12 August 2010

Ten Public Finance Reforms For Colorado

Colorado has an inherently unstable system of public finance that is ill equipped to handle business cycles, and has a host of other problems that create bad land use incentives, encourage environmental damaging development of open space when it doesn't make economic sense to do so, makes it hard to fiscally manage local governments sensibly, distorts its economy, and fails to provide appropriate education funding to all students who should be benefiting from public education at both the K-12 and higher education levels.

This post explains the problems and proposed ten public finance reforms to put Colorado on a firmer fiscal foundation, with only slight adjustments to overall tax levels, without running long term deficits, and with only modest adjustments to the mix of taxes paid by people in Colorado. These are long term, permanent solutions, rather than band aid solutions. These proposals concentrate cyclic public finance pressures in state government and gives state government the tools it needs to handle these pressures better, and reduces the harm that recessions inflict on the state economy. They also seek to make our state and local public finance system less complex, and to reduce the impact that state and local government policies have on economic decision making in the state.

Tax Revenue Is Not Stable

Tax revenues are very cyclic. Sales tax revenues amplify booms and busts because the tax base excludes the necessities that can't be cut even in hard economic times. Income tax revenues fall with payrolls and declining investment income during recessions. Tourism and gambling based tax revenues fall during recessions with a fall in luxury and business travel. Real estate transfer taxes in resort communities are similarly highly cyclic.

There are less cyclic taxes as well.

Property taxes change only when properties are reassessed, have a near 100% collection rate because the first priority liens they create on real estate with high interest rates are good investments and can be sold quickly to private investors, and can be designed so that revenues remain constant even when property values change.

Colorado also imposes its property taxes on vehicles as part of vehicle registration charges, and this tax base is also quite stable, although collection rates aren't quite as immaculate and the revenues from these taxes gradually declines if people postpone buying new vehicles.

"Sin taxes" on liquor and tobacco, are relatively more stable, because people buy these products in good times and in bad, and because these taxes are based on volume rather than product price. There is a long term decline in tobacco purchases, however.

Gas taxes vary more, but cyclic variation in gas taxes is less of a concern because reduced driving also reduces need for the road maintenance that they finance.

Severance taxes on oil and gas production fluctuate wildly with the oil and gas commodity markets, but these fluctuations aren't neatly cyclic. Sometimes they boom during recessions. Sometimes they slump when the economy is strong.

Government Service Demand Is Not Stable

Demand for government services is also cyclic. In recessions, people need more government assistance. In economic booms, people need less. As in the case of taxes, however, some forms of demand for government services is greater than others.

The demand for means tested financial assistance programs, like free and reduced price school lunches, food stamps, cash welfare payments, Medicaid, unemployment assistance, housing assistance, homeless shelters, foreclosure assistance, and grants to non-profits that help the poor all rise when the economy is weak, and fall when the economy is strong.

Demand for higher education is cyclic. When the labor market is weak, people go to college and retrain.

Other government programs are relatively indifferent to business cycles. Water and sewer services, fire protection, public education, libraries, parks, the conduct of elections, and police protection all cost about the same amount to provide in good times and bad. The cost of operating prisons tends to be stable, because, contrary to common perceptions, crime isn't strongly cyclic.

Demand for a few government programs is counter-cyclic. Demand for building permits and inspections, zoning changes, and public works construction for new developments, for example, rises when the economy is strong, and falls when the economy is weak.

Against Balanced Budgets

So, in general, the tax revenues fall and demand for government services rises during recessions, and tax revenues rise while demand for government services falls during economic booms. Given this predictable trend, the notion that government expenses should match government revenues in good times and bad doesn't make good sense. Expressed this way, balanced budgets in every fiscal year are bad policy.

While medium and long term balance between government revenues and government expenses is desirable, short term and annual balance between government revenues and government expenses is a bad idea.

Distortions In Land Use Incentives

As noted, not all tax revenue streams are cyclic, and a lot of important forms of public spending are stable. In general, local government programs tend to have stable spending, while state and federal government programs tend to have cyclic budget demands.

Local governments in Colorado, however, have not heeded this fact. Most municipal governments in the state, and many counties, are highly reliant on sales taxes for revenues, although they do collect some relatively stable property taxes, despite having programs (other than state funded and locally administered programs) that require stable funding. Colorado's state government gets the vast majority of is tax revenues from taxes whose revenues are cyclic, and has strongly cyclic funding demands.

Municipal reliance on sales taxes and the structure of Colorado's property tax system also creates problems in land use policy. Reliance on sales tax revenues gives Colorado municipalities a disproportionately strong incentive to encourage retail development over other forms of commercial, industrial and residential development.

Colorado's Gallagher Amendment further complicates the matter, by imposing higher property taxes on business property than on residential property. As a result, residential property development, already discouraged because they don't generate sales tax revenues, are further discouraged because the property taxes generated from residential development aren't sufficient to cover the demand for the government services that residential development generates.

The Gallagher Amendment also further enhances the incentives for urban planners in local governments to encourage retail development, whose sales tax and disproportionately high property taxes cause these developments to generate far more tax revenue than their demand for public services. The Gallagher Amendment also makes non-retail commercial development and industrial development reasonably attractive, because these land uses generate little demand for government services but generate disproportionately high property tax revenues.

Commercial and industrial development also generates tax revenues from "head taxes" that many local governments employ on a fixed amount per employee per year basis, while residential development does not.

The land use biases created by the tax system affects the entire state economy and the state's environment.

Existing local governments have strong incentives to enter into a race to the bottom to attract the retail businesses that generate the most taxes in excess of their demand for services, and to a lesser extent commercial and industrial properties, while discouraging residential development.

As a result, developers respond to these regulatory incentives by putting new residential development in "green fields" rather than doing infill development, because there are no zoning and building code restrictions to prevent this from happening there.

Also, developers often don't bear the full costs of building new infrastructure in "green field" areas, whose roads and new schools are often heavily subsidized by state government departments that have little or no ability to tax or prevent new land uses. Likewise, infill developers rarely profit from the fact that infrastructure is already in place to support their development, reducing costs to the public.

The bottom line that flows from these facts is that new residential development is frequently built in unincorporated areas that require costly new infrastructure, rather than in more densely populated existing municipalities that have infrastructure already in place. This also eats away at a finite supply of open space in the state.

In sum, Colorado's current tax system creates incentives to over encourage retail development, to discourage infill residential development, to unnecessarily invest in new government infrastructure, and to unnecessarily develop open space.

Distortions To the Colorado Economy and Tax Base

These incentives distort not only our land use, but our economy. It encourages a retail heavy economy that also over invests in the construction industry, and in particular the part of the construction industry that builds suburban housing and new government infrastructure.

A system that allows retail businesses to reduce their share of the tax burden by locating themselves in unincorporated areas, a particularly great problem with auto dealership whose products are high value, infrequent purchases where inconvenience is outweighed by the tax savings of making a vehicle purchase in a low tax jurisdiction.

Tax Administration Problems

The great variety of local sales tax bases and sales tax rates in Colorado complicate tax administration and make operating retail businesses with multiple locations in Colorado much more cumbersome. For example, groceries are not subject to state sales taxes, but are subject to local sales taxes in many localities, and grocery store chains have to manage a system where there are differences in sales tax rules for every store.

It is also much easier to collect sales and use taxes in taxable mail order and interstate transactions when there is a single uniform state sales tax than it is when there are different sales taxes for every locality.

Educational Inequality

The per capita size of the property tax base varies dramatically from one locality to another. Some jurisdictions, like resort towns, have an immense property tax base for each resident. Others, like rural communities, have a very large property tax base for each resident.

In the area of providing local government services, this isn't problematic. Local residents who benefit from local governments services are expected to pay for those services and if that means that high property taxes must be imposed (and perhaps passed onto customers from outside of the locality in the case of businesses), so be it. But, in the case of public education, this is a problem.

Colorado has a state constitutional mandate to provide free and adequate public education to every child. So, disparities in the availability of property tax funding for education for K-12 education varies dramatically from place to state in the state, and the state government has a legal duty to make up this shortfall.

The current state school funding formula goes part of the way in solving this problem, but at best, the existing system is clunky and provides school systems no say in residential property development that dramatically impact the financial burdens imposed on those school systems. In reality, the current system also creates gross inequalities between large property tax base and low property tax base areas that cause the state to fail to meet its duty to provide an adequate free public education to every child.

Funding schools with local property taxes also greatly complicates the state policy of allowing any child to choice into any public school to the extent that space is available.

Political Mismatches

Local sales taxes are also popular because they seek to get people who don't live in a community and impose few service demands upon it to pay its taxes. A political disconnect between the people who approve sales taxes and those who pay them is problematic. While taxation without representation is not a strictly prohibited by the state or federal constitutions, it is still a bad way to run a democracy.


A number of reforms to the state and local tax system in Colorado could enhance its fiscal stability during recessions, reduce tax induced demand to develop open space, reduce unreasonable competition between jurisdiction for retail development, and reduce tax compliance and administration costs.

The first two proposals really need to be enacted together to prevent unnecessary disruption in overall property tax rates. But, otherwise, each proposed reform could be enacted independently of the others, and the fourth proposal for a rainy day fund and credit facility in times of declining state revenues, could be broken up into two independent proposals.

1. Replace local sales taxes with a single uniform state sales tax at a rate comparable to the existing combined state and local tax rate; replace the lost local government revenue with property taxes.

Making sales taxes an exclusively state responsibility and making sales taxes uniform, and replacing the revenue lost at the local level with property taxes, would provide several benefits:

* Local governments would have much more stable and predictable revenue streams, reducing their reliance on inherently inaccurate economic forecasting and allowing them to maintain basic government services, which mostly have stable spending demands, undisturbed even in hard economic times.

* The sales tax base would grow as businesses could no longer reduce sales taxes by locating themselves in low tax jurisdictions.

* The incentive for local governments to disproportionately encourage retail development in a way that distorts the economy and land use decisions would be reduced. This would also help the environment by reducing the tax incentive to develop open space.

* The cost to government and retail sales businesses, particularly those with multiple locations, of administering the sales and use tax would be greatly reduced and sales tax enforcement would be easier.

2. Replace local property tax funding of schools entirely with state funding of schools.

* The reduction in the property tax burden created by ending local property taxes for K-12 education almost exactly matches the increase in the property burden created by requiring local governments to replace sales tax revenues with property tax revenues.

* It is much easier to develop a sensible school funding formula that provides an adequate free public education to every child when a highly varied mix of local and state funding doesn't have to be considered.

* Statewide school choice will meet much less resistance from schools if their funding isn't tied to local property taxes.

* Amendment 23, guaranteeing school districts stable funding would remain in place, matching their stable expenses, so that the task of balancing cyclic state government revenues and expenses would fall to state government alone, rather than having to be addressed in parallel by every school district in the state.

3. Develop a coordinated development fee that adequately reflect the public cost of new infrastructure.

New development should be required to pay development fees to each local government impacted by the development that reflect the public cost of that development, collected by the local government in charge of building permits and distributed to the governments impacted in proportion to the public costs they incur.

* School districts would receive a share of the development fee that would mostly reflect the cost of building new buildings, equipment and buses and the hiring process new teachers (but not their continued employment and operations which would be covered by state school funding formula money). This would be lower in places already served by schools that aren't full, and higher in places where new schools will have to be built, even if that means that development fees are different in different parts of the same school district. State standards would determine how these costs are determined.

* State government would receive a share of the development fee that would mostly reflect the cost of building new state roads and opening new government offices.

* Local governments would receive a share of the development fee that would reflect the cost of building new local roads, other new local infrastructure, and tap fees to secure water for the new development. State standards would determine how these costs are determined and they could vary in different places within a locality.

* Developers would be required to show proof of ability to pay development fees before beginning construction.

* Development fees don't stop growth that can pay for itself, but it does discourage growth that would be economically irrational by privatizing externalities associated with development. They do encourage smart growth by favoring development in places where it has less public fiscal impact. This check on economically irrational development would discourage the kind of housing bubbles that were intensely harmful where they occurred whose collapse caused the financial crisis.

4. The state government should establish a rainy day fund and credit facility.

In times when state revenue exceeds the highest revenue in the medium term past (perhaps five or ten years), a significant share of the revenue increase (at least as much as the TABOR amount which would be discontinued) should be required to be devoted to a rainy day fund.

In times when state revenue falls, the state could tap the rainy day fund, and to the extent that it was insufficient, borrow money with medium term debt (with a repayment period of perhaps five to ten years) without voter approval, to make up a large share of the decline, perhaps half to three-quarters of it, so that state budget cuts in hard economic times would be muted.

A state budget that is under no stress during hard economic times would probably be even better as an automatic stablizer in hard economic times, but politically, some amount of shared sacrifice by government during hard economic times is probably necessary. But, a "Little Hoover" economic policy of fiscal restraint during recessions is bad policy from a macroeconomic perspective.

A rainy day fund and credit facility would give the state the power to balance out flush good years and scarce bad years in a way that wouldn't take superhuman political will. Funding a rainy day fund is much more politically acceptable when it comes as a share of increased revenues in good times, rather than forcing a cut to existing spending levels for government programs.

State legislators could still cut spending in hard times, but they wouldn't be required to do so by fiscal necessity even if it didn't make sense to do so. Indeed, even in a TABOR style regime, voters might actually favor debt to allow government programs to remain funded in recessions, but recessions rarely happen with enough advanced warning to allow the electoral process to run its course in time for voters to act.

5. Allow temporary tax increases to make up declining state revenues.

TABOR's requirement of voter approval for tax increases also means that in tight budgetary times, legislators cannot balance tax increases and spending cuts. Yet, usually, a mix of the two would be the best choice.

Allowing temporary tax increases up to the amount of a state revenue drop as reduced by rainy day funds, would even the scale between taxes and spending, while limiting that authority in the absence of voter approval to amounts necessary to maintain the status quo of government spending.

This way, legislators could use a combination of spending cuts, temporary tax increases, rainy day fund spending, and medium term borrowing to address dropping state revenues in recessions, rather than having to address the entire burden of recessionary declines in revenue through spending cuts no matter how much hard that would do to state programs and the state economy. This would make it easier to secure a "soft landing" when the state economy is weak.

6. Gas taxes should be increased to, at least fully cover the cost of road maintenance.

Gas taxes are basically a fair users fee. Those who use roads pay for their upkeep in rough proportion to the amount that they use roads. The general fund, which has to cope with a great many cyclic pressures, should not have to deal with road maintenance as well when there is a more fair alternative.

7. The Gallagher Amendment should be adjusted to reduce the disparity between residential and business property tax rates.

Property taxes make intuitive rough justice sense as a way to pay for local government when only residential property is considered. Home value is a proxy for a family's fare share of local government costs, which are paid either directly through property taxes, or indirectly through rent a part of which is used to pay property taxes.

Businesses complicate this sense of rough justice. Everyone pays property taxes indirectly through increased prices charged by businesses for their goods and services, in addition to paying them directly on homes that they own and indirectly on homes that they rent. The right proportions are not obvious.

What is obvious is that the current balance produces property taxes that are insufficient to cover the cost of providing local government services to residential property, while providing more tax revenues than the cost of providing local government services to businesses.

When a state formula allocates the share of the total property tax base to be paid by businesses and residential properties respectively, as the Gallagher Amendment does, it is important to get these revenues and expenses in balance, because the share of the property tax base that is residential varies greatly from one property taxing jurisdiction to another.

A mismatch of tax collections and local government service provision costs artificially discourages local governments from permitting residential development in property taxing jurisdictions.

But, dramatically changing the Gallagher formula immediately and dramatically, for example, by ending all distinctions between residential and business tax burdens in a way that greatly increases individual property taxes would be a problem, because families have built their family budget around expectations of what their property taxes will be under Gallagher and can't easily make different residential choices in the meantime.

Addressing the current imbalance doesn't have to be an all or nothing matter. There is no sacred reason that business property tax rates and residential property tax rates have to be exactly balanced to the cost of providing local government services to those kinds of property. The harm filters through to local governments through this imbalance mostly arises through the filter of the incentives local governments have in formulating land use regulations and economic development plans, so the connection isn't direct or exact.

Ending the funding of local school districts with property taxes also reduces the pressure to get the balance exactly right, because local governments adjust other taxes and charges, like head taxes and licensing fees, if the property tax rates for residential or business property are a little too high or a little too low.

Also, an increase in local government property tax revenues by increasing local property taxes to make up for the loss of state sales taxes that is made manageable by eliminating property taxes that fund local school districts would help bridge the gap between the cost of providing services associated with residential property and the tax revenues generated by residential property.

A good measure of the reliance interest that homeowners have in the financial decision associated with buying a home is the average time between sales of a home, which is about ten years.

Right now, residential property pays property taxes based upon 29% of their actual valuation. A reasonable adjustment to Gallagher might increase this percentage from 29% to 49% over ten years, with the valuation increased by two percentage points each year over ten years, and nominal property tax rates reduced each year to make total property tax revenues neutral over the adjustment period. At that point, the public could look at the situation and see if further adjustment is needed.

8. Unemployment insurance rates and benefits should be increased, and unemployment eligibility should be liberalized.

There is already one rainy day fund in the current tax system. Unemployment taxes go into a trust fund that is used to pay unemployment benefits. But, those taxes are low, the benefits paid for meager, and many unemployed people don't qualify for them, so the inadequately buffer our economy.

Stingy unemployment benefits come at a price. Weak unemployment benefits are one of the main reasons that demand for government services is so cyclic. People who are unemployed and receive inadequate unemployment benefits experience great financial distress that results in other cyclic burdens on government. Unemployed people lose their homes causing property values to fall. They buy less causing sales tax revenues to fall and shrinking the economy. They drop health insurance because they can't afford to pay for continuation of health insurance coverage and get themselves or their children on Medicaid funded health care instead. Their families fall apart putting long term burdens on a variety of social services and law enforcement resources.

Certainly, there needs to be an incentive to find work. But, the current system, which involves immense red tape regarding efforts to look for work, is overkill given that benefit levels are far below what is necessary to give people strong incentives to return to work. And, every unemployed person who is excluded from employment because their termination is not a type that qualifies for benefits still has needs that must be met somehow or other.

Even with a much more generous unemployment system, there would still be a strong incentive to find work, but the burdens placed on the public and the health of the economy by unemployment would be significantly reduced.

Also, the way the current system is run, in which great care is taken to make sure that anyone who has obtained employment should be stripped of benefits immediately, discourages unemployed workers from securing partial solutions to their problem, like temporary jobs, marginal partial employment or seriously pursuing jobs that might not work out and result in a swift unemployment benefitless termination.

A cheaper to administer system might replace short term unemployment benefits limited largely to people who are laid off, with a simple, no fault severance payment system that provides a certain percentage (perhaps two-thirds or three-quarters of pay) up to a dollar cap that is higher than under current law (perhaps $650 a week), in installments for a fixed time period (perhaps the three months), regardless of re-employment with another employer during the three month period, to anyone whose employment ends regardless of the reason and ends the link between particular employer unemployment rates and claims experience for these no fault claims. The dollar cap (but not the percentage of the total cap) might be removed for long time employees. This no fault severance pay system could fund lost income during family leaves like maternity leaves, as well as true unemployment.

Essentially, this kind of no fault system would be a mandatory emergency fund savings plan, with an insurance element that allows it to pay out even if the payroll tax deductions haven't matched the amount payable yet.

Long term unemployment might look more like the current system and start after three months from termination have passed, but could last longer, perhaps two years rather than the six months that is the norm when unemployment levels are normal and Congress does not extend benefits.

Unemployment benefits might also be supplemented by a system that paid a large percentage (perhaps two-thirds or three-quarters) of any continuation of coverage costs incurred for health insurance.

9. Establish a state property tax and use it to reduce state sales taxes.

Another way to make the state budget more stable, is to replace a cyclic revenue base with a stable one on a revenue neutral basis. Even a state property tax small enough to cover only a one percentage point reduction in a unified state sales tax would add a great deal of stability to the state's revenue streams, thereby reducing the strain on other counter-cyclic automatic stablizers like rainy day funds. Property taxes are the most stable revenue stream while sales taxes are the least stable revenue stream.

Alternately, to ease the burden on homeowners, part of the state property tax could be used to fund a reduction in state sales tax, and another part could be used to fund a homestead exemption available to all homeowners, which would disproportionately help the homeowners most heavily impacted by other measures in the package.

10. Shift significant higher education funding to need based scholarships.

Across the board vouchers for in state students for higher education is an inefficient way to fund higher education. It provides substantial aid to large numbers of college students whose families can afford to pay for college without financial assistance, and provides inadequate aid to college students who can't afford to pay for college.

This inefficiency is heightened in hard economic times when large numbers of people, affluent and less affluent alike, who would have otherwise entered the workforce go back to school instead.

While this would require administrative work to determine financial need, this administrative effort is something that is being done in any case, so the added administrative burden of this approach is minimal.

Directing funds to people whose attendance at college depends upon financial support, with more generous need based packages, would provide more benefit for the same funds than the current system.

Also, a need based system would significantly reduce the demand for higher education funding in good times, because there would be fewer students then and there would also be fewer students from financially stressed families because fewer families would be financially stressed. This in turn would help build up the rainy day fund for hard times.

No comments: