areas to strong areas is the primary mechanism available for restoring full employment within the U.S. If capital and labor mobility are low it will take longer for the U.S. to recover from area-specific negative economic shocks.

State or area-specific economic shocks are more likely in large countries like the U.S. that have very diverse local economies. This makes labor and capital mobility more important in the U.S. than in smaller, less economically diverse countries such as Denmark or Switzerland, since those countries are less susceptible to area-specific economic shocks.

Why labor mobility is low

There is some consensus about policies that can increase labor mobility. Many people, including former President Barack Obama, my colleagues at the Mercatus Center and others, have pointed out that state occupational licensing makes it harder for workers in licensed professions to move across state borders. There is similar agreement that land-use regulations increase housing prices which makes it harder for people to move to areas with the strongest economies.

Reducing occupational licensing and land-use regulations would increase labor mobility, but actually doing these things is not easy. Occupational licensing and land-use regulations are controlled at the state and local level, so currently there is little that the federal government can do.

Moreover, as Mr. Schleicher points out in his paper, state and local governments created these regulations for a reason and it’s not clear that they have any incentive to change them. Like all politicians, state and local ones care about being re-elected and that means, at least to some extent, listening to their constituents. These residents usually value stability, so politicians who advocate too strongly for growth may find themselves out of office. Mr. Schleicher also notes that incumbent politicians often prefer a stable, immobile electorate because it means that the voters who elected them in the first place will be there next election cycle.

Occupational licensing and land-use regulations make it harder for people to enter thriving local economies, but other policies make it harder to leave areas with poor economies. Nearly 13% of Americans work for state and local governments and 92% of them have a defined-benefit pension plan. Defined-benefit plans have long vesting periods and benefits can be significantly smaller if employees split their career between multiple employers rather than remain at one employer. Thus over 10% of the workforce has a strong retirement-based incentive to stay where they are.

Eligibility standards for public benefits and their amounts also vary by state, and this discourages people who receive benefits such as Temporary Assistance for Needy Families (TANF) from moving to states that may have a stronger economy but less benefits. Even when eligibility standards and benefits are similar, the paperwork and time burden of enrolling in a new state can discourage mobility.

The federal government subsidizes home ownership as well, and homeownership is correlated with less labor mobility over time. Place-based subsidies to declining cities also artificially support areas that should have less people. As long as state and federal governments subsidize government services in cities like Atlantic City and Detroit people will be less inclined to leave them. People-based subsidies that incentivize people to move to thriving areas are an alternative that is likely better for the taxpayer, the recipient and the country in the long run.

How to increase labor mobility

Since state and local governments are unlikely to directly address the impediments to labor mobility that they have created, Mr. Schleicher argues for more federal involvement. Some of his suggestions don’t interfere with local control, such as a federal clearinghouse for coordinated occupational-licensing rules across states. This is not a bad idea but I am not sure how effective it would be.

Other suggestions are more intrusive and range from complete federal preemption of state and local rules to federal grants that encourage more housing construction or suspension of the mortgage-interest deduction in places that restrict housing construction.

Local control is important due to the presence of local knowledge and the beneficial effects that arise from interjurisdictional competition, so I don’t support complete federal preemption of local rules. Economist William Fischel also thinks the mortgage interest deduction is largely responsible for excessive local land-use regulation, so eliminating it altogether or suspending it in places that don’t allow enough new housing seems like a good idea.

I also support more people-based subsidies that incentivize moving to areas with better economies and less place-based subsidies. These subsidies could target people living in specific places and the amounts could be based on the economic characteristics of the destination, with larger amounts given to people who are willing to move to areas with the most employment opportunities and/or highest wages.

Making it easier for people to retain any state-based government benefits across state lines would also help improve labor mobility. I support reforms that reduce the paperwork and time requirements for transferring benefits or for simply understanding what steps need to be taken to do so.

Several policy changes will need to occur before we can expect to see significant changes in labor mobility. There is broad agreement around some of them, such as occupational licensing and land-use regulation reform, but bringing them to fruition will take time. As for the less popular ideas, it will be interesting to see which, if any, are tried.

Today’s public policies exacerbate our differences

The evidence that land-use regulations harm potential migrants keeps piling up. A recent paper in the Journal of Urban Economics finds that young workers (age 22 – 26) of average ability who enter the labor force in a large city (metropolitan areas with a population > 1.5 million) earn a wage premium equal 22.9% after 5 years.

The author also finds that high-ability workers experience additional wage growth in large cities but not in small cities or rural areas. This leads to high-ability workers sorting themselves into large cities and contributes an additional 3.2% to the urban wage-growth premium.

These findings are consistent with several other papers that have analyzed the urban wage premium. Potential causes of the wage premium are faster human capital accumulation in denser, more populated places due to knowledge spillovers and more efficient labor markets that better match employers and employees.

The high cost of housing in San Francisco, D.C., New York and dozens of other cities is preventing many young people from earning more money and improving their lives. City officials and residents need to strike a better balance between maintaining the “charm” of their neighborhoods and affordability. This means less regulation and more building.

City vs. rural is only one of the many dichotomies pundits have been discussing since the 2016 election. Some of the other versions of “two Americas” are educated vs. non-educated, white collar vs. blue collar, and rich vs. poor. We can debate how much these differences matter, but to the extent that they are an issue for the country our public policies have reinforced the barriers that allow them to persist.

Occupational licensing makes it more difficult for blue-collar manufacturing workers to transition to middle-class service sector jobs. Federal loan subsidies have made four-year colleges artificially cheap to the detriment of people with only a high school education. Restrictive zoning has made it too expensive for many people to move to places with the best labor markets. And once you’re in a city, unless you’re in one of the best neighborhoods your fellow citizens often keep employers and providers of much needed consumer staples like Wal-Mart out, while using eminent domain to build their next playground.

Over time people have sorted themselves into different groups and then erected barriers to keep others out. Communities do it with land-use regulations, occupations do it with licensing and established firms do it with regulatory capture. If we want a more prosperous America that de-emphasizes our differences and provides people of all backgrounds with opportunity we need more “live and let live” and less “my way or the highway”.

Does Tax Increment Financing (TIF) generate economic development?

Tax increment financing, or TIF, is a method of financing economic development projects first used in California in 1952. Since then, 48 other states have enacted TIF legislation with Arizona being the lone holdout. It was originally conceived as a method for combating urban blight, but over time it has become the go-to tool for local politicians pushing economic development in general. For example, Baltimore is considering using TIF to raise $535 million to help Under Armor founder Kevin Plank develop Port Covington.

So how does TIF work? Though the particulars can vary by state, the basic mechanism is usually similar. First, an area is designated as a TIF district. TIF districts are mostly industrial or commercial areas rather than residential areas since the goal is to encourage economic development.

Usually, in an effort to ensure that TIF is used appropriately, the municipal government that designates the area as a TIF has to assert that economic development would not take place absent the TIF designation and subsequent investment. This is known as the ‘but-for’ test, since the argument is that development would not occur but for the TIF. Though the ‘but-for’ test is still applied, some argue that it is largely pro forma.

Once an area has been designated as a TIF district, the property values in the area are assessed in order to create a baseline value. The current property tax rate is applied to the baseline assessed value to determine the amount of revenue that is used for the provision of local government goods and services (roads, police, fire, water etc.). This value will then be frozen for a set period of time (e.g. up to 30 years in North Carolina), and any increase in assessed property values that occurs after this time and the subsequent revenue generated will be used to pay for the economic development project(s) in the TIF district.

The key idea is that municipalities can borrow against the projected property value increases in order to pay for current economic development projects. A simple numerical example will help clarify how TIF works.

In the table below there are five years. In year 1 the assessed value of the property in the TIF district is $20 million and it is determined that it takes $1 million per year to provide the government goods and services needed in the area (road maintenance, sewage lines, police/fire protection, etc.). A tax rate of 5% is applied to the $20 million of assessed value to raise the necessary $1 million (Tax revenue column).

TIF example table

The municipality issues bonds totaling $1 million to invest in an economic development project in the TIF district. As an example, let’s say the project is renovating an old business park in order to make it more attractive to 21st century startups. The plan is that improving the business park will make the area more desirable and increase the property values in the TIF district. As the assessed value increases the extra tax revenue raised by applying the 5% rate to the incremental value of the property will be used to pay off the bonds (incremental revenue column).

Meanwhile, the $1 million required for providing the government goods and services will remain intact, since only the incremental increase in assessed value is used to pay for the business park improvements. Hence the term Tax Increment Financing.

As shown in the table, if the assessed value of the property increases by $2 million per year for 4 years the municipality will recoup the $1 million required to amortize the bond (I’m omitting interest to keep it simple). Each $1 million dollars of increased value increase tax revenue by $50,000 without increasing the tax rate, which is what allows the municipality to pay for the economic development without raising property tax rates. For many city officials this is an attractive feature since property owners usually don’t like tax rate increases.

City officials may also prefer TIF to the issuance of general obligation bonds since the latter often require voter approval while TIF does not. This is the case in North Carolina. TIF supporters claim that this gives city officials more flexibility in dealing with the particular needs of development projects. However, it also allows influential individuals to push TIF through for projects that a majority of voters may not support.

While TIF can be used for traditional government goods like roads, sewer systems, water systems, and public transportation, it can also be used for private goods like business parks and sports facilities. The former arguably provide direct benefits to all firms in the TIF district since better roads, streetscapes and water systems can be used by any firm in the area. The latter projects, though they may provide indirect benefits to nearby firms in the form of more attractive surroundings and increased property values, mostly benefit the owners of entity receiving the development funding. Like other development incentives, TIF can be used to subsidize private businesses with taxpayer dollars.

Projects that use TIF are often described as ‘self-financing’ since the project itself is supposedly what creates the higher property values that pay for it. Additionally, TIF is often sold to voters as a way to create jobs or spur additional private investment in blighted areas. But there is no guarantee that the development project will lead to increased private sector investment, more jobs or higher property values. Researchers at the UNC School of Government explain the risks of TIF in a 2008 Economic Bulletin:

“Tax increment financing is not a silver bullet solution to development problems. There is no guarantee that the initial public investment will spur sufficient private investment, over time, that creates enough increment to pay back the bonds. Moreover, even if the investment succeeds on paper, it may do so by “capturing” growth that would have occurred even without the investment. Successful TIF districts can place an additional strain on existing public resources like schools and parks, whose funding is frozen at base valuation levels while growth in the district increases demand for their services.”

The researchers also note that it’s often larger corporations that municipalities are trying to attract with TIF dollars, and any subsidies via TIF that the municipality provides to the larger firm gives it an advantage over its already-established, local competitors. This is even more unfair when the local competitor is a small, mom-and-pop business that already faces a difficult challenge due to economies of scale.

There is also little evidence that TIF regularly provides the job or private sector investment that its supporters promise. Chicago is one of the largest users of TIF for economic development and its program has been one of the most widely studied. Research on Chicago’s TIF program found that “Overall, TIF failed to produce the promise of jobs, business development or real estate activity at the neighborhood level beyond what would have occurred without TIF.”

If economic development projects that rely on TIF do not generate additional development above and beyond what would have occurred anyway, then the additional tax revenue due to the higher assessed values is used to pay for an economic development project that didn’t really add anything. Without TIF, that revenue could have been used for providing other government goods and services such as infrastructure or better police and fire protection. Once TIF is used, the additional revenue must be used to pay for the economic development project: it cannot be spent on other services that residents might prefer.

Another study, also looking at the Chicago metro area, found that cities that adopt TIF experience slower property value growth than those that do not. The authors suggest that this is due to a reallocation of resources to TIF districts from other areas of the city. The result is that the TIF districts grow at the expense of the municipality as a whole. This is an example of the TIF working on paper, but only because it is pilfering growth that would have occurred in other areas of the city.

Local politicians often like tax increment financing because it is relatively flexible and enables them to be entrepreneurial in some sense: local officials as venture capitalists. It’s also an easier sell than a tax rate increase or general obligation bonds that require a voter referendum.

But politicians tend to make bad venture capitalists for several reasons. First, it’s usually not their area of expertise and it’s hard: even the professionals occasionally lose money. Second, as Milton Friedman pointed out, people tend to be more careless when spending other people’s money. Local officials aren’t investing their own money in these projects, and when people invest or spend other people’s money they tend to emphasize the positive outcomes and downplay the negative ones since they aren’t directly affected. Third, pecuniary factors don’t always drive the decision. Different politicians like different industries and businesses – green energy, biotech, advanced manufacturing, etc. – for various reasons and their subjective, non-pecuniary preferences may cause them to ignore the underlying financials of a project and support a bad investment.

If TIF is going to be used it should be used on things like public infrastructure – roads, sewer/water lines, sidewalks – rather than specific private businesses. This makes it harder to get distracted by non-pecuniary factors and does a better – though not perfect – job of directly helping development in general rather than a specific company or private developer. But taxpayers should be aware of the dangers of TIF and politicians and developers should not tout it as a panacea for jump-starting an area’s economy.

Northern Cities Need To Be Bold If They Want To Grow

Geography and climate have played a significant role in U.S. population growth since 1970 (see here, here, here, and here). The figure below shows the correlation between county-level natural amenities and county population growth from 1970 – 2013 controlling for other factors including the population of the county in 1970, the average wage of the county in 1970 (a measure of labor productivity), the proportion of adults in the county with a bachelor’s degree or higher in 1970 and region of the country. The county-level natural amenities index is from the U.S. Department of Agriculture and scores the counties in the continental U.S. according to their climate and geographic features. The county with the worst score is Red Lake, MN and the county with the best score is Ventura, CA.

1970-13 pop growth, amenities

As shown in the figure the slope of the best fit line is positive. The coefficient from the regression is also given at the bottom of the figure and is equal to 0.16, meaning a one point increase in the score increased population growth by 16 percentage points on average.

The effect of natural amenities on population growth is much larger than the effect of the proportion of adults with a bachelor’s degree or higher, which is another strong predictor of population growth at the metropolitan (MSA) and city level (see here, here, here, and here). The relationship between county population growth from 1970 – 2013 and human capital is depicted below.

1970-13 pop growth, bachelors or more

Again, the relationship is positive but the effect is smaller. The coefficient is 0.026 which means a 1 percentage point increase in the proportion of adults with a bachelor’s degree or higher in 1970 increased population growth by 2.6 percentage points on average.

An example using some specific counties can help us see the difference between the climate and education effects. In the table below the county where I grew up, Greene County, OH, is the baseline county. I also include five other urban counties from around the country: Charleston County, SC; Dallas County, TX; Eau Claire County, WI; San Diego County, CA; and Sedgwick County, TX.

1970-13 pop chg, amenities table

The first column lists the amenities score for each county. The highest score belongs to San Diego. The second column lists the difference between Green County’s score and the other counties, e.g. 9.78 – (-1.97) = 11.75 which is the difference between Greene County’s score and San Diego’s score. The third column is the difference column multiplied by the 0.16 coefficient from the natural amenity figure e.g. 11.75 x 0.16 = 188% in the San Diego row. What this means is that according to this model, if Greene County had San Diego’s climate and geography it would have grown by an additional 188 percentage points from 1970 – 2013 all else equal.

Finally, the last column is the actual population growth of the county from 1970 – 2013. As shown, San Diego County grew by 135% while Greene County only grew by 30% over this 43 year period. Improving Greene County’s climate to that of any of the other counties except for Eau Claire would have increased its population growth by a substantial yet realistic amount.

Table 2 below is similar to the natural amenities table above only it shows the different effects on Greene County’s population growth due to a change in the proportion of adults with a bachelor’s degree or higher.

1970-13 pop chg, bachelor's table

As shown in the first column, Greene County actually had the largest proportion of adults with bachelor’s degree or higher in 1970 – 14.7% – of the counties listed.

The third column shows how Greene County’s population growth would have changed if it had the same proportion of adults with a bachelor’s degree or higher as the other counties did in 1970. If Greene County had the proportion of Charleston (11.2%) instead of 14.7% in 1970, its population growth is predicted to have been 9 percentage points lower from 1970 – 2013, all else equal. All of the effects in the table are negative since all of the counties had a lower proportion than Greene and population education has a positive effect on population growth.

Several studies have demonstrated the positive impact of an educated population on overall city population growth – often through its impact on entrepreneurial activity – but as shown here the education effect tends to be swamped by geographic and climate features. What this means is that city officials in less desirable areas need to be bold in order to compensate for the poor geography and climate that are out of their control.

A highly educated population combined with a business environment that fosters innovation can create the conditions for city growth. Burdensome land-use regulations, lengthy, confusing permitting processes, and unpredictable rules coupled with inconsistent enforcement increase the costs of doing business and stifle entrepreneurship. When these harmful business-climate factors are coupled with a generally bad climate the result is something like Cleveland, OH.

The reality is that the tax and regulatory environments of declining manufacturing cities remain too similar to those of cities in the Sunbelt while their weather and geography differ dramatically, and not in a good way. Since only relative differences cause people and firms to relocate, the similarity across tax and regulatory environments ensures that weather and climate remain the primary drivers of population change.

To overcome the persistent disadvantage of geography and climate officials in cold-weather cities need to be aggressive in implementing reforms. Fiddling around the edges of tax and regulatory policy in a half-hearted attempt to attract educated people, entrepreneurs and large, high-skill employers is a waste of time and residents’ resources – Florida’s cities have nicer weather and they’re in a state with no income tax. Northern cities like Flint, Cleveland, and Milwaukee that simply match the tax and regulatory environment of Houston, San Diego, or Tampa have done nothing to differentiate themselves along those dimensions and still have far worse weather.

Location choices reveal that people are willing to put up with a lot of negatives to live in places with good weather. California has one of the worst tax and regulatory environments of any state in the country and terrible congestion problems yet its large cities continue to grow. A marginally better business environment is not going to overcome the allure of the sun and beaches.

While a better business environment that is attractive to high-skilled workers and encourages entrepreneurship is unlikely to completely close the gap between a place like San Diego and Dayton when it comes to being a nice place to live and work, it’s a start. And more importantly it’s the only option cities like Dayton, Buffalo, Cleveland, St. Louis and Detroit have.

Exit, voice, and loyalty in cities

Economist Albert Hirschman’s 1970 book Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States presents a theory of how consumers express their dissatisfaction to firms and other organizations after a decline in product or service quality. In terms of interjurisdictional competition exit is demonstrated by migration: dissatisfied residents migrate to a community that better matches their preferences for local government services, externality mitigation, and fiscal policy. Voice, on the other hand, requires staying in place and is usually manifested through voting. Other methods such as protests, letters, and public comments directed at officials may also be effective ways to create change.

Loyalty plays a role in whether voice or exit is employed. Someone who is loyal to a city will be less likely to exit due to a given deterioration in quality. Hirschman argues that loyalty serves an important function by limiting the use of exit and activating voice. If exit is too easy, the quality-conscious people most capable of using voice to elicit change at the local level will tend to leave early, sparking a “brain drain” and generating cumulative deterioration. If some of the most quality-conscious residents are loyal they will remain in place, at least initially, and try to fix a city’s problems from within i.e. they will use some method of voice.

The presence of loyalty within a city’s population has implications for city population decline and growth. The diagram below, based on one from Hirschman’s book on p. 90, shows the relationship between city quality and population.

Exit and loyalty diagram

Quality deteriorates as one moves up the y-axis and population increases along the x-axis, which enables a depiction of the relationship between quality and population similar to that of a traditional demand curve.

The example begins at point A. If quality declines from Q1 to Q2, the population will decline from Pa to Pb. The relatively small population decline relative to the decline in quality is due to the presence of loyalty. Loyalty can be conscious, meaning that the loyal residents are aware of the quality decline and are staying to try to improve the situation, or it can be unconscious, meaning that some residents are unaware that quality is deteriorating. These unaware residents appear loyal to outsiders, but in reality they have just not perceived the decline in quality. Perhaps the decline has not impacted their particular neighborhood or is so gradual that many people don’t realize it is happening. Hirschman notes that unconscious loyalty will not spark voice since by definition the resident is unaware that decline is occurring.

As quality continues to decline from Q2 to Q3 it becomes more observable and even the most loyal residents accept the fact that voice will not save their city. Additionally, the unconscious “loyal” residents will finally notice the decline. Both groups of people will then exit the city in order to reside somewhere else. This leads to a larger drop in population and is shown in the diagram as a movement from Pb to Pc.

This pattern is repeated as a city recovers. An initial quality improvement from Q3 to Q2 induces a relatively small amount of migration back to the city (Pc to Pd), since most people will need confirmation that the city has actually started down a path of sustainable improvement before they will return. Further improvement from Q2 to Q1 will generate a larger increase in population, represented by a movement from point D to point A (Pd to Pa).

What is interesting about this theoretical analysis is that it generates two different populations for the same level of quality. At quality Q2 the city’s population will be relatively large (Pb) if the city is declining in quality and it will be relatively small (point Pd) if the city’s quality is improving. This means that a declining city such as Detroit, Baltimore, Cleveland, Buffalo, etc. will have to make substantial quality improvements before they will see a large influx of people. So even if a city such as Cleveland returns to its 1970 level of relative quality we shouldn’t expect a drastic increase in population, as this model predicts that its population will be less than its actual 1970 population since it will be on the returning curve (CDA) rather than the exiting curve (ABC).

A city that is consistently losing population over a long period of time faces a variety of problems such as increased crime, declining housing values, a decline in the quality of public services, and higher costs in the provision of public services. Fixing these problems is often expensive and this model implies that the costs required for increasing quality from Q1 to Q2 will not result in substantial population gain, which means per capita costs to taxpayers are unlikely to decline by much and may even increase as the city begins to improve. This model predicts that revitalizing America’s struggling cities is a more difficult task than many politicians and policy makers are acknowledging.

Where’s the growth?

In a famous Wendy’s commercial from 1984, three elderly women are examining a hamburger with a rather large bun when one of them asks “Where’s the beef?” in order to express her disappointment that the burger is all bun and no meat. When it comes to the economy growth is like the beef of a burger – without it all you’re left with is fluff and filler.

For the last 8 years the US economy has been mostly fluff and filler. Sure unemployment is down, but that is largely due to a lower labor force participation rate. Wage growth has been anemic and total GDP growth remains below the pre-recession long-run average of 3%.  GDP per capita growth is weak too.

Within a country as large as the US different regions are going to have different levels of GDP per capita and different growth rates for a variety of reasons including labor force characteristics, industry composition, weather, and geography. In order to examine the differences across the US, the graph below depicts the natural log of real GDP per capita in 2009 dollars for the 9 census divisions from 2001 to 2014. Because the natural log is on the y-axis the slope of the line corresponds to the growth rate between years. The black line is the US Metropolitan Area average and does not include rural areas.

ln real per cap gdp by cen div 2001-14

I created the census division average by generating a population weighted average of the real per capita GDP of the Metropolitan Statistical Areas located in each division. The weights are adjusted for each year in the data. Also, since the averages discussed in this post do not include rural areas one can think of them as the urban average in each census division. The population data for the weights and the real GDP per capita data are from the BEA.

As shown in the graph, the highest average real GDP per capita is in the New England division (orange) while the lowest is in the East South Central (purple), although as of 2014 the Mountain is not far ahead.

The slopes of the lines are steeper on average prior to the recession, indicating that the regions were growing faster during the pre-recession period. This is particularly noticeable in the Mountain and South Atlantic division, where real GDP per capita growth has essentially been zero (flat line) since 2009. Growth has also slowed considerably in the Pacific division (dark blue). Only in the East North Central (yellow) and West South Central (brown) does it appear that growth has reached or eclipsed its pre-recession rate.

The next graph below shows the average real per capita GDP by census division in three separate years – 2001, 2007, and 2014. This makes it easier to see the changes in levels over time.

real per cap gdp by cen div 2001,07,14

Real GDP per capita was higher in 2014 than in 2007 (year prior to the recession) in only three divisions – the Mid Atlantic, West North Central, and West South Central. The rest of the country has experienced either no gain or a decrease in the case of the South Atlantic and Mountain divisions. Together these graphs are hardly evidence of a strong economy.

High per capita GDP is not a perfect measure of economic prosperity but it is strongly correlated with many of the other things people care about. Countries with a higher level of per capita GDP are healthier, freer, and happier. The data presented here show that the US economy is struggling when it comes to growth, especially in the South Atlantic and Mountain divisions where people have become worse off on average. Whoever the next president is, he or she needs to come up with an answer to the question – Where’s the growth?

 

Ignoring the adverse effects of the minimum wage may cost taxpayers billions

Today the Obama administration issued a statement calling for a ‘First Job’ funding initiative to connect young Americans with jobs.

The statement laments how difficult it is for young people to find employment and emphasizes how important a first jobs is for future career success:

“After the worst economic crisis of our lifetimes, the United States is in the midst of the longest streak of private-sector job growth in our history, with more than 14 million new jobs created during the past 70 months. But for too many young people, getting a first job—a crucial step in starting their career—is challenging.

When a young person struggles to get their first job, it can have a lasting negative impact on her lifetime income as well as her motivation, pride, and self-esteem.”  

I brought up this same issue 3 months ago in a previous blog post that highlighted the differences in teenage unemployment across cities. And unsurprisingly there are substantial differences – in 2012 teenage unemployment was over 45% in Atlanta and only about 26% in Houston.

So what’s the proposal? A $5.5 BILLION grab bag of grants, skills investment, and direct wage payments to put young people to work. Naturally, the most obvious solution to the teenage unemployment problem is never mentioned – eliminating the minimum wage. In fact, nowhere is it hinted at that the minimum wage may be contributing to teenage unemployment, despite several recent studies affirming this theory.

From a 2013 study:

“Thus, for older workers, the two effects offset one another, and there is little impact on their long-term employment rate. For teenagers, the extra reduction in hiring implies that their employment rates decline. The results are very similar for males and females.”

From a 2015 study:

Using three separate state panels of administrative employment data, we find that the minimum wage reduces job growth over a period of several years”

From a 2015 study:

We find that a higher minimum wage level is associated with higher earnings, lower employment and reduced worker turnover for those in the 14–18 age group. “ (My bold)

From a 2015 study:

I apply the estimator to estimate the impact of the minimum wage on the employment rate of teenagers. I estimate an elasticity of -0.10 and reject the null hypothesis that there is no effect.”

This glaring omission is unconscionable in light of the abundant evidence that the minimum wage harms the least skilled, least experienced workers, which includes teenagers.

As a Prof. David Neumark stated in a recent WSJ op-ed:

“…let’s not pretend that a higher minimum wage doesn’t come with costs, and let’s not ignore that some of the low-skill workers the policy is intended to help will bear some of these costs.”

An all too common occurrence in US policy is that government intervention causes a problem that the government then tries to solve with additional intervention, completely ignoring the possibility that the initial intervention was the source of the problem. In this case, price controls at the bottom of the labor-market ladder have prevented young people from getting on the first rung, so now the government wants to wheel over a $5.5 billion dollar stool to give them a boost.

While this series of imprudent events is not surprising, it’s still frustrating.