Government Spending Across the World How the United States Compares

University of New Hampshire CARSEY RESEARCH Carsey School of Public Policy National Issue Brief #144 Fall 2019

Michael Ettlinger, Jordan Hensley, and Julia Vieira

Introduction and Summary Comparisons of how much the governments of different countries spend, and on what, can illuminate the range of fiscal policy options available and provide insight into which approaches work best. They also can tell us what different countries value. Levels of spending vary greatly across the world. In France the spending of all levels of government accounts for more than half (56.4 percent) of the economy, as measured by gross domestic product (GDP). At the other end of the spectrum is Colombia, where government spending is less than a third (31.9 percent) of GDP. At 38.0 percent, the federal, state, and local governments of the United States rank twenty-fourth among the twenty-nine countries for which recent comparable data are available.1 Where countries rank in government spending is mostly driven by how much they spend on “social protection,” a category that includes most non- health-care social safety net and social insurance programs (for definitions of spending categories, though low compared to other wealthy countries and see Appendix A, “OECD Spending Categories”). twenty-second in social protection. Countries that are high spending overall spend a lot By either measure, the United States ranks near on social protection, while countries that are low the top in military, health care, education, and law spending overall spend less. The United States ranks enforcement spending. Its military spending is equal last in social protection spending. to the combined military expenditures of the next Spending as a share of the economy is, however, seven highest-spending countries combined.2 just one way of measuring the differences between The choices the United States and other countries countries. While it is a good measure of how much make regarding how they spend their resources are countries spend relative to their available financial influenced by many factors. There is no single right resources, it is not necessarily a good measure for choice. The decisions governments make reflect dif- comparison of the level of service provided by gov- ferent values and priorities which lead them to differ- ernments. A better measure for that is spending per ent conclusions as to the best way to reach their goals capita—the amount spent per person. By that measure, and face their distinctive challenges. Some countries the United States ranks closer to the middle overall, value projecting military power throughout the world. 2 CARSEY SCHOOL OF PUBLIC POLICY

Others see military spending as a Measuring Government But most countries similar to the diversion of resources from more Spending constructive uses that benefit their United States are included. There are society and residents. Some coun- Comparing how governments in some relatively poor countries in the tries see universal access to college different countries spend is not as OECD statistics, but they are not the education as of broad value, while easy as it might seem. There are poorest countries in the world, and others do not believe one person’s multiple ways to measure spending, the public has a say in their fiscal mat- tax dollars should pay for another and governments present their data ters through democratic processes. person’s college education. Some differently and spend in their own Even with the benefit of OECD- countries invest heavily in preserv- currencies. Also, some countries compiled data from roughly ing their culture’s arts and letters; are so different from each other similar countries, the best way to others are less interested. Some that comparisons are not illumi- compare spending among coun- countries see alleviating poverty as a nating. Comparing U.S. spending tries depends on the question moral imperative; others see spend- to that of the small island nation for which one seeks an answer. ing on the poor as rewarding failure of Nauru would tell us little about We measure spending in three and breeding dependency. The list the relative values and convictions ways: as a share of the economy of choices is almost endless. of the two countries, as they face (spending as a share of GDP); per The reasons countries make the enormously different challenges person; and, solely for military decisions they make are as varied as with enormously different assets. spending, the amount in U.S. the countries themselves, but one In this paper we address some dollars (converting from foreign common factor is attitudes toward of these challenges by employing currencies for other countries). taxation. In the long run, the overall statistics from the Organization Each of these measures has its level of spending, with the conse- for Economic Cooperation and advantages, which are discussed in quential impact on each category of Development (OECD). Using the sections where each is used. spending, is related to the level of OECD statistics helps in two ways: One final complication is that taxes. Some countries believe that • The statistics available from some countries, the United States lower taxes are the path to economic the OECD are presented in a being a prime example, deliver success, or they question the morality consistent form for most of the some government spending in of taking money from their citizenry, countries included, making the form of targeted tax breaks. and thus tax and spend relatively valid comparisons possible. For example, while the United little. Other countries believe that States does have housing support • The countries from which the public investments in such things programs that are implemented OECD collects and organizes as infrastructure and education are through checks sent to landlords data (member countries of the the key to economic vitality, and that and residences directly built or OECD plus a handful more) public spending is the path to a more leased, its most expensive hous- are primarily economically just, fair, and safe nation. For those ing support program has been the developed, democratic, free- countries, drawing taxes from private deduction for mortgage interest market countries with a role hands for a greater good is warranted. on federal and state income tax in the world economy—crite- forms. The data for including these ria which make comparisons “tax expenditures” in comparisons 3 mostly “apples-to-apples.” among countries are extremely There are some countries, Canada hard to come by, but at the end of for example, that would be useful to this paper we investigate whether include but that are missing from the excluding tax expenditures from OECD statistics because of challenges our analysis has a significant in collecting recent comparable data. impact on our conclusions. CARSEY SCHOOL OF PUBLIC POLICY 3

Government Spending TABLE 1. TOTAL PUBLIC SPENDING AS A SHARE OF GDP, 2017 as a Share of GDP For the twenty-nine countries for which the OECD reports recent comparable statistics, government spending as a share of GDP ranges from a high of 56.4 percent in France to a low of 31.9 percent in Colombia. The median for the twenty-nine countries in 2017 was 42.4 percent. Table 1 divides these countries into high spending (more than 48 percent of GDP), middle spending (40 percent to 48 percent of GDP), and low spending (less than 40 percent of GDP).4 Figure 1 shows the same information in Note: High-income countries are in green, middle-income countries purple, lower-income countries red, graphic format. Governments in lowest-income countries darker red. Source: OECD, “General Government Spending (Total),” 2019. the United States—federal, state, and local—spend 38.0 percent of FIGURE 1. TOTAL GOVERNMENT SPENDING AS A SHARE OF GDP, 2017 national GDP, placing the country twenty-fourth of the twenty-nine countries. What does the share of national income devoted to government spending tell us about a country? Overall, it can tell us how much a country values the goods and services that can be provided only through government and the extent to which it believes that other goods and services are most effec- tively and fairly delivered by government versus the private sector. Spending as a share of the economy in specific categories can be telling about the values of a country, as it shows how much a country Note: Color coding is by spending level: yellow = low, blue = middle, orange = high. spends toward achieving a goal Source: OECD, “General Government Spending (All),” 2019; and authors’ calculations. relative to its capacity to spend. It can be the equivalent of asking a about which economists are very outcomes for their people or household what shares of its income interested. One of the most con- whether countries are better off it puts toward charity, saving for the tested questions in public econom- when taxes and spending are low. future, or buying the latest in-home ics is whether governments that One might expect that a richer appliances and electronics. choose to allocate a greater share of country would spend a smaller Government spending as a share their economy to taxes and public share of its income on public of the economy is also a statistic spending produce better economic services than a poorer country. 4 CARSEY SCHOOL OF PUBLIC POLICY

After all, just as a rich family that green, middle-income countries the amount spent on social protec- spends 10 percent of its income on (per capita GDP from $35,000 tion. Of the 24.5 percent of GDP food eats better than a poor fam- to $44,000) are in purple, low- difference between France and ily that spends 20 percent, a rich income countries (per capita Colombia in overall spending, country can get more out of gov- GDP less than $35,000 and more 15.3 percent is the spending dif- ernment spending that is a smaller than $20,000) are in red, and the ference in social protection. Of the percentage of its national income. lowest-income countries (less 18.4 percent of GDP difference in The data show, however, that many than $20,000 per capita) are in total spending between France and richer countries actually choose darker red.5 As can be seen in the the United States, 16.7 percent is to spend a greater share of their table, seven of the eight countries social protection. The nine highest- income through government than in the high spending group are spending countries overall are also do poorer countries, making the wealthy countries. In the middle- the nine highest-spending in social choice to use the added income spending group, five of the eleven protection. Of the ten countries available to provide superior public are low-income countries and two in the low-spending group, nine services instead of leaving money in are middle income, but there are of them are in the bottom ten in the private sector for individuals or also four of the richest countries, social protection spending as well. businesses to spend on their own. including the Netherlands (fifth Thus, the single biggest explana- This policy reflects the assignment richest) and Germany (eighth). tion for what makes a country higher of a high value to services that are Half the low-spending group are spending or lower spending is the better provided through the com- low- or lowest-income coun- generosity the country shows to the bined efforts of the public, through tries, but there are also the high- portion of its population it determines the instrumentality of government, income United States (fourth) and in need of social protection. (Social relative to the value assigned to per- Switzerland (second), and middle- protection spending is discussed in sonal spending for personal benefit. income Japan (fourteenth), Israel more detail in the section looking at For example, wealthier countries (seventeenth), and the Czech specific areas of spending.) may decide that they value superior Republic (eighteenth). public education, universal health High- and low-spending coun- Government Spending care, high-quality public transpor- tries group together geographically Per Capita tation, public parks, drug counsel- even more than by national income ing, poverty reduction, or military levels. All in the highest-spending Spending as a share of countries’ dominance more than the types group are Western European. The economies is a good way to compare of purchases made individually, middle group comprises both their commitment to different types such as better or larger houses or Western and Eastern European of spending relative to their capacity automobiles. countries. The low-spending group to spend, but it is a poor measure There is not, however, a uni- includes European countries but for comparing the level of services, form relationship between the also all the non-European coun- goods, and investment provided by wealth of countries and the tries. It is perhaps not surpris- government. The 8.1 percent of its share of their national income ing that countries in geographic economy that Costa Rica spends they spend through govern- proximity, which are more likely to on education provides less fund- ment. In Table 1 countries are have shared cultural and economic ing for education—to pay teachers, color-coded by whether they pasts, would make similar choices buy books, build schools—than the are, by the standards of this in value-laden decisions regarding 4.6 percent that the richer United OECD collection of generally government spending. Kingdom spends. For that reason, in well-off countries, high-income, The specific spending that this section, we use a measure that middle-income, low-income, or drives these overall results tells us gets closer to comparing what resi- lowest-income. The high-income about the choices being made. The dents, collectively, actually receive. countries (per capita GDP exceed- strongest determinant of whether a That measure is per capita spending $44,000) are highlighted in country is low or high spending is ing in U.S. dollars (converting from CARSEY SCHOOL OF PUBLIC POLICY 5 native currencies using purchasing Analysis of Specific Areas look at several of the largest areas power parity), which is the amount of Spending of public spending and how the spent divided by the total population United States compares to other of the country. Per capita spending The prior sections have compared countries. does not, of course, measure perfectly U.S. spending to other countries Figure 2 shows the difference as the quality or quantity of service pro- overall and identified social pro- a share of GDP between what the vided—countries may operate more tection spending as the prime dif- United States spends and the aver- or less efficiently, may have different ferentiator of national government age for the twenty-nine countries requirements (more or fewer school- spending levels. In this section we for which data are available for age children per capita, for example), and may offer somewhat different TABLE 2. TOTAL PUBLIC SPENDING PER CAPITA, 2017 services—but it is a strong indicator. If a country is spending 30 percent less per person in a particular category than another country, it is most likely providing much less of that service than the country to which it is being compared. Table 2 divides the countries by this measure into high spending (more than $24,000 per capita), middle (from $15,000 to $24,000), and low (less than $15,000). Not surprisingly, by this measure richer countries are high spending and poorer countries are low spending, reflecting their different capacities and the fact that when a citizenry Note: High-income countries are in green, middle-income countries purple, lower-income countries red, lowest-income countries darker red. Source: OECD, “General Government Spending (Total),” 2019; and authors’ has higher income it wants higher calculations. quality in both what it buys privately and what it gets through FIGURE 2. U.S. GOVERNMENT SPENDING VERSUS 29-COUNTRY AVERAGE, AS government. The fact that the PERCENT OF GDP United States, for example, spends 19 percent more in dollars per capita than the average of the twenty- nine countries—which includes twenty-five countries that are lower income—is not surprising. Of high-income countries, however, the United States ranks low in this measure—eleventh of the thirteen, spending 14 percent less than the average. Again, social protection spending is a key driver: The United States ranks last among Source: OECD, “General Government Spending (All),” 2019; and authors’ calculations. high-income countries in that category and spends less than half the average for the group. 6 CARSEY SCHOOL OF PUBLIC POLICY

each of the spending categories. TABLE 3. SPENDING AS A SHARE OF GDP, U.S. VERSUS 29-COUNTRY AVERAGE Table 3 shows the levels of spending and where the United States ranks for each category. Most notable for the United States, in addition to the low level of spending on social protection, are the high levels of defense and health spending. These will be discussed more fully in the coming sections.6 Turning to the per-capita measure, Figure 3 and Table 4 show the percent more or less that U.S. governments spend per person

relative to all other countries in Source: OECD, “General Government Spending (All),” 2019; and authors’ calculations. the analysis and relative to just high-income countries. Even by FIGURE 3. PERCENT DIFFERENCE IN PER CAPITA SPENDING, U.S. VERSUS this measure that approximates 29-COUNTRY AVERAGE AND VERSUS OTHER HIGH-INCOME COUNTRIES the level of service provided, the United States spends 38 percent less per person in social protection than the average for all countries in the analysis (ranking twenty-first) and 57 percent less than the high- income countries (ranking last). Also notable is that, relative to the twenty-nine-country group, the United States spends: • Over three times more than the average per person on the military, trailing only Israel. Source: OECD, “General Government Spending (All),” 2019; and authors’ calculations. • Seventy percent more per person on public order and safety, TABLE 4. GOVERNMENT SPENDING PER CAPITA, U.S. VERSUS 29-COUNTRY making it the highest-spending AVERAGE AND VERSUS 13 HIGHEST-INCOME COUNTRIES country in this category. • Almost two times more per person on health, also placing it at number one. The following sections delve into each spending category in more detail.

Source: OECD, “General Government Spending (Categories),” 2019; and authors’ calculations. CARSEY SCHOOL OF PUBLIC POLICY 7

Social Protection TABLE 5. SOCIAL PROTECTION SPENDING AS SHARE OF GDP, 2017 By any measure, the United States spends a great deal less than other countries on social protection. Government programs that fall in this category include non-health programs to support those with low incomes and the elderly and disabled—essentially safety net and social insurance programs (for the OECD description, see the listing of category definitions in Appendix A). For example, in the United States, the Supplemental Nutrition Assistance Program (SNAP, for- merly known as food stamps) is included, as is Social Security. The Note: High-income countries are in green, middle-income countries purple, lower-income countries red, United States has long been less lowest-income countries darker red. Source: OECD, “General Government Spending (Social Protection),” 2017. fiscally committed to this area of spending than other economically TABLE 6. SOCIAL PROTECTION SPENDING PER CAPITA, 2017 advanced countries. Spending 7.6 percent of its GDP on social protection, the United States ranks at the very bottom of the category (Table 5). Finland at 24.9 percent and France at 24.3 percent are the highest-spending countries, while the median for all the countries is 16.2 percent. Social protection is the only spending category for which U.S. spending is greatly lower than other countries. The difference explains how the United States can spend so much more than other countries on its military and health services while still spending so much less than other countries overall. Note: High-income countries are in green, middle-income countries purple, lower-income countries red, On a per capita basis (Table 6), lowest-income countries darker red. Source: OECD, “General Government Spending (Social Protection),” 2019; and authors’ calculations. the United States spends a third less than the twenty-nine-country Note that the reason the United vary so much, OECD measures of average. The next lowest-spending States spends less on social pro- high-income country is the United poverty in 2017 show the United tection is not a lack of poverty. States having lower poverty than Kingdom, which spends almost 50 Although comparing poverty percent more per capita more than only Costa Rica and Israel of these across countries is complicated countries.7 United States. and controversial because percep- tions of acceptable living standards 8 CARSEY SCHOOL OF PUBLIC POLICY

Health FIGURE 4. GOVERNMENT HEALTH SPENDING AS A SHARE OF GDP, 2017 Governments in the United States spend more on health care than governments in any other country as both a share of GDP (Figure 4) and per capita. Indeed, despite being a low-spending country overall, the United States outspends the OECD average by 2.9 percent of GDP in health. In per capita terms, the country spends almost double the average. The higher spending by U.S. governments is consistent with the very high spending in the United States on health care overall. Counting both public and private health care, the United States spent 17.1 percent of its economy on health care in 2017 while the next closest country, Source: OECD, “General Government Spending (Health),” 2019; and authors’ calculations. Switzerland, spent 12.3 percent.8 The high amount U.S. governments Military both put limits on a country’s military spend reflects the high costs in the The United States is second only to prowess, but the threats to a coun- health market in general. Israel in military spending as a share try’s safety or territorial integrity are It is worth nothing that, in most of GDP and per capita; by the latter largely independent of the country’s countries, governments provide at measure it spends over three times economic or demographic size. In least basic health care for all resi- the twenty-nine-country average. addition, the effectiveness at project- dents while spending by U.S. govern- There is no particular relationship ing military power is a function of the ments is primarily for the elderly and between the wealth of the countries total amount spent rather than how lower income. A key objective of the being compared, or their region, much is spent relative to the size of Affordable Care Act (ACA) passed and how much they spend in the the economy or its population. The in 2010 was to bring U.S. health care defense category. Rather, different total amount the United States spends spending more in line with the rest countries spend more or less based on its military relative to other coun- of the world. Although health spend- on real or imagined threats, politi- tries—not as a share of its economy or ing slowed initially after passage, it cal imperatives, and their desire to per person—is why the United States has picked up in recent years. It is project power beyond their borders. is the dominant military power. still an open question whether, and While the amount spent per In terms of the total amount the extent to which the ACA, will person or as a share of national spent on the military, the United accomplish that goal (assuming it income measures how much of a States spends 50 percent more isn’t gutted or repealed) and what the burden defense spending creates for than all the other OECD countries impact on the quality of care would Americans, it does not give a good combined. Of course, since other be from constraining spending. measure of how calibrated the spend- OECD countries are not the most It is beyond the scope of this ing is to the objective. Compared to likely U.S. military adversaries, paper to delve into why health care the other categories of spending, there comparing spending to them is costs are so high in the United is a looser connection between the not revealing as to whether the States9 and why, for all this spend- purpose being served by the spending level of U.S. military spending ing, health outcomes are not better and either the size of the economy or is well-calibrated to its purpose. than in other countries.10 the size of the population. Obviously, CARSEY SCHOOL OF PUBLIC POLICY 9

Looking more globally, the United FIGURE 5. MILITARY EXPENDITURE BY COUNTRY, 2019 States spends more on its military than the next seven countries combined, OECD or not (Figure 5).11 Worldwide, the United States is responsible for 36 percent of global military spending, though it has 4.2 percent of the world population and accounts for 24 percent of the world’s economy.12 Education

The United States ranks sixth in Source: Stockholm International Peace Research Institute, SIPRI Military Expenditure Database, April 2019. education spending as a share of GDP relative to the twenty-nine countries for which detailed data Other Spending Categories (a) The government matches are available, 0.8 percent of GDP In the remaining categories for which the first $5,000 per year you above the average. It is not surpris- OECD data are available for the put into a retirement savings ing that the United States would United States and the other countries, account with a $5,000 gov- be somewhat above the level of the United States ranks twenty-third ernment check into the same account. other countries, as it has a higher as a share of GDP in economic affairs, share of its population age 5–19 a low ranking that may be due to its (b) The government allows you (20 percent) than every country of preference for handling these mat- a 50 percent refundable tax the twenty-nine except for Israel, ters through the tax code (see the tax credit on the first $10,000 you 13 Colombia, and Costa Rica. Costa expenditure section below); twelfth in put into retirement savings.16 Rica and Israel lead this category in general public services; ninth in hous- A real-life example of a tax spending as a share of GDP. ing and community amenities; and expenditure in the United States is Public Order and Safety twenty-eighth in recreation, culture, the subsidy for buying a home by and religion (a small share of spend- allowing a tax deduction for interest This category—law enforcement ing across the countries).15 and fire protection being the paid on home mortgages. There are also large tax breaks for retirement most prominent elements— Tax Expenditures is another area where the United savings, health insurance, and many States stands out, ranking first To all of the above there is a caveat. other areas where incentives have per capita and eighth as a share Governments achieve their spending been deemed to be of societal value. of GDP. In contrast to other objectives not just by means of spend- There are a number of reasons categories, the countries ranking ing money as conventionally con- countries choose to achieve spend- above the United States in spend- ceived and tabulated by the OECD ing goals through the tax system ing as a share of GDP are pre- data used here—essentially by writing instead of direct spending, including dominantly Eastern European or checks. Governments also achieve political motivations and simplicity Latin American and all are lower spending goals by giving targeted tax of administration. There are benefits income. The United States impris- breaks, and those “tax expenditures” and drawbacks to both approaches—a ons far more people than other are not reflected in these data. strong criticism of tax expenditures countries, which contributes to Functionally, both approaches in the income tax is that they tend to the country’s high ranking.14 are the same. Let’s say a govern- benefit higher-income people more 17 ment wants to subsidize saving for than lower-income people. The U.S. retirement. There is no difference mortgage interest deduction is an between the following two policies: example of this. If one were designing a program to assist people in home 10 CARSEY SCHOOL OF PUBLIC POLICY

ownership from scratch it is unlikely are not so high, relative to direct have been generally lowered in lieu that one would design it in a way that spending, to change the fact that of more targeted tax breaks. helped those with higher incomes the United States is a relatively low- It appears, therefore, that while more than lower incomes—includ- spending country, at least relative the choice of using tax expenditures ing many who have no need for a to these six other countries. instead of direct spending for some tax break to facilitate owning their The four areas that contribute purposes makes the United States own homes. Yet the interaction of the most substantially to the United look like a lower-spending country mortgage interest deduction with the States being high in tax expenditures, than it really is, it is still a low-spend- rest of the structure of the personal relative to five of the other countries, ing country. The 2010 study suggests income tax does exactly that. are health, housing, intergovernmen- this, and changes in U.S. law have Does excluding tax expenditures tal relations, and retirement. U.S. tax probably reduced the impact of tax from our analysis distort our find- expenditures on housing and inter- expenditures since then.20 ings? The lack of readily available governmental relations are almost data makes it a difficult question to certainly lower now than in the year Taxes answer with precision. The OECD that the study examined (2008 for did release a study in 2010 that the United States), for a number of Lastly, we briefly compare the tax compared the levels of tax expendi- reasons. First, because of changes levels available from OECD data. tures among seven OECD coun- in tax law fewer people are itemiz- Not surprisingly the United States tries, including the United States.18 ing deductions on their tax returns, ranked low in taxes—twenty-ninth The author cautioned against draw- reducing the use of the home mort- of the thirty-three countries for ing too strong conclusions from gage interest deduction.19 The tax law which data are available—with only his analysis, given the challenges also capped the deduction for state Korea, Turkey, Chile, and Mexico he faced in reconciling different and local taxes, reducing tax expen- ranking lower (Figure 6). The United countries’ systems (or lack thereof) ditures in the intergovernmental States may move even further down of accounting for tax expenditures. relations category. The United States the list after the tax law changes of Nevertheless, for broad compari- is also generous with business-related 2017 become evident in the data. sons it is helpful. tax expenditures, but these are also The lower ranking in taxes than in Of the seven countries—Canada, likely lower now as business taxes spending reflects a relatively high Germany, Korea, the Netherlands, level of government borrowing. Spain, the United Kingdom, and the United States—the United FIGURE 6. TAXES AS A SHARE OF GDP, 2017 States was second to the United Kingdom in the level of income tax expenditures as a share of GDP, at 5.97 percent. The average for the group was 3.65 percent. In terms of direct spending, among these seven countries the United States ranked sixth. Adding tax expenditures to direct spending, for a measure of total spending for all seven countries, the United States still ranked sixth. In other words, although the United States has high tax expenditures, they Source: OECD (2019), Tax revenue (Total); and authors’ calculations. CARSEY SCHOOL OF PUBLIC POLICY 11

Conclusion Endnotes 7. “Poverty Rate (Total)” (Paris: 1. Ranking is for 2017. Note that Ireland OECD, 2019), doi: 10.1787/0fe1315d- The differences between how much is excluded from the ranking because of en, https://data.oecd.org/inequality/ countries spend are a reflection problems related to the measurement poverty-rate.htm. The OECD measure of both intentional choices and of its GDP arising from tax avoidance sets the poverty line at half the accidental inefficiencies. There schemes in which multinational median household income of the total is no single right answer. But the corporations shift ownership of their population for the country. This is, thus, bottom line of this analysis is that intellectual property to their Irish a relative measure. Many of those who the United States is a low-tax, low- subsidiaries. See “Irish GDP Up by are classified as poor in the United States 26.3% in 2015?” (Paris: Organization by this measure have a higher quality of spending country relative to the for Economic Cooperation and life than some who are above poverty in other countries examined, particu- Development (OECD), 2016), http:// countries that are significantly poorer larly when compared to its fellow www.oecd.org/sdd/na/Irish-GDP-up- overall than the Unites States. higher-income countries. in-2015-OECD.pdf. 8. OECD Stat, “Global Health What is true of the overall level 2. This ranking includes countries for Expenditures Database” (Paris: of spending is true of the levels for which comparisons are not otherwise OECD, 2019), https://stats.oecd.org/ each spending category—decisions made in this report because, for other BrandedView.aspx?oecd_bv_id=health- are made and accidents happen. The purposes, they are not meaningfully data-en&doi=data-00349-en. United States has been, rightly or comparable. Switzerland also has a system that is wrongly, intentional in its high level 3. “Report of the Chair of the ostensibly even more private than that of spending on the military and low Working Group on the Future Size of the United States, but it is highly spending on social programs. How and Membership of the Organisation regulated: all residents are required the United States has ended up with to Council: Framework for the to have private insurance, coverage offered by that insurance is set by the the highest level of government Consideration of Prospective Members” (Paris: OECD, 2017), government, and insurance providers’ spending on health care, and yet is profits on basic insurance are restricted. among its most fractured providers, http://www.oecd.org/mcm/ 9. “U.S. Health Care Spending Highest is a much more complicated story. documents/C-MIN-2017-13-EN.pdf. Among Developed Countries” (Baltimore, The purpose of this report is not 4. The breakpoints for high, middle, and low are somewhat arbitrary. They MD: Bloomberg School of Public Health, to decide these issues but to put Johns Hopkins, 2019), https://www.jhsph. them on the table in the context of are chosen at points of discontinuity in the spending levels of countries and edu/news/news-releases/2019/us-health- worldwide norms. No country has roughly divide the list in thirds. care-spending-highest-among-developed- a monopoly on fiscal wisdom, but countries.html. 5. The breakpoints for high, middle, looking to other countries offers the 10. OECD Stat, “Health Status: opportunity to assess the merits of low, and lowest income are somewhat arbitrary. They are chosen at points Mortality, Life Expectancy” (Paris: different approaches to address the of larger gaps in the incomes between OECD, 2019), https://stats.oecd.org/ shared and distinctive challenges countries and split the list into four BrandedView.aspx?oecd_bv_id=health- that all countries face. groups. Note also that, by global data-en&doi=data-00349-en. standards, these twenty-nine countries 11. “U.S. Defense Spending Compared are skewed heavily toward wealthier to Other Countries,” Peter G. Peterson countries. Thus, “middle-income” Foundation, 2019; “World Military countries on this list are quite high Expenditure Grows to $1.8 Trillion in income by global standards. 2018” (Stockholm International Peace 6. Note that the OECD environmental Research Institute, 2019), https://www. protection spending category is sipri.org/media/press-release/2019/ excluded from Figures 2 and 3 and world-military-expenditure-grows-18- Tables 3 and 4. U.S. spending in this trillion-2018. category is not reported separately and 12. Databank, “GDP” (Washington is included in other categories. Among DC: World Bank, 2019), https://data. the twenty-nine countries median worldbank.org/indicator/ny.gdp.mktp.cd. spending on environmental protection is low, at 0.6 percent of GDP. 12 CARSEY SCHOOL OF PUBLIC POLICY

13. “Demographic Yearbook 2017” (New York: NY: United Nations, Appendix A 2018), https://unstats.un.org/unsd/ demographic-social/products/dyb/ OECD Spending Categories dyb_2017/. 14. “World Prison Brief” (London: Defense: Military defense; civil defense; foreign military aid; research Institute for Criminal Policy Research, and development (R&D) in defense; and defense not elsewhere 2019), https://www.prisonstudies.org. classified (n.e.c.). 15. The low ranking in the last “catch- Economic affairs: General economic, commercial, and labor affairs; all” category could be due to a variety agriculture, forestry, fishing, and hunting; fuel and energy; mining, of factors, including the separation of church and state in the United States and manufacturing, and construction; transport; communication; other the low level of government spending on industries; R&D in economic affairs; and economic affairs n.e.c. public broadcasting and sports. Education: Pre-primary and primary education; secondary education; 16. A refundable tax credit is one in post-secondary nontertiary education; tertiary education; education which the excess of the credit amount not definable by level; subsidiary services to education; R&D in over the pre-credit tax liability is paid education; and education n.e.c. out to the taxpayer. Environmental protection: Waste management; waste water 17. Jacob Funk Kirkegaard, “The True management; pollution abatement; protection of biodiversity and Levels of Government and Social Expenditures in Advanced Economies” landscape; R&D in environmental protection; and environmental (New York, NY: Peterson Institute for protection n.e.c. International Economics, 2015), https:// General public services: Executive and legislative organs, financial www.piie.com/publications/pb/pb15-4. and fiscal affairs, external affairs; foreign economic aid; general pdf; “2010 Tax Expenditures in OECD services; basic research; R&D in general public services; general public Countries (Paris: OECD, 2010), https:// services n.e.c; public debt transactions. doi.org/10.1787/9789264076907-2-en. 18. “2010 Tax Expenditures in OECD Health: Medical products, appliances, and equipment; outpatient Countries (Paris: OECD, 2010). services; hospital services; public health services; R&D in health; and 19. Jim Tankersley and Ben Casselman, health n.e.c. “As Mortgage Interest Deduction Housing and community amenities: Housing development; Vanishes, Housing Market Shrugs” New community development; water supply; street lighting; R&D in York Times, Aug. 4, 2019, https://www. housing and community amenities; housing and community amenities nytimes.com/2019/08/04/business/ n.e.c. economy/mortgage-interest-deduction- tax.html. Public order and safety: Police services; fire-protection services; law 20. Note that there is another OECD courts; prisons; R&D in public order and safety; and public order and resource on tax expenditures—the safety n.e.c. Social Expenditure Database—but, for Recreation, culture, and religion: Recreational and sporting services; reasons explained in Appendix B, it cultural services; broadcasting and publishing services; religious and does not afford useful information to other community services; R&D in recreation, culture, and religion; the question at hand. recreation, culture, and religion n.e.c. 21. “OECD Social and Welfare Statistics” (Paris: OECD, no date), Social protection: Sickness and disability; old age; survivors; family https://doi.org/10.1787/socwel-data-en. and children; unemployment; housing; social exclusion n.e.c.; R&D in social protection; and social protection n.e.c. Source: “Government at a Glance 2017,” Annex C, OECD 2017. CARSEY SCHOOL OF PUBLIC POLICY 13

Appendix B

Options for analyzing international tax expenditures Another relevant OECD resource on tax expenditures is its Social Expenditure Database.21 This database contains data on social spending in both the public and private sectors and includes the effects of tax expenditures. It shows the United States as ranking higher on a broad measure of social spending (broader than the social protection category used here) than our analysis, but for what we are examining here it is not useful: the data are presented in a way that do not allow isolating government spending (inclusive of tax expenditures) from private spending for social purposes (e.g., private health insurance) or the impact of taxes on public benefits (e.g., income tax imposed on social security payments). The purpose of our report is to examine how governments differently spend the funds they raise from taxes, fees, and borrowing. Thus, this database is not helpful. As an aside, we are unconvinced that netting out taxes paid on social benefits or including private spending for social purposes is the most illuminating way to examine social spending. Because the United States does not have broad-based consumption taxes offsetting social spending, it ranks higher in this database. But residents of other countries are getting public services beyond social spending in return for those higher taxes, so they are not necessarily receiving reduced benefits because of taxes, as the database implies. With respect to including private social spending, including some forms of compensation (employer provided health insurance, for example) also makes the United States rank higher in this database, but the provision of these benefits privately instead of through government has important impli- cations, especially on how those benefits are distributed, that should not be blurred. And attributing some forms of compensation to social purposes but not others (e.g., wages that pay for food, clothing, and shelter) has an arbitrary quality to it. See Jacob Funk Kirkegaard, “The True Levels of Government and Social Expenditures in Advanced Economies” (New York, NY: Peterson Institute for International Economics, 2015) for a somewhat different perspective. 14 CARSEY SCHOOL OF PUBLIC POLICY

About the Authors Acknowledgments

Michael Ettlinger is the founding director of the Carsey This paper is an evolution of a previous work, School of Public Policy. “Comparing Public Spending and Priorities Across OECD Countries,” by Sabina Dewan and Michael Jordan Hensley is a policy analyst at the Carsey School Ettlinger (Washington DC: Center for American of Public Policy. Progress, 2009), https://www.americanprogress.org/ Julia Vieira is a Master in Public Policy student at the issues/economy/reports/2009/10/02/6746/comparing- Carsey School of Public Policy. public-spending-and-priorities-across-oecd-countries/. The authors would like to thank that report’s co-author, Sabina Dewan, for her contribution to the analytic framework of this brief.

University of New Hampshire Carsey School of Public Policy

The Carsey School of Public Policy at the University of New Hampshire is nationally recognized for its research, policy education, and engagement. The school takes on the pressing issues of the twenty-first century, striving for innovative, responsive, and equitable solutions. Huddleston Hall • 73 Main Street • Durham, NH 03824 (603) 862-2821 TTY Users: dial 7-1-1 or 1-800-735-2964 (Relay N.H.) carsey.unh.edu