Social programs

Lane Kenworthy, The Good Society
April 2021

Public social programs have existed in their present form for only a century or so, yet they are among the core institutions of modern society.1 What are these programs? What do they do? How have they developed and changed over time? What similarities and differences do we observe across the world’s rich longstanding-democratic countries?

Skip to:

WHAT DO GOVERNMENT SOCIAL PROGRAMS DO?

It’s only a slight exaggeration to say that government in the world’s affluent nations is essentially an insurance company with an army.2 Public social programs are, for the most part, insurance programs, and they are much of what modern governments do.3

Risk is pervasive in life. Prominent sources include loss of a job, accident, illness, old age, disability, family needs, neighborhood dysfunction, discrimination, and low wages, among others.4 Ideally, we want to minimize the incidence of risk events and minimize the impact when we fall victim to one.

As individuals, we try to steer clear of risk events. But it’s difficult to do so. One in five American children lives in a household with income below the poverty line (about $24,000 for a family of four). More than 90% of Americans end up unemployed at some point during their working life, 80% more than once. About 10% of prime-working-age Americans are diagnosed with cancer, and 10% of males with cardiovascular disease. About 20% of Americans have a disability, and between 5% and 10% are rendered permanently unable to work at some point. One in ten 25-year-olds won’t live to age 65, often leaving behind a family dependent on their income. Most of us at one time or another fall victim to illness, accidents, and family troubles.5

Since we can’t escape risk, we attempt to insure ourselves against potential harm or loss in the event we do get hit by one. We save money, for example, in case we lose our job or outlive our working years. Families are another risk-alleviation mechanism: binding ourselves in long-term family partnerships (marriage) helps to spread childcare duties and safeguard against financial difficulty. Voluntary organizations — community groups, charities, and more — provide food, housing, and other types of assistance to people in need. Markets and firms are another source of risk mitigation: they enable us to purchase health insurance in case we need expensive medical care, auto insurance in case we get in an accident, and homeowners insurance in case our house is robbed, damaged, or destroyed.

Some types of risk are most effectively and/or efficiently addressed via a government program. Consider income in old age. Many individuals don’t save enough on their own, because they don’t feel they can afford to given their current incomes, because the consequences are far in the future, and because they have no clue how long they’ll live. Others may save enough but make poor investment choices. Employers can provide a pension to their workers, but some people have irregular paid work or are self-employed or work for a small firm that doesn’t offer a pension. Families can help by adding an additional earner, but not everyone gets or stays married, and couples face the same uncertainty and resource constraints as individuals. People could form voluntary savings clubs to provide encouragement and advice about saving and investing, but in practice few do. Charities and community groups may help to support retirees who have inadequate income, but historically they have proved inadequate to the task. Their efforts can be immensely helpful, but their limited size and scope means they can’t cover everyone who needs assistance. A solution is to create a government pension program that collects money from firms, workers, and/or taxpayers and distributes the revenue to retirees. Program details vary, but the basic idea is simple and quite effective.

A similar logic applies to some other risks. Unemployment compensation insures against the risk of losing your job. Disability payment programs insure against the risk of suffering a physical, mental, or psychological condition that renders you unable to earn a living. Other public services and benefits also are insurance programs, even if we don’t usually think of them as such. Public schools insure against the risk that private schools are unavailable, too expensive, or poor in quality. Special education services insure against the risk of having a disability that inhibits participation in school. Retraining and job placement programs insure against the risk that market conditions make it difficult to find employment. The Earned Income Tax Credit (EITC) insures against the risk that your job pays less than what’s needed for a minimally decent standard of living. Social assistance programs such as SNAP (Food Stamps) and TANF (Temporary Assistance for Needy Families) insure against the risk that you find yourself unable to get a job but ineligible for unemployment or disability compensation. Affirmative action programs insure against the risk of being in a group that is, or formerly was, discriminated against.

Government social programs are created and expanded for three main reasons. First, people like insurance. As a person’s income or assets increase, she will tend to spend more on insurance. The same is true of nations; as they get richer, they tend to allocate a larger portion of their income — their gross domestic product (GDP) — to insurance.

Second, people care about fairness, and public social programs enhance fairness, particularly by helping to equalize opportunity. Here too we see an income effect: as societies get richer, their members become more willing to support programs that enhance fairness, even if they don’t anticipate that they themselves will benefit from such programs.

Third, in many instances government social programs work well. They can ensure comprehensive coverage in a way private efforts can’t. They can reduce administrative costs and impose cost controls on service providers. They can spread the cost across a larger pool (all citizens), thereby reducing it for each individual. They can enhance stability, as participants need not worry that their insurer will go out of business.

Given these considerations, we should expect to see an expansion of public social programs and an increase in their generosity as countries get richer. At the same time, nations differ in a variety of ways — political institutions, economic structures, organization of interest groups, public opinion, and more — that are likely to produce differences in the amount and types of social programs.

THE RISE AND PERSISTENCE OF THE WELFARE STATE

Large-scale public insurance programs originated in the late 1800s, became commonplace in the 1930s, and were significantly expanded over the ensuing half century.6 Since the mid-1980s there has been a good bit of change in social programs, but limited increase in their overall size and generosity. Figure 1 shows this long-run pattern in the world’s rich longstanding democratic countries. It uses the most common measure of social program generosity — how much money a country spends on these programs, expressed as a share of GDP. The data sources differ for different periods, so the trend lines may not be perfectly comparable, but they very likely give an accurate picture of over-time developments.

Figure 1. Public social programs
Share of GDP. Data source: Esteban Ortiz-Ospina and Max Roser, “Public Spending,” Our World in Data, using data for 1880-1930 from Peter Lindert, Growing Public, volume 1, Cambridge University Press, 2004, data for 1960-1979 from OECD, “Social Expenditure 1960-1990: Problems of Growth and Control,” OECD Social Policy Studies, 1985, and data for 1980ff from OECD, Social Expenditures Database. “Asl” is Australia; “Aus” is Austria.

Let’s take a closer look at the over-time story. The Social Policy Indicators Database (SPIN) from the Swedish Institute for Social Research at Stockholm University has reliable estimates of social program generosity beginning in 1930. Figure 2 shows a measure of benefit generosity for public pensions in eighteen rich nations. The point here isn’t to emphasize pensions per se, but rather to give a sense of the long-run pattern for a representative public insurance program. In most of the world’s rich countries public pensions were created as a national program during or shortly after the Great Depression in the 1930s. They were made steadily more generous through the mid-1980s. Since then their generosity has leveled off and even decreased slightly.

Figure 2. Generosity of a typical public insurance program
Public pension average net replacement rate for a single person or couple who earned an average production worker’s wage. The countries are Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, New Zealand, Norway, Sweden, Switzerland, the United Kingdom, and the United States. Data source: Social Insurance Entitlements Dataset (SIED), provided as part of the Social Policy Indicators Database (SPIN), version June 2015, series px2indst.

Data for a larger number of programs are available beginning in 1960. Simon Birnbaum, Tommy Ferrarini, Kenneth Nelson, and Joakim Palme use the SPIN data to create an indicator of the generosity of public insurance programs aimed at compensating for risks during childhood, the working-age years, and old age. The measure, like that shown in figure 2, is the average replacement rate — the benefit amount as a percentage of the person’s previous earnings. This is affected by eligibility criteria (some people receive no benefit at all), benefit level, and duration of receipt. In figure 3 we see a similar over-time story. Program generosity increased steadily up to the mid-1980s. In some countries it continued to rise beyond that point, but in many it flattened out and in a few it decreased.

Figure 3. Public transfer program generosity
Average replacement rate for public insurance programs aimed at risks during childhood (child allowances/tax credits and paid parental/maternity leave), working age (unemployment and sickness), and old age (pensions). The countries are Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, New Zealand, Norway, Sweden, Switzerland, the United Kingdom, and the United States. Data source: Simon Birnbaum, Tommy Ferrarini, Kenneth Nelson, and Joakim Palme, The Generational Social Contract, Edward Elgar, 2017, using data from the Social Policy Indicators (SPIN) database.

A related measure of social program generosity comes from Lyle Scruggs, Detlef Jahn, and Kati Kuitto. Their measure covers three programs: pensions, unemployment insurance, and sickness insurance. For each country they score the size of the benefit, its duration, and the share of the relevant group that is covered. They then combine these into an overall measure of program generosity. They use these to calculate an average measure of generosity across the three programs. These data, available beginning in 1971, are shown in figure 4. They too suggest a rise in generosity through the mid-1980s, followed by a flattening out.

Figure 4. Generosity of public pensions, unemployment compensation, and sickness compensation
The generosity measure is an average for the three programs. It includes program eligibility criteria, coverage, waiting period, benefit replacement rate, and benefit duration. The countries are Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, New Zealand, Norway, Sweden, the United Kingdom, and the United States. Data source: Lyle Scruggs, Detlef Jahn, and Kati Kuitto, Comparative Welfare Entitlements Data Set 2, version 2014‐03, series totgen.

Finally, we can return to spending on public social programs. This is a more comprehensive measure than those shown in figures 2-4 in that it covers all social programs, including services. It includes government spending on benefits and services related to old age, unemployment, active labor market, health, incapacity, family, housing, and a variety of other programs. (We might reasonably add education, but it usually isn’t included in measures of social programs.7) This measure’s drawback is that it is influenced by not just program generosity but also need — the share of the population that is elderly, the share that is unemployed, and so on. Also, reliably comparable data don’t begin until 1980. Figure 5 shows that in most countries public social spending has increased modestly since the mid-1980s.

Figure 5. Public expenditures on social programs
Share of GDP. Gross expenditures. Includes spending on old age, unemployment, active labor market, health, incapacity, housing, family, and other social programs. The countries are Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States. Data source: OECD, Social Expenditure Database.

DIFFERENCES ACROSS COUNTRIES

Broadly speaking, welfare states in the rich democratic nations have followed a common trajectory, growing fairly rapidly from 1930 to around 1985 and then more slowly since then. But that isn’t to say that these countries have ended up in the same place. They have long differed in the expansiveness and generosity of social programs, and that difference has not disappeared in recent decades.

Figure 6 shows a measure of overall benefit generosity: the average replacement rate for programs aimed at compensating for risks during childhood, during the working-age years, and during old age (the same as in figure 3). The Nordic and continental European countries tend to be on the high end and the English-speaking nations, including the United States, on the low end.

Figure 6. Public transfer program generosity
Average replacement rate for public insurance programs aimed at risks during childhood (child allowances/tax credits and paid parental/maternity leave), working age (unemployment and sickness), and old age (pensions). Data source: Simon Birnbaum, Tommy Ferrarini, Kenneth Nelson, and Joakim Palme, The Generational Social Contract, Edward Elgar, 2017, using data from the Social Policy Indicators (SPIN) database. “Asl” is Australia; “Aus” is Austria.

Figure 7 shows public expenditures on social programs (the same measure as in figure 5). Here too the Nordic and continental countries top the list while the English-speaking nations, along with Switzerland and South Korea, are at the bottom.

Figure 7. Public expenditures on social programs
Share of GDP. Gross expenditures. Includes spending on old age, unemployment, active labor market, health, incapacity, housing, family, and other social programs. Data source: OECD, Social Expenditure Database. Thick line: United States. “Asl” is Australia; “Aus” is Austria.

The United States is commonly considered a laggard in government social programs. It has fewer public programs (no national sickness insurance or paid parental leave). Some of the programs it has aren’t especially generous. It relies less on universal programs available to all citizens and more on means-tested programs available only to individuals or households with low income. It encourages employment but provides limited support for those who struggle in the labor market. America’s laggard status isn’t surprising, given the country’s position with respect to many of the key causes of public social program generosity: citizen antipathy to big government, weak labor unions, no true left political party, lots of veto points in the political system.8

Yet this is only part of the story. In the United States a sizable amount of money is spent on private social programs. The biggest of these are firms’ expenditures on health insurance and pensions for their employees. Also, the large spending levels in some affluent countries are misleading because the benefits that are paid to individuals or households are taxable, so some of the money goes right back to the government. Moreover, households spend some of their benefits on items that are subject to consumption taxes, which means even more of the money returns to the government. A more complete measure of social spending is therefore public net (rather than gross) expenditures plus private expenditures.9

These data are shown on the vertical axis in figure 8, with the more familiar public gross spending data on the horizontal axis.10 There is some shifting of positions as we move from the horizontal axis measure to the vertical axis on this graph. For instance, Denmark and Sweden drop down a bit in the ranking. That’s because they are among the most aggressive in “clawing back” cash benefits by subjecting those benefits to income tax and by imposing heavy consumption taxes.

The United States stands out as the country for which the choice of measure makes the biggest difference. Public gross social expenditures total just 19% of GDP in the US, putting it near the low end among the rich nations. America’s total for public net plus private social expenditures, by contrast, is 29% of GDP, and it’s one of the top spenders according to this measure. The difference owes a little to America’s lesser taxation of benefits, but it’s mainly a function of its large quantity of private social spending, which far exceeds that in any of the other nations.

Figure 8. Social program expenditures
Share of GDP. 2017. Includes spending on old age, unemployment, active labor market, health, incapacity, housing, family, and other social programs. Data source: OECD, Social Expenditure Database. “Asl” = Australia; “Aus” = Austria.

So is America a laggard? If we’re interested in total social spending, the answer is no. If we focus on public social spending, the answer is yes.

The distinction matters. Public social programs tend to benefit either the entire population or the most needy. Private programs disproportionately benefit people with higher incomes, because they are provided mainly to persons employed by large firms. Public programs thus tend to be more effective in mitigating risk and improving economic security.11

Is the US welfare state categorically distinct from others, as Gøsta Esping-Andersen suggested in his 1990 book The Three Worlds of Welfare Capitalism? Or does it lie at one end of a continuum, different in degree rather than in kind? Researchers continue to debate this question, and both perspectives have merit.12 But the categorical perspective leads many observers to miss an important point: While the US welfare state lags behind those of the leading countries when it comes to the size and scope of public social programs, it has been advancing for most of the past century. Countries such as Denmark and Sweden spend significantly more on public social programs, but the difference between these countries and the United States today is much smaller than the difference between the US today and the US a century ago (figure 1 above).

TYPES OF PROGRAMS

Let’s turn from the aggregate picture to specific types of social programs.13 The OECD, which is the source of the expenditure data for social policies, organizes them into eight categories: old age, unemployment, active labor market, health, incapacity, housing, family, and “other.” The data are available beginning in 1980. I’ll focus on public gross expenditures, since for most countries these give a pretty accurate portrait of total spending, and also because public net spending data aren’t available for most programs in most countries.

Income in our retirement years can come from a variety of sources, most notably personal savings, an employer pension, or sale of assets (such as a home). But for many people these aren’t sufficient to ensure adequate resources, so all rich nations have one or more public programs to mitigate poverty in old age. Public pensions come in various forms. Earnings-based programs pay an amount proportional to average or peak earnings during the work career. Typically these guarantee a benefit amount (“defined benefit”), but sometimes only the contribution amount is predetermined, with the benefit determined by both contribution and performance of the investment (“defined contribution”). “Basic” pension schemes pay a flat rate or an amount determined by the number of years of work. “Targeted” programs adjust the payment based on the person’s income or assets in retirement. “Minimum” pensions adjust based on other sources of pension income, rather than on total income or assets. Many countries combine two or more of these programs.14

Public pensions are funded mainly via social contributions or payroll taxes earmarked for this program. In the United States, the Social Security program is funded by a tax of 12.4% on an employee’s wage, with half paid by the employee and half by the employer. In some countries the contribution is considerably larger, totaling more than 40% of the wage.

Figure 9 shows public expenditures on old-age programs as a share of GDP. Much of this is on pensions, but this category also includes spending on services for the elderly such as rehabilitation and home-help. It also includes cash benefits and/or services to various categories of survivors. (Payments to the widows of dead soldiers were one of the earliest public assistance programs in many rich nations.) The biggest spenders on old-age programs are the continental and southern European countries, including Italy, France, Austria, and Portugal. The United States is toward the low end, but it isn’t particularly exceptional.

Old-age programs are one of the two biggest-spending social program categories. As of the most recent year of data, the amount ranges as high as 16% of GDP among the countries shown in figure 9. In many countries, old-age program expenditures have increased since the mid-1980s. In most, however, this rise isn’t a result of rising program generosity. Instead, it’s a function of population aging — a growing share of the population is age 65 or older. Figure 10, which is based on program details rather than the quantity of expenditures, confirms that public pension generosity has been largely flat since around 1980 (see also figure 2).

Figure 9. Public expenditures on old-age programs
Share of GDP. Cash expenditures on old-age and survivors pensions plus expenditures on services for the elderly such as day care, rehabilitation, in-home help, residential care in an institution, and others. Data source: OECD, Social Expenditure Database. Thick line: United States. “Asl” = Australia; “Aus” = Austria.

Figure 10. Generosity of public pensions
Includes program eligibility criteria, coverage, waiting period, benefit replacement rate, and benefit duration. Data source: Lyle Scruggs, Detlef Jahn, and Kati Kuitto, Comparative Welfare Entitlements Data Set 2, version 2014‐03, series pgen. “Asl” = Australia; “Aus” = Austria.

Unemployment insurance provides temporary cash assistance to people who have lost a job involuntarily. Access to the program usually hinges on fulfilling criteria such as minimum length of employment and minimum earnings level. In the United States, only about 40% of unemployed persons qualify for unemployment benefits.

The benefit level typically is proportional to prior earnings. Replacement rates have varied from as low as 10% to as high as 90% in the rich nations. In the United States, they have averaged around 60% over the past generation.

Duration of benefit receipt also varies widely. In the US, the formal limit is 26 weeks. In states where the unemployment situation has worsened significantly, this is extended by 13 or 20 weeks, and during national recessions the federal government typically extends it to a full year. In some European countries, eligibility in the past has been in effect unlimited in duration. In recent decades many countries have moved to shorten this.

The funds for unemployment insurance usually come from social contributions or payroll taxes or from general revenues. In the United States, employers pay a dedicated tax to the federal and state governments, and state governments administer the program. In some countries the unemployment benefit system is administered by unions and employers rather than by the government. In four of these — Belgium, Denmark, Finland, and Sweden — access to the system requires being a member of a union.

Public expenditures on unemployment compensation are shown in figure 11. They’ve typically accounted for less than 2% of GDP in most nations, and less than 1% in many. Unemployment spending tends to move with the economy — falling during growth periods, rising during recessions. If we ignore these fluctuations, the trend in most countries has been flat.

(The expenditure charts in this section all use the same range of values on the vertical axis in order to facilitate comparison across policies.)

Figure 11. Public expenditures on unemployment compensation
Share of GDP. Expenditure on cash benefits for people to compensate for unemployment. Data source: OECD, Social Expenditures Database. Thick line: United States. “Asl” = Australia; “Aus” = Austria.

The flat over-time pattern also is in evidence if we use a measure based on program details, shown in figure 12. Noteworthy here is that the United States is in the middle of the pack, rather than at the low end. That’s because the unemployment insurance eligibility criteria and replacement rate in the US are average relative to other rich nations. The stinginess of the US program lies mainly in its short duration. The low expenditure level in the United States (figure 11) also owes to the fact that many Americans find a new job comparatively rapidly.15

Figure 12. Generosity of unemployment compensation
Includes program eligibility criteria, coverage, waiting period, benefit replacement rate, and benefit duration. Data source: Lyle Scruggs, Detlef Jahn and Kati Kuitto, Comparative Welfare Entitlements Data Set 2, version 2014‐03, cwed2.org, series ugen. “Asl” = Australia; “Aus” = Austria.

To help people who are unemployed or out of work for other reasons, rich nations offer more than just monetary compensation. They also provide a range of services such as retraining, help with job placement, assistance in moving to a new location, and special services for particularly vulnerable groups such as young persons and the disabled. These services are often referred to as “active labor market” programs. Research on the effectiveness of retraining and job placement assistance has yielded mixed conclusions, but a recent meta-analysis finds generally positive estimates of medium-term impact on likelihood of employment.16

Figure 13 shows that public expenditures on active labor market programs tend to be relatively modest, though Denmark, Sweden, and Belgium spend 1% to 2% of GDP on them. Despite the centrality of paid work in American culture, the US is at the low end of the spectrum in the resources it devotes to these kinds of employment supports.17

Figure 13. Public expenditures on active labor market programs
Share of GDP. Includes training, job placement assistance, hiring subsidies, and direct job creations in the public sector. Data source: OECD, Social Expenditures Database, calculated as total expenditures on labor market programs minus expenditures on out-of-work income. Thick line: United States. “Asl” = Australia; “Aus” = Austria.

After old-age programs, health care is the largest category of public social expenditure, accounting on average for 7% of GDP as of the most recent year of available data. Figure 14 shows the country patterns. Health also is the category in which spending has increased most in recent decades. (For more detail, see “Health Care.”)

Figure 14. Public expenditures on health
Share of GDP. Data source: OECD, Social Expenditures Database. Thick line: United States. “Asl” = Australia; “Aus” = Austria.

We can see the rise in health spending even more clearly if private expenditures are included, as in figure 15. Here the United States has steadily pulled away from all other affluent democracies since around 1980, and it now stands quite apart. Public spending on health is relatively high in the US, and when we add private spending the US total far exceeds that of every other country.

Figure 15. Total (public and private) expenditures on health
Share of GDP. Gross expenditures. Data source: OECD. “Asl” = Australia; “Aus” = Austria.

A related but separate category is for incapacity due to disability, sickness, or occupational injury. Figure 16 shows that the Nordic countries are the biggest spenders in this area, at 4-5% of GDP. The United States is at the low end, with expenditures totaling about 1% of GDP. That’s partly because the US has no public sickness insurance (paid sick leave) program, as figure 17 indicates. Incapacity programs have seen little change in generosity over the past generation. The lone noteworthy exception is the Netherlands; its disability program, which was exceptionally generous in the 1980s, has been scaled back in recent decades.

Figure 16. Public expenditures on incapacity (disability, sickness, occupational injury) programs
Share of GDP. Public spending on incapacity refers to spending due to sickness, disability, and occupational injury. It includes disability cash benefits that are comprised of cash payments on account of complete or partial inability to participate gainfully in the labour market due to disability. The disability may be congenital or the result of an accident or illness. Services for disabled persons include day care, rehabilitation, home help, and other benefits in kind. Sickness cash benefits compensate lost earnings owing to a temporary inability to work due to illness. (This excludes paid leave related to sickness or injury of a dependent child, which is recorded under family cash benefits.) Expenditures for occupational injury and disease include cash payments such as paid sick leave, special allowances, and disability-related payments such as pensions, if they are related to specific occupational injuries and diseases. Data source: OECD, Social Expenditures Database. Thick line: United States. “Asl” = Australia; “Aus” = Austria.

Figure 17. Generosity of sickness compensation
Includes program eligibility criteria, coverage, waiting period, benefit replacement rate, and benefit duration. Data source: Lyle Scruggs, Detlef Jahn, and Kati Kuitto, Comparative Welfare Entitlements Data Set 2, version 2014‐03, series sgen. “Asl” = Australia; “Aus” = Austria.

All rich countries provide some help with housing, mainly in the form of public housing and rental assistance. As figure 18 shows, spending in this area is modest almost everywhere, at 1% of GDP or less. It has changed little over time.

Figure 18. Public expenditures on housing programs
Share of GDP. Data source: OECD, Social Expenditures Database. Thick line: United States. “Asl” = Australia; “Aus” = Austria.

Spending on families includes cash benefits and tax allowances, provision or subsidization of childcare and preschool, and other family services.18 As figure 19 shows, here too the Nordic countries are among biggest spenders. The United States, despite its cultural emphasis on strong families, is at the bottom. Assistance to families is, along with health care, the chief area in which social programs have tended to become more generous in recent decades. The spending total is smaller than for health, and the rise has been more modest, but it is real.

Figure 19. Public expenditures on family programs
Share of GDP. Broadly speaking there are three types of public spending on family benefits: (1) Child-related cash transfers (cash benefits) to families with children, including child allowances, with payment levels that in some countries vary with the age of the child, and sometimes are income-tested; public income support payments during periods of parental leave; and income support for single-parent families. (2) Public spending on services for families (benefits in kind) with children, including direct financing and subsidizing of providers of childcare and early education facilities, public childcare support through earmarked payments to parents, public spending on assistance for young people and residential facilities, public spending on family services, including center-based facilities and home help services for families in need. (3) Financial support for families provided through the tax system, including tax exemptions (e.g. income from child benefits that is not included in the tax base); child tax allowances (amounts for children that are deducted from gross income and are not included in taxable income), and child tax credits, amounts that are deducted from the tax liability. Data source: OECD, Social Expenditures Database. Thick line: United States. “Asl” = Australia; “Aus” = Austria.

The last of the OECD’s categories of social spending is a catch-all “other” group. Figure 20 shows that, apart from in Canada, these expenditures are relatively small, and they’ve been largely constant over the past generation.

Figure 20. Public expenditures on other social programs
Share of GDP. Data source: OECD, Social Expenditures Database. Thick line: United States. “Asl” = Australia; “Aus” = Austria.

To summarize: Pensions and health are by far the largest sources of expenditure on public social programs. Health and family are the only two areas with noteworthy increases in spending during the past generation; most others have been flat or decreased modestly. The United States is a big spender on health care; when private expenditures are included, America’s health spending dwarfs that of any other rich nation, and even its public spending is among the highest. In almost all other areas of public social program expenditure, the US is at or near the bottom of the pack.

A CLOSER LOOK AT AMERICA’S CASH AND NEAR-CASH PROGRAMS

For the least well-off, government cash and near-cash transfers are an important source of income. No matter how much we might want working-aged households to be self-sufficient, there inevitably will be a nontrivial fraction that have significant labor market disadvantages — disability, family constraint, geographic vulnerability to structural unemployment, and more — and so rely partly or wholly on income from public social programs. It isn’t surprising, therefore, that countries in which the level of public minimum income protection is higher tend to have lower incidence of material deprivation.19 Nor should we be surprised that in recent decades, rising incomes for low-end households have tended to depend on increases in government transfers.20

America’s largest cash transfer program is Social Security, which was created in 1935. The share of elderly (age 65 and over) Americans receiving Social Security increased rapidly until the mid-1970s, when it reached 90%. It has remained at that level since. As figure 21 indicates, about 13% of Americans are Social Security recipients. As figure 22 shows, the average benefit level has increased steadily. It currently is about $16,000 a year.

(As in the previous section, I use the same range of values on the vertical axes of all charts in this section in order to facilitate comparison. On the horizontal axes, I note the year 1980, which is sometimes thought to be a point at which the upward trajectory of US social policy generosity slowed or reversed.)

Figure 21. Social Security recipients
Share of the population. Social Security Old-Age. Data source: Social Security Administration, Annual Statistical Supplement, 2016, table 3.C5.

Figure 22. Social Security benefit level
Average yearly benefit. Monthly benefit multiplied by 12. Inflation adjustment is via the CPI-U-RS. “k” = thousand. Social Security Old-Age. Data source: Social Security Administration, Annual Statistical Supplement, 2016, table 5.C2.

Early death of a family breadwinner is far less common than in earlier eras, but it still happens. In most instances the family receives Social Security Survivors benefits. About 6 million Americans receive these benefits, and the program pays out twice as much each year as all private life insurance policies combined.21 Figure 23 shows the population share receiving these benefits, and figure 24 shows the benefit level for the surviving spouse and for children.

Figure 23. Social Security Survivors recipients
Share of the population. Data source: Social Security Administration, Annual Statistical Supplement, 2016, tables 5.F4 and 5.F8.

Figure 24. Social Security Survivors benefit level
Average yearly benefit. Monthly benefit multiplied by 12. Inflation adjustment is via the CPI-U-RS. “k” = thousand. Data source: Social Security Administration, Annual Statistical Supplement, 2016, tables 5.F4 and 5.F8.

About 40% of unemployed Americans qualify for unemployment insurance compensation. In a typical year, that means approximately 1% of the population receives this benefit, as figure 25 shows. The amount depends on a person’s prior earnings. As figure 26 suggests, in recent years it has averaged close to $18,000 on an annual basis, though most recipients get this money for half a year at most. The average benefit level has increased only a little since the 1970s, mainly because earnings for Americans in the lower half have been stagnant.22

Figure 25. Unemployment insurance recipients
Share of the population. Data source: US Department of Labor, Employment and Training Administration.

Figure 26. Unemployment insurance benefit level
Average yearly benefit. Weekly benefit multiplied by 52. Inflation adjustment is via the CPI-U-RS. “k” = thousand. Data source: US Department of Labor, Employment and Training Administration.

Figure 27 shows that about 3% of Americans receive disability benefits from Social Security Disability Insurance (SSDI) and around 2% from Supplemental Security Income (SSI). Both shares have increased steadily over the past generation. As figure 28 shows, SSDI pays the average beneficiary about $14,000 a year, SSI $7,000. Neither program’s benefit level has changed much since the mid-1970s.

Figure 27. Disability benefit recipients
Share of the population. Data source: Social Security Administration, Annual Statistical Supplement, 2016, table 7.A9.

Figure 28. Disability benefit level
Average yearly benefit. Average for two categories: blind and disabled. Monthly benefit multiplied by 52. Inflation adjustment is via the CPI-U-RS. “k” = thousand. Data source: Social Security Administration, Annual Statistical Supplement, 2016, table 7.A5.

America has two principal social assistance programs: Temporary Assistance for Needy Families (TANF) and the Supplemental Nutritional Assistance Program (SNAP, or “food stamps”). TANF was created by the 1996 welfare reform, replacing Aid to Families with Dependent Children (AFDC), which had been created in the 1930s. That mid-1990s reform gave states more leeway in deciding eligibility criteria and benefit levels, and it placed a five-year lifetime cap on eligibility. As figure 29 shows, the share of Americans (parents and their children) receiving AFDC-TANF has fallen steadily and sharply since the mid-1990s, from a high of 5% to just 1% as of 2015.

The AFDC-TANF benefit level has been declining too. But as figure 30 indicates, that continues a trend dating all the way back to 1970. The average benefit has fallen, in inflation-adjusted terms, from $12,000 in 1970 to just $5,000 as of 2015.

Figure 29. AFDC-TANF recipients
Share of the population. Data source: US Department of Health and Human Services, Office of Family Assistance, “AFDC Caseload Data” and “TANF Caseload Data.”

Figure 30. AFDC-TANF benefit level
Average yearly benefit per recipient family. Inflation adjustment is via the CPI-U-RS. “k” = thousand. Data source: for pre-2000, Social Security Administration; for 2000ff, US Department of Health and Human Services, Office of Family Assistance, “Characteristics and Financial Circumstances of TANF Recipients.”

Food stamps (SNAP) has moved in the other direction. It became a nationwide program in 1974. As figure 31 shows, in that year 6% of Americans received the benefit. By the mid-1990s that share had risen to 10%. In the wake of the 2008-09 Great Recession, and on the heels of a legislated increase in scope, the share increased to 13-15% in the years 2010 to 2015. The benefit too has risen, though only a little. It was about $3,000 per year in the mid-1970s and $4,000 in the early 2010s, as figure 32 shows.

Figure 31. SNAP recipients
Share of the population. Data source: US Department of Agriculture, Food and Nutrition Service.

Figure 32. SNAP benefit level
Average yearly benefit per recipient family. Inflation adjustment is via the CPI-U-RS. “k” = thousand. Data source: US Department of Agriculture, Food and Nutrition Service.

The Earned Income Tax Credit (EITC) was created in 1975. The EITC subsidizes pretax income for low-earning households by as much as 45%. The credit is refundable; if it amounts to more than the household owes in federal income taxes, the household receives the difference as a cash refund. It therefore functions like a cash benefit.

Nearly one in four Americans receives the EITC. This share rose sharply between the late 1980s and the mid-1990s and again in the 2000s, as figure 33 indicates. These increases were a result of changes in eligibility criteria, increases in the benefit amount, and stagnant wage levels for Americans on the lower rungs of the wage ladder. The average amount recipient households get is $2,300 per year. As figure 34 shows, this amount increased significantly between 1987 and 1996. Since then it has been flat.

Figure 33. EITC recipients
Share of the population. These numbers are estimates, calculated using the number of EITC recipient households and the average number of persons per household. Data source: Tax Policy Center, “EITC Recipients.”

Figure 34. EITC benefit level
Average yearly benefit per recipient household. Inflation adjustment is via the CPI-U-RS. “k” = thousand. Data source: Tax Policy Center, “EITC Recipients.”

The federal government has been involved in the provision of housing assistance to low-income Americans since the 1930s, when it began construction of public housing units. Today this assistance primarily takes the form of vouchers that recipients use to help defray the cost of rent, a tax credit to developers who build housing for low-income renters, and a million or so public housing units. As figure 35 shows, about 5% of Americans benefit from these programs. The expenditures amount to around $9,000 per recipient household, as figure 36 indicates.

Figure 35. Low-income housing assistance recipients
Share of the population. These numbers are estimates, calculated using the number of housing assistance recipient households and units and the average number of persons per household. Data source: Robert Collinson, Ingrid Gould Ellen, and Jens Ludwig, “Low-Income Housing Policy,” Working Paper 21071, National Bureau of Economic Research, 2015, table 3, using data from the Department of Housing and Urban Development’s Annual Performance Report.

Figure 36. Low-income housing assistance expenditures
Expenditures per recipient household. Inflation adjustment is via the CPI-U-RS. “k” = thousand. Expenditures for the Low-Income Housing Tax Credit are included beginning in 2004. Data source: Robert Collinson, Ingrid Gould Ellen, and Jens Ludwig, “Low-Income Housing Policy,” Working Paper 21071, National Bureau of Economic Research, 2015, tables 5 and 6, using data from the Office of Management and the Budget and the Joint Committee on Taxation.

FROM DECOMMODIFICATION TO EMPLOYMENT PROMOTION

Social scientists and policy makers traditionally have viewed the welfare state as a means of achieving economic security and redistribution. Social programs provide money and services to individuals and households in order to insure against market-based risks and reduce market-generated inequality and poverty. In recent decades, an additional goal has become prominent: employment. Policy reforms and innovations have more and more aimed to increase paid work.23 This approach, commonly referred to as “activation” or “social investment,” isn’t novel. Sweden in the 1950s put in place a set of active labor market programs, including retraining, job placement assistance, and public sector jobs as a last resort. What’s new is the centrality of employment in modern social policy.

The employment promotion turn has a number of causes.24 The chief one, arguably, is the fact that the retirement of the baby boom generation will increase pension and health care costs at the same time that capital mobility has made increasing tax rates more difficult. Getting more citizens into paid work means greater payroll and income tax revenue without an increase in tax rates, and it also means lower expenditures on social assistance, unemployment compensation, and related programs. Other considerations and aims pushing policy makers toward employment promotion include poverty reduction, social inclusion, fairness, women’s independence and fulfillment, subjective well-being, and encouragement by external organizations such as the OECD and the EU.

The policy tools used to boost employment are wide-ranging. Some impose requirements, while others provide incentives. Some work on the demand side, attempting to increase hiring by employers, while others try to increase the supply of those seeking employment. Some are aimed broadly, at the entire working-age population, whereas others are targeted at particular groups for which the employment rate is low, such as women, those near retirement age, the young, persons with limited education and/or skills, immigrants, and people with disabilities. Some aim to increase employment directly, while others attempt to improve people’s capabilities, under the assumption that this will indirectly bring about greater employment.

In every rich country, persons with more education are more likely to be employed, so a common strategy for increasing employment is to improve educational attainment and/or the quality of schooling. Policy makers have attempted to do this in a variety of ways: subsidizing early education; improving elementary and secondary schooling via increased funding, greater centralization or decentralization of decision making, heightened teacher accountability, school choice, and others; increasing opportunity for college attendance via reduced costs or better access to grants and loans. Some countries also have enhanced opportunity for “lifelong learning” via retraining, subsidies for return to schooling, access to online education, and others.

Government benefits with easy eligibility conditions, generous payments, and lengthy duration may discourage employment. Some nations have tightened eligibility criteria, reduced benefit levels, made benefit receipt conditional on participation in training or employment (“workfare”), and/or shortened the duration of receipt.

Public assistance with finding new employment has long been a key feature of Swedish active labor market policy. Since the mid-1990s its use has become more widespread. A trend in recent years has been toward individualized assistance: clients are assigned an individual caseworker who evaluates their abilities and needs, helps with benefit receipt, oversees the search for new work and placement with the firm, and monitors employment outcomes. In areas where key sources of employment are not easily accessible via public transport, governments sometimes provide subsidies to defray the cost of transportation to and from work.

Cash transfers and tax credits to people in paid work but with low earnings are increasingly prominent in affluent countries. The US and the UK began using such employment-conditional earnings subsidies in the 1970s, and in recent decades most of the other rich longstanding-democratic countries have adopted some version of them.25

Public employment can be used as a “last resort” in providing jobs for those unable to find work in the private sector. In the Nordic countries and France, the government typically has accounted for 25% to 30% of total employment, making the public sector a common source of employment whether as a last resort or a first option.

Part-time jobs are an attractive option for some people, especially second earners in households. In a number of countries they now account for a quarter or more of all employment, and in the Netherlands a third. Some of this is due to inability to find a satisfactory full-time position, but surveys suggest much of it is by choice, particularly where part-time employees have the same wage and benefit status as full-timers. The Dutch employment success story since the early 1980s is largely one of part-time employment growth. Flexible work schedules also tend to be attractive to potential employees, again especially those who would be a second earner within a household. Protections for such employees as well as financial incentives to employers for expanding flexible work-time options can help to attract them.

Taxes can matter. Some countries have moved to shift the tax burden away from payroll taxes in order to reduce nonwage labor costs. Also, in some nations the tax system penalizes a couple with two earners relative to those with just one earner. Reducing this disincentive can help to boost employment, particularly among women.

In jobs where productivity is low and difficult to increase, employers may hire fewer people if wages are too high. Lowering the wage floor may therefore increase employment. In countries with a statutory minimum wage, policy makers can do this directly. In the United States and the Netherlands, the inflation-adjusted value of the statutory minimum has been allowed to fall since the early 1980s. In nations without a statutory minimum wage, one response has been the formation of formal or informal social pacts, in which government encourages wage restraint in exchange for changes in certain social programs, taxes, or active labor market programs.

Strong employment protection regulations make it more difficult and/or costly for employers to fire employees. Since the mid-1990s, most of the continental European and Nordic countries have reduced the stiffness of employment protection regulations, mainly by easing restrictions on fixed-term (temporary) employment.

Family-friendly policies, particularly paid parental leave and quality affordable early education, can boost women’s employment. Originally confined to the Nordic countries along with Belgium and France, these programs now are embraced to one degree or another by many affluent nations.26

Most of the rich democratic countries have higher employment rates now than they did a generation ago. Moreover, Sweden, Denmark, and Norway, consistent practitioners of employment-promoting social programs since the 1970s, tend to be at or near the top in employment. While we don’t know for certain how much of the common employment rise or of the Nordic countries’ success owes to this approach to social policy, these patterns suggest grounds for optimism that it can help.

The US experience also is suggestive. The United States has pursued a “market liberal” approach to employment growth: a low wage floor, very limited labor market regulations, relatively stingy government benefits, comparatively low taxes, steady deregulation of product markets, and limited support for retraining, job placement, and work-family balance. Up to the turn of the century the “great American jobs machine” was comparatively successful; the US was among the rich world’s leaders in raising its employment rate. But since then America’s employment performance has been comparatively poor.27

While most rich democracies have been converging toward a “social investment” approach, a number of key policy questions don’t yet have clear-cut answers. Should early education be universal or targeted to the poor? Should paid parental leave be for six months or one year or three years? Should it include a “daddy quota”? Should there be a statutory minimum wage? If so, how high? What is the best mix of carrots and sticks for social assistance recipients reluctant to enter or reenter paid work? Should low market income be supplemented by an employment-conditional earnings subsidy? Is individualized assistance more helpful in the early years, the K-12 years, or later in the life course? What is the right balance between employment protection for workers and flexibility for employers?

Getting the policy details right requires experimentation, adjustment, and learning from best practice. And since countries vary in political structure, economic institutions, culture, and in many other ways, optimal policies and policy combinations may well vary too.

SUMMARY

Most of us like security. As we become more affluent, we are willing to pay more to insure against major risks. Government social programs cushion against some risks more effectively and efficiently than individuals, families, voluntary organizations, or markets. They also enhance fairness. And they tend to work well. As a result, once countries achieve sustained economic growth, they tend to adopt and expand such programs. The welfare state, as it is often called (disparagingly by some, admiringly by others), is one of the core institutions of modern life.

By the standards of contemporary rich democratic nations, America’s social programs provide inadequate protection against life’s risks. Our overall social spending is quite high, but private employer-sponsored social programs (pensions, health insurance, dental insurance, paid sick leave, paid parental leave, short-term disability insurance, group life insurance) play a much more prominent role here than elsewhere. Private social spending excludes many Americans, especially those who work for smaller firms or who aren’t employed, and it tends to go disproportionately to persons with higher incomes. Public programs exclude fewer people, and their benefits tend to be tilted toward those with greater need and fewer resources.28 The United States has fewer public social programs than other affluent countries (no sickness insurance or paid parental leave), and the generosity of those it has tends to be on the low end.

Will the US remain a laggard? Given our political and economic institutions, it is tempting to say yes. And yet, if the pattern of the past century continues, the United States may well move toward significantly greater use of public insurance and public services.29 Policy makers, perhaps with a push from organized interest groups or the populace, will recognize the benefits of expanded social programs and will attempt to move the country in that direction. Often they will fail. Sometimes, however, they will succeed. Progress will be incremental, coming in fits and starts, but it will have staying power. New programs and expansions of existing ones will tend to persist, because programs that work well become popular and because the US policy making process makes it difficult for opponents of social programs to remove them. Small steps and the occasional big leap, coupled with limited backsliding, will have the cumulative effect of significantly increasing the breadth and generosity of government social programs.


  1. Francis Castles, Stephan Leibfried, Jane Lewis, Herbert Obinger, and Christopher Pierson, eds., Oxford Handbook of the Welfare State, Oxford University Press, 2010; David Garland, The Welfare State: A Very Short Introduction, Oxford University Press, 2016. 
  2. Mark Thoma, “Who First Said the US Is ‘an Insurance Company with an Army’?,” Economist’s View, 2013. 
  3. Michael J. Graetz and Jerry L. Mashaw, True Security: Rethinking American Social Insurance, Yale University Press, 1999; Nicholas Barr, The Welfare State as Piggy Bank, Oxford University Press, 2001; David A. Moss, When All Else Fails: Government as the Ultimate Risk Manager, Harvard University Press, 2002; John Quiggin, “The Risk Society: Social Democracy in an Uncertain World,” Centre for Policy Development, 2007. 
  4. I.M. Rubinow in the 1930s termed the first four of these the “four horsemen of the apocalypse,” which “ride roughshod over lives and fortunes of millions of wage workers of every modern industrial community.” Quoted in Mike Konczal, “The Voluntarism Fantasy,” Democracy Journal, 2014. 
  5. Theodore R. Marmor, Jerry L. Mashaw, and John Pakutka, Social Insurance, CQ Press, 2014; Matthew W. Brault, “Americans with Disabilities: 2010,” Current Population Reports P70-131, Census Bureau, 2012. 
  6. Alexander Hicks, Social Democracy and Welfare Capitalism, Cornell University Press, 1999; Evelyn Huber and John Stephens, Development and Crisis of the Welfare State, University of Chicago Press, 2001; Stein Kuhnle and Anne Sander, “The Emergence of the Western Welfare State,” in Oxford Handbook of the Welfare State, edited by Francis G. Castles et al, Oxford University Press, 2010. 
  7. Irwin Garfinkel, Lee Rainwater, and Timothy Smeeding, Wealth and Welfare States, Oxford University Press, 2010. 
  8. Alexander Hicks, Social Democracy and Welfare Capitalism, Cornell University Press, 1999; Evelyn Huber and John Stephens, Development and Crisis of the Welfare State, University of Chicago Press, 2001; Alberto Alesina and Edward L. Glaeser, Fighting Poverty in the US and Europe, Oxford University Press, 2004. 
  9. Christopher Howard, The Hidden Welfare State, Princeton University Press, 1997; Jacob S. Hacker, The Divided Welfare State, Cambridge University Press, 2002; Neil Gilbert, “Comparative Analyses of Stateness and State Action: What Can We Learn from Patterns of Expenditure?,” in United in Diversity? Changing Social Models in Europe and America, edited by Jens Alber and Neil Gilbert, Oxford University Press, 2010; Willem Adema, Pauline Fron, and Maxime Ladaique, “How Much Do OECD Countries Spend on Social Protection and How Redistributive Are Their Tax/Benefit Systems?,” International Social Security Review, 2014. 
  10. This updates and extends figure 1 in Jens Alber, “What the European and American Welfare States Have in Common and Where They Differ: Facts and Fiction in Comparisons of the European Social Model and the United States,” Journal of European Social Policy, 2010. 
  11. Lane Kenworthy, “Public Insurance and the Least Well-Off,” The Good Society. 
  12. Francis G. Castles and Deborah Mitchell, “Worlds of Welfare and Families of Nations,” in Families of Nations, edited by Francis G. Castles, Dartmouth, 1993; Gøsta Esping-Andersen, Social Foundations of Postindustrial Economies, Oxford University Press, 1999; Alexander Hicks and Lane Kenworthy, “Varieties of Welfare Capitalism,” Socio-Economic Review, 2003; Wil A. Arts and John Gelissen, “Models of the Welfare State,” in Oxford Handbook of the Welfare State, edited by Francis G. Castles et al, Oxford University Press, 2010; Martin Schröder, Integrating Varieties of Capitalism and Welfare State Research: A Unified Typology of Capitalisms, Palgrave Macmillan, 2013. 
  13. See also Francis G. Castles, “What Welfare States Do: A Disaggregated Expenditure Approach,” Journal of Social Policy, 2008: 45-62; Herbert Obinger and Uwe Wagschal, “Social Expenditure and Revenues,” in Oxford Handbook of the Welfare State, edited by Francis G. Castles et al, Oxford University Press, 2010. 
  14. OECD, Pensions at a Glance, various editions. 
  15. Markus Gangl, “Scar Effects of Unemployment: An Assessment of Institutional Complementarities,” American Sociological Review, 2006. 
  16. David Card, Jochen Kluve, and Andrea Weber, “What Works? A Meta Analysis of Recent Active Labor Market Program Evaluations,” Journal of the European Economic Association, 2018. 
  17. Burt S. Barnow and Jeffrey Smith, “Employment and Training Programs,” Working Paper 21659, National Bureau of Economic Research, 2015. 
  18. The OECD’s definition of family expenditures is as follows: “Family benefits spending refer to public spending on family benefits, including financial support that is exclusively for families and children. Spending recorded in other social policy areas, such as health and housing, also assist families, but not exclusively, and it is not included in this indicator. Broadly speaking there are three types of public spending on family benefits: (1) Child-related cash transfers (cash benefits) to families with children, including child allowances, with payment levels that in some countries vary with the age of the child, and sometimes are income-tested; public income support payments during periods of parental leave and income support for sole parents families. (2) Public spending on services for families (benefits in kind) with children, including direct financing and subsidizing of providers of childcare and early education facilities, public childcare support through earmarked payments to parents, public spending on assistance for young people and residential facilities, public spending on family services, including centre-based facilities and home help services for families in need. (3) Financial support for families provided through the tax system, including tax exemptions (e.g. income from child benefits that is not included in the tax base); child tax allowances (amounts for children that are deducted from gross income and are not included in taxable income), and child tax credits, amounts that are deducted from the tax liability.” 
  19. Kenneth Nelson, “Counteracting Material Deprivation: The Role of Social Assistance in Europe,” Journal of European Social Policy, 2012. 
  20. Lane Kenworthy, “A Decent and Rising Income Floor,” The Good Society. 
  21. Marmor et al, Social Insurance, 102. 
  22. Lane Kenworthy, “Shared Prosperity,” The Good Society. 
  23. Lane Kenworthy, Jobs with Equality, Oxford University Press, 2008; Nathalie Morel, Bruno Palier, and Joakim Palme, eds., Towards a Social Investment Welfare State?, Policy Press, 2012; Anton Hemerijck, Changing Welfare States, Oxford University Press, 2013; Hemerijck, ed., The Uses of Social Investment, Oxford University Press, 2017; Lane Kenworthy, “Employment Promotion,” in Oxford Handbook of the Welfare State, 2nd edition, edited by Daniel Béland et al, Oxford University Press, 2021; Anna Diamantopoulou, Agnieszka Chłoń-Domińczak, Bernhard Ebbinghaus, Elena Granaglia, Anton Hemerijck, Hans-Peter Klös, Catherine Mathieu, Pasi Moisio, Jozef Pacolet, Yves Stevens, Dorottya Szikra, and Anu Toots, The Future of Social Protection and of the Welfare State, European Commission, Social Protection Unit, 2023. 
  24. Gøsta Esping-Andersen, Social Foundations of Postindustrial Economies, Oxford University Press, 1999; Esping-Andersen, The Incomplete Revolution: Adapting to Women’s New Roles, Polity, 2009; Maurizio Ferrera, Anton Hemerijck, and Martin Rhodes, “The Future of Social Europe: Recasting Work and Welfare in the New Economy,” report prepared for the Portuguese Presidency of the European Union, 2000; Fritz W. Scharpf and Vivien A. Schmidt, eds., Welfare and Work in the Open Economy, Oxford University Press, 2000; Gøsta Esping-Andersen with Duncan Gallie, Anton Hemerijck, and John Myles, Why We Need a New Welfare State, Oxford University Press, 2002; Wim Kok, Carlos Dell’Aringa, Federico Duran Lopez, Anna Eckström, Marta João Rodrigues, Christopher Pissarides, Annette Roux, and Günther Schmid, Jobs, Jobs, Jobs: Creating More Employment in Europe, report of the European Commission’s Employment Taskforce, 2003; Kenworthy, Jobs with Equality, 2008; OECD, Extending Opportunities: How Active Social Policy Can Benefit Us All, 2005; OECD, “Is Work the Best Antidote to Poverty?,” in OECD Employment Outlook, 2009. 
  25. Lane Kenworthy, “Do Employment-Conditional Earnings Subsidies Work?” ImPRovE Working Paper 15-10, Herman Deleeck Centre for Social Policy, University of Antwerp, 2015. 
  26. Janet C. Gornick and Marcia K. Meyers, Families That Work, Russell Sage Foundation, 2003; Heather Boushey, Finding Time: The Economics of Work-Life Conflict, Harvard University Press, 2016; OECD, “OECD Family Database”. 
  27. Lane Kenworthy, “Employment,” The Good Society. 
  28. Kenworthy, “Public Insurance and the Least Well-Off.” 
  29. Lane Kenworthy, Social Democratic America, Oxford University Press, 2014, ch. 5.