Bank Credit Analyst
Following the debacle of the failed attempt to repeal and replace Obamacare, the Trump Administration is focusing on another important part of its policy platform: reforming taxes and reshaping government spending. In theory, the legislative obstacles should be easier to overcome than with the controversial health care bill, but many challenges still lie ahead. Meanwhile, the assumptions underpinning many of the key measures are questionable. The Administration's fiscal proposals are based on the following assertions: The level of U.S. taxes puts the U.S. at a competitive disadvantage and is a hindrance to faster economic growth. Military and infrastructure spending needs to rise sharply after having been cut back too severely in recent years. The federal government, outside of defense, has become bloated and needs to be drastically pruned. Entitlement spending remains politically untouchable. Proposed tax changes will be broadly deficit neutral after allowing for the revenue boost from faster economic growth. The above assertions supporting the administration's policy platform are a mix of facts, fallacies and fantasies. A frustrating aspect of economic debates is that it often is relatively easy to cherry pick data to support any particular argument one wants to make. In other words, there are plenty of alternative facts to choose from. In this report, I will endeavor to illuminate the debate about fiscal policy with unvarnished official statistics, untainted by partisan biases. Are U.S. Taxes Too High? Taxes are a necessary evil if a country's residents want their government to provide some services such as defense, policing, schooling, and old-age benefits etc. In a democracy, the exact level of services provided by a government is a choice that can be voted on at election time. Sometimes, politicians campaign on a platform of increased government spending (and implicitly higher taxes) and at other times, the opposite is true. As a presidential candidate, Donald Trump campaigned on a promise to reduce the government's involvement in the economy and society in general, with a corresponding reduction in tax burdens. The government's revenue grab takes many forms beyond just taxing incomes and can occur at the federal, state or local level. There are taxes on spending, assets, imports and employment, and a multitude of fees ranging from park entrance charges to speeding tickets. Chart 1 shows total U.S. tax and fee revenues from all levels of government, expressed as a share of GDP since 1980.1 The most striking thing about the chart is how little the ratio has changed over the past quarter century. Government revenues have averaged around 27% of GDP over the period and the only years with a marked divergence from that level were the late 1990s when the tech-driven stock market boom triggered unusually strong capital gains tax receipts and in 2009/10 when the economic collapse and temporary tax cuts led to a plunge in revenues. The other interesting point to note is that, according to OECD data, the U.S. is the lowest taxed industrial country, except for Ireland. Taxes and social security contributions as a share of GDP are more than ten percentage points below the unweighted average of 21 other industrial countries. And this gap has been relatively constant over the years (Chart 2). The unweighted average for European countries is almost 39% of GDP. Chart 1U.S. Total Tax Burdens Chart 2U.S. Tax Burdens: An International Perspective As noted earlier, whether a country's overall tax burdens are high or low is largely a reflection of voter preference. In the majority of countries outside the U.S., the government is the main or even sole provider of health care and that often is used to explain the lower level of U.S. taxes. Yet, it is not widely realized that U.S. government spending on health care as a percent of GDP is higher than the industrial country average (Chart 3).2 The point is that the U.S.'s low ranking in terms of global tax burdens does not simply reflect the lack of a universal government-funded health care system. Low taxes are a very good thing if they are sufficient to finance the required level of government services and provide positive incentives for economic growth. However, there is a loose but positive correlation between the level of tax burdens and structural budget deficits. In other words, the countries with low tax burdens have tended to have higher average cyclically-adjusted budget deficits (Chart 4). Again, that is choice that voters can make: choosing lower taxes today at the expense of rising debt burdens that will have costs in the future. The U.S. has been at the extreme end of the spectrum so far this century with the combination of low taxes and large deficits. Chart 3Government Spending on Health Care Chart 4Lower Tax Burdens Generally Mean Larger Fiscal Deficits The data I have shown highlight that the U.S. is a low-tax country from an international perspective and that overall tax burdens have not changed dramatically over time. Nonetheless, there is plenty of scope for reforming taxes in order to improve economic incentives and efficiency. The Case For Tax Reform There is a disconnect between low overall U.S. tax burdens and the facts that the country has the highest marginal corporate tax rate in the industrial world and that so many people feel over-taxed. The principal explanation is the skewed nature of the U.S. tax system with its heavy dependence on taxes on income rather than consumption. The U.S. is the only industrial country in the world without a national value added tax (VAT), and state and local sales taxes are low by international standards. This means that taxes on goods and services account for less than 18% of general government tax revenues in the U.S. compared with an unweighted average of almost 33% for all OECD countries (Table 1). As a result, the U.S. is forced to rely more on taxes on income and profits. These account for almost 48% of tax revenues in the U.S., 14 percentage points higher than the OECD average. General perceptions about tax burdens probably are more affected by income tax rates than by taxes on goods and services, many of which are hidden from view. Table 1The Structure of Government Tax Receipts Problems are compounded by the skewed distribution of income tax payments. For example, although the marginal U.S. corporate tax rate is around 39%,3 many large companies with overseas subsidiaries pay a significantly lower rate. According to Internal Revenue Service (IRS) corporate tax return data, the largest businesses (annual receipts above $100 million) paid an average federal rate of 22.8% on their taxable income in 2013 (the latest year for which detailed corporate returns are available), compared with 32.2% for companies with sales between $10 million and $100 million and 27.5% for those with sales of less than $10 million. It is no wonder that many multinationals are keen to shelter income overseas. There is a case for reforming the corporate tax code to equalize the playing field between multinationals and those with domestic operations. When it comes to personal taxes, there also are distortions. As is well known, there are many hard-to-justify allowances including those on carried interest and on mortgages up to the value of $1 million. Even if the government wanted to use the tax system to subsidize home ownership (which many countries have stopped doing), it would make sense to cap the benefit at the mortgage required to finance a median-priced home. The national median price for a single-family home currently is $230,000. A key problem is the fact that many people do not earn enough to pay much income tax, so the burden falls heavily on a relatively narrow group. The average personal federal tax rate has not changed very much over the past 35 years (Chart 5), but Table 2 shows the remarkably skewed nature of personal tax payments by income level. In 2014 (the latest year for detailed IRS personal data), 148 million tax returns were filed, but more than one-third had no taxable income. Almost 45% of filers reported gross adjusted income of less than $30,000 and, overall, this group received net tax refunds. At the other end of the scale, those with incomes above $200,000 represented only 4.2% of filed returns yet accounted for almost 63% of total federal taxes paid. It is no surprise that many high-income earners feel over-taxed. It is harder to justify the fact that 55% of respondents to a recent Fox News poll said that taxes were too high. The message is that taxes can never be low enough! Chart 5The Average Federal Personal Tax Rate Table 2The Skewed Nature of Personal Income Taxes An obvious way to improve the tax structure would be to eliminate some deductions and use the savings to reduce marginal rates. An even more significant change would be to broaden the tax base by introducing a VAT, using the revenue to dramatically lower income tax rates. The regressive nature of a VAT can be countered by exempting certain items such as food, energy, and children's clothing. The main argument against a VAT is that, once introduced, it becomes an easy way to raise revenue and an initial rate of say 5% eventually could end up at European levels (20%). The proposal for a new Border Adjustment Tax would be a step toward rebalancing tax burdens toward consumption and away from incomes. However, there is considerable opposition to such a move and its future is in doubt. To conclude, the data do not support the notion that the U.S. is overly taxed - either compared to its own history or relative to other countries. But the system has many distortions and there is a strong case for increased taxes on consumption, using the revenues to reduce marginal income tax rates in both the corporate and personal sector. The Case For More Spending On Infrastructure And Defense Unlike tax reform, increased infrastructure spending is not a contentious issue. As Larry Summers likes to quip, anyone flying to New York and driving into Manhattan can see infrastructure spending needs all around, from dreary airports to dodgy bridges and pothole-filled roads. Real government spending on non-defense structures (a proxy for infrastructure) has risen by only 20% over the past 50 years, a drop of almost 30% in per capita terms (Chart 6). As a share of GDP, infrastructure spending has almost halved in the past half century. The administration has talked about boosting infrastructure spending by $1 trillion over the next ten years. If we assume a constant baseline of spending averaging 1.6% of GDP (the 2016 level), an additional $1 trillion would equate to an additional 30% rise in overall infrastructure expenditure over the decade. But even with this increase, spending would still only be 2% of GDP, a relatively modest level by historical standards. The administration's infrastructure proposal is quite reasonable in terms of its scale and desirability. Of course, there is no guarantee that it will materialize. The administration's plan to significantly increase defense spending is a more debatable issue. The number of military personnel has been in a sharp downtrend since the end of the Vietnam War. In the past 35 years or so, a key driver has been the impact of technology with machines replacing people, but defense spending as a share of GDP also has been in a structural downtrend (Chart 7). At the same time, real spending per military employee has been in a strong uptrend, reflecting the switch in strategy away from boots on the ground towards sophisticated equipment. From an international perspective, U.S. defense spending remains very high compared to other countries. According to the SIPRI Military Expenditure Database, in 2015, the U.S. spent as much as the next eight largest military spenders combined.4 Yet, the combined GDP of those eight countries was 55% above that of the U.S. Chart 6Government Infrastructure Spending Chart 7Trends In Defense Spending The Budget Control Act of 2011 put tough spending caps on discretionary spending and these have not been repealed. According to the Congressional Budget Office (CBO), under current law, defense outlays as a share of GDP would fall from 3.2% of GDP to 2.6% by fiscal 2027. The Trump Administration has proposed a $54 billion increase in defense spending authority for fiscal 2018, implying an increase of around 9% from the 2017 level. And while we do not have details, we can assume that the longer-term plan is to reverse the downtrend in spending as a share of GDP. What is the right level of defense spending? The world remains a dangerous place, but the U.S. already outspends other countries by a huge margin. At the end of the day, financial constraints mean it boils down to a choice between defense and other spending programs. Voters may state a preference for increased defense spending, but that likely would change if other programs were crowded out. Is The Federal Government Bloated? Chart 8Federal Non-Defense Discretionary Spending Spending on entitlements is widely regarded as untouchable from a political perspective and it is no surprise that Trump has promised to defend these programs. Given the administration's platform of tax cuts and increased military and infrastructure spending, containing budget deficits implies tough constraints on non-defense discretionary spending. This includes spending by the Departments of Agriculture, Education, Energy, Homeland Security, Health and Human Services, Justice, State and Veterans Affairs. Such spending has already declined sharply during the past several decades, both as a share of total government outlays and as a share of GDP (Chart 8). The administration seeks further drastic cuts in the years ahead. There is a general perception that much of government spending is wasteful, implying huge savings can be made. At the same time, surveys show that people do not want cuts in areas such as security, veterans affairs, education and health. The problem is that spending by the Departments of Education, Health and Human Services, Homeland Security, Justice and Veterans Affairs account for more than half of non-defense discretionary spending. Thus, pressures for spending cuts fall heavily on other areas. But this often is not practical given that many of these other programs are so small. For example, spending on foreign aid represents less than 0.2% of GDP and less than 5% of non-defense discretionary spending. As for federal employment being bloated, it should be noted that civilian federal employment has shown no net change over the past 50 years, despite the marked growth in the population and economy over the period. Federal employment currently accounts for less than 2% of total employment, down from 4% in 1970 (bottom panel of Chart 8). There inevitably are areas of wasteful government spending and it is appropriate to look for savings. However, it is not reasonable to believe that there can be tax cuts and increases in defense spending and domestic security, while protecting entitlements programs and preventing a massive rise in the budget deficit. And that is even without adding in the cost of the proposed border wall with Mexico. Entitlement Spending Is The Major Problem Social Security has been called the third rail of American politics - touch it and you are dead. No politician seeking election would dare campaign on a platform of major cuts to the program in the form of reduced benefits, higher contributions, means testing, or an increase in the age eligibility limit. And the same is broadly true for Medicare. Voter dislike of government involvement in the provision of health care does not seem to extend to those over the age of 65! The combination of rising life expectancy and a decline in the ratio of taxpayers to retirees will place growing financial strains on the Social Security and Medicare systems. In 1970, there were 5.4 people between the ages of 20 and 64 for every person 65 and older. That ratio has since dropped to 4 and will be down to 2.6 within the next 20 years (Chart 9). Spending on entitlements (Social Security, Medicare, Medicaid, income security, and government pensions) is on an unsustainable trajectory. In fiscal 2016, these programs equaled 74% of federal revenues and the CBO estimates that this will rise to 84% by 2027, absent any change to current law (Chart 10). If we also allow for net interest costs, total mandatory spending is projected to exceed revenues within the next 12 years or so, meaning that deficit financing will be required for all discretionary spending. Chart 9The Demographic Fiscal Headwind Chart 10The Entitlement Problem Politicians operating in a world of two-year election cycles have no incentive to support short-term pain for long-term gain. At some point, markets will force change, but it is hard to know exactly when that will happen. According to the CBO's latest estimates, current policies imply that the federal deficit will average 4% of GDP over the next decade, rising to 6.2% and 8.4% over the subsequent two 10-year periods. As a result, the debt-to-GDP ratio rises from 77% currently to 113% by 2037 and 150% by 2047.5 Of course, that is a long way in the future and much can happen to undermine these projections - for the better or for the worse. Long-run fiscal projections are subject to a wide margin of error because, in addition to legislative changes, they are very sensitive to assumptions about economic growth, inflation and interest rates. The CBO's baseline estimates published in mid-2009 had Medicare spending rising from 3.1% to 7.2% of GDP between 2010 and 2037. The latest CBO report has 2037 Medicare spending at a much lower 5.3% of GDP, representing massive savings from the 2009 estimate. Unfortunately, total federal revenues as a share of GDP were revised down by an even greater amount, with the result that expected deficits and debt levels have been revised up sharply since the 2009 report, despite the slower path of Medicare spending (Chart 11). Chart 11Long-Term Fiscal Projections: Prone to Revisions One can point to Japan as an example of how a high government debt-to-GDP ratio need not imply economic disaster. Japan's gross debt currently stands at 250% of GDP and there has not been any difficulty in financing its ongoing deficits. However, two qualifications are necessary. First, it is too soon for Japan to claim victory: its horrible demographic profile points to an ever-worsening fiscal position and there likely will be a crisis at some point. Secondly, Japan finances its deficits internally which protects it from the whims of foreign investors. Although the dollar's status as reserve currency also gives the U.S. protection, the country's ongoing large current account deficit creates vulnerability to financing problems if overseas investors lose confidence in the U.S. fiscal outlook. Concluding Thoughts Public discussions of fiscal policy invariably morph into partisan arguments about the appropriate size and role of the government in the economy. It quickly becomes frustrating when the warring factions then use misleading or outright wrong data to support their positions. In the spirit of the adage that "everyone is entitled to their own opinions, but there is only one set of facts," I have focused this paper on published and reputable data about government revenues and spending. Several points emerge: One may want taxes to come down, but it is a FACT that the U.S. is a low-tax country by international standards, and tax burdens have not noticeably risen over time. It is a FACT that the U.S. tax system has serious distortions and is crying out for some reform. But what these reforms should be is open to debate, and are a matter of opinion. There is a strong case for increased infrastructure because it is a FACT that spending has fallen sharply over the years. It is less obvious that a major rise in defense spending is warranted. It would be a matter of preference rather than incontrovertible need. It is a FALLACY to describe overall non-defense discretionary spending as massively bloated and out-of-control. Of course, there are many places where the government can make cuts and improve efficiency, but squeezing this category of spending will provide only limited savings. It is FANTASY to think that entitlement programs can be maintained over the long run in their current form. The longer that reforms are delayed, the bigger the cutbacks will have to be. Government deficits and debt do matter, but it is virtually impossible to predict when financing problems might occur. There is no particular level of the debt-to-GDP ratio that will trigger a crisis because much depends on the domestic and global economic and financial environment. But, to quote the late Herb Stein, "if something cannot go on forever, it will stop." The Trump administration's fiscal desires are a mix of sensible policies, wishful thinking and impracticalities. Hopefully, there will be progress with boosting infrastructure, and making some positive reforms to the tax code. However, there will be serious challenges to tax changes once special interests get involved. The end point may very well be outright tax cuts without reform, and that would be much less desirable. On the spending side, increased defense spending is a perfectly legitimate choice, but the planned severe cuts to non-defense discretionary spending are impractical. The good news is that the odds of such severe cuts being implemented are very low. It seems almost certain that federal deficits will head higher over the coming few years. Using dynamic scoring to suggest that the economy will improve by enough to make tax cuts and spending increases virtually self-financing will have little credibility outside of the administration and will be challenged by the calculations of the CBO and Joint Committee on Taxation. If the government is successful in implementing major fiscal stimulus then the biggest problem might be overheating the economy. As I discussed in a recent report, the U.S. economy already is operating close to full capacity and it will not take much to create a classic late-cycle build-up of inflationary pressures.6 That would set the scene for enough Fed tightening in 2018 to give high odds of a recession in 2019. Martin H. Barnes, Senior Vice President Economic Advisor mbarnes@bcaresearch.com 1 The totals exclude government interest receipts and transfers from the Federal Reserve to the Treasury as those largely represent transactions related to intra-government holdings of Treasury securities. 2 Overall the U.S. devotes a much larger share of its GDP to health care than other countries. According to OECD data, total health care spending represented 16.9% of U.S. GDP in 2015, compared to an unweighted average of 10% for other industrial countries. Within these totals, the government share was 8.4% in the U.S and 7.6% elsewhere, with the private sector making up the difference. 3 This comprises a top federal rate of 35% and state and local taxes of 6%, fully deductible against federal taxes. 4 SIPRI stands for the Stockholm International Peace Research Institute. Details available at https://www.sipri.org/databases/milex 5 For more information, please see The 2017 Long-Term Budget Outlook, Congressional Budget Office, March 2017. Available at www.cbo.gov 6 Please see BCA Special Report, "Beware the 2019 Trump Recession," dated March 7, 2017 available at bcaresearch.com
