Governments are bigger than ever. They are also more useless
Why voters across the rich world are miserable
You may sense that governments are not as competent as they once were. Upon entering the White House in 2021, President Joe Biden promised to revitalise American infrastructure. In fact, spending on things like roads and rail has fallen. A flagship plan to expand access to fast broadband for rural Americans has so far helped precisely no one. Britain’s National Health Service soaks up ever more money, and provides ever worse care. Germany mothballed its last three nuclear plants last year, despite uncertain energy supplies. The country’s trains, once a source of national pride, are now often late.
You may also have noticed that governments are bigger than they once were. Whereas in 1960 state spending across the rich world was equal to 30% of GDP, now it is above 40%. In some countries growth in the state’s economic power has been still more dramatic. Since the mid-1990s Britain’s government spending has risen by six percentage points of gdp, while South Korea’s has risen by ten points. All of which raises a paradox: if governments are so big, why are they so ineffective?
The answer is that they have turned into what can be called “Lumbering Leviathans”. In recent decades governments have overseen an enormous expansion in spending on entitlements. Because there has not been a commensurate increase in taxes, redistribution is crowding out spending on other functions of government. This, in turn, is damaging the quality of public services and bureaucracies. The phenomenon may help explain why people across the rich world have such little faith in politicians. It may also help explain why economic growth across the rich world is weak by historical standards.
America, which has some of the best fiscal data, shows how one government became a Lumbering Leviathan. In the early 1950s we estimate that state spending on public services, including everything from paying teachers’ salaries to building hospitals, equalled 25% of the country’s gdp (see chart 1). At the same time, entitlements spending, broadly defined, was a small line item, with outlays on both pensions and other sorts of welfare equivalent to about 3% of GDP. Today the situation is very different. The American government’s outlays on entitlements have swelled and spending on public services has crashed. Both now equal around 15% of gdp.
Other countries have followed a similar path. We have examined long-run GDP data, looking at how much governments spend each year on social benefits and transfers. This includes standard entitlements, such as pensions and tax credits, but also the provision of transfers “in kind”, such as discounts on health insurance and help with housing. Both types have become a lot bigger. On average across the OECD, social expenditure in countries with available data rose from 14% of GDP in 1980 to 21% in 2022 (see chart 2).
Moreover, conventional statistics understate the scale of the change. Governments have accumulated mind-boggling off-balance-sheet obligations to dole out money in the future. Adapting work by James Hamilton of the University of California, San Diego, we estimate that America’s federal government has made pledges of compensation to different groups worth, in aggregate, six times America’s GDP (see chart 3). In addition to reported public debt, Uncle Sam guarantees people’s bank deposits, health-care payouts and mortgages. He will also need to make good on promises to future retirees. In the history of the modern state, this represents a uniquely large financial commitment.
Some of the growth in entitlement spending has been unavoidable. In 2022 there were 33m people over the age of 85 in the rich world, representing 2.4% of the total population—a huge rise on the 5m, representing 0.5% of the total population, around in 1970. Governments have not helped themselves by failing to raise the retirement age: the average person in the rich world currently retires at age 64, no older than in the late 1970s. But it would have been hard (and unwise) to have stopped pension spending from growing.
Because entitlements for the old tend to be universal—European countries, for example, have little private pension provision—more cheques are going to the well-heeled. We estimate that in the oecd between a fifth and a third of entitlements spending, broadly defined, goes to the richest 20% of households. The American government spends about $400bn, or roughly half the budget of the Department of Defence, on transfers for the top income quintile. In 2019 an average household in the top 1% received $16,000 in transfers from Uncle Sam, including from things such as Social Security and Medicare.
Transfers to the working-age population have increased even faster, making the system more redistributive. In 1980 the bottom fifth of American earners received means-tested transfers equal to a third of their gross earnings. By the late 2010s the figure had doubled, before the covid-19 pandemic sent it higher still (see chart 4). A similar pattern is evident in Canada and Finland, two other countries with good data. Spending often follows a ratchet effect. For instance, since the 1970s the share of Americans on food stamps has doubled, to one in eight people. In recessions the number of recipients rises like a rocket; thereafter it falls like a feather.
Across the board, governments have become more generous in times of trouble. During the pandemic they shovelled money to affected workers and companies, as well as to many that were carrying on largely as normal. Throughout the energy crisis of 2022, lots of governments threw caution to the wind. Even the German government, historically among the more tight-fisted, allocated 4.4% of GDP to measures shielding households and firms from its effects. More recently, some have lost the plot. In Italy a project to encourage homeowners to make their homes greener has spiralled out of control, with the government so far disbursing support worth more than €200bn (or 10% of GDP).
Nordic nirvana
A rise in entitlement spending is not necessarily a problem if governments are able to adequately and efficiently fund themselves. Textbook economics says that the societal cost of redistribution comes from the distorted incentives tax and welfare spending can create. These cannot be judged merely by the size of redistribution—the design of the system is what matters most. Indeed, Scandinavian countries have long sustained big states alongside thriving market economies, in part by funding redistribution with high rates of VAT, one of the least-distorting taxes, and by keeping down taxes on capital, which are particularly harmful to growth.
But in recent years politicians have preferred to act as if extra spending can happen with little more taxation of any kind. From the 1960s to the 1990s the tax take, as a share of rich-world GDP, rose steadily. Since the 2000s it has hardly grown. A database of tax reforms maintained by the IMF, and last updated in 2018, suggests that whereas in the 1970s and 1980s reforms were evenly split between revenue-raisers and revenue-cutters, more recent ones have focused on cutting taxes.
By 2022 some 85% of reforms to rich countries’ personal-income-tax bases caused them to narrow, while just 15% broadened them. The biggest reform of the past decade was President Donald Trump’s enormous tax cut in 2017. Neither Mr Trump nor Kamala Harris, the Democratic nominee, promise sober fiscal stewardship in the years to come. To the extent that today’s governments implement measures to raise revenue, they tend to take the form of crafty workarounds. According to our calculations, in 2022 American federal, state and local governments raised $80bn from fines, fees, penalty taxes and settlements—nearly three times as much, relative to GDP, as in the 1960s and 1970s.
Politicians who fail to raise revenues face two choices. One is to run large fiscal deficits: this year rich-world governments will run an aggregate deficit of 4.4% of GDP, even with the global economy in decent shape. Another is to fund more generous entitlements by making cuts elsewhere. Demand for public services has grown hugely. Yet in 2022 the median rich country spent 24% of GDP on them, the same as in 1992. Public-sector employment, as a share of the total, has drifted down since the late 1990s. Everything from state-provided health care to education and public safety has suffered.
Another historical role of government—now waning—was to provide an efficient bureaucracy. It is tough to measure this quantitatively, but researchers have had a go. Data produced by the Berggruen Institute, a think-tank, and the University of California, Los Angeles, combine objective measures, such as tax revenue, and subjective measures, such as perceptions of corruption, to devise a cross-country measure of “state capacity”. In the G7 group of advanced economies this measure is falling. The same is true of the “rigorous and impartial public-administration index”, produced by V-Dem, another think-tank, which illustrates the extent to which public officials respect the law.
The effects of declining state capacity show up everywhere. Some are small-bore. In America the time lag between a residential project being granted permission to build and construction beginning has doubled since the 1990s. Builders face long waiting times as they form-fill and box-tick. In Britain employment tribunals are facing huge delays owing to a shortage of judges, with hearings on everything from unfair dismissal to racial discrimination now scheduled as far ahead as 2026. Five years ago the website of Australia’s passport office said that the processing time for an application was “three weeks”; two years ago it said “up to six weeks”; by last year it said “minimum of six weeks”.
Governments also seem less willing and able to pull off big projects. It is just about impossible to imagine that the Empire State Building could be built in a year—and yet, in the 1930s, it was. Moreover, throughout the 20th century governments invested both money and intellect in science and R&D, seeking to shift economic growth into a higher gear. Initiatives like DARPA, undertaken in America to devise and spread groundbreaking technologies, hinted at the scale of governments’ ambitions. In the 1950s and 1960s governments, including Germany’s and Japan’s, built millions of units of public housing and millions of miles of road and rail.
Now politicians just want to get from one day to the next. Spending on short-term fixes takes precedence over difficult, long-term projects. Mr Biden talks up his industrial policy, which is supposed to revive manufacturing jobs and reduce America’s dependence on China. In practice, the fiscal outlays associated with the policy are trivial. Elsewhere in the rich world public investment is well down, while governments have slashed r&d departments. Across the OECD the state now accounts for less than 10% of total R&D spending, a sharp change from the post-war norm (see chart 5). Governments are no longer hotbeds of innovation. Almost all of the recent artificial-intelligence developments have emerged from the private sector.
When it comes to growth-boosting reform, such as tweaks to labour laws, governments have almost entirely lost interest. A paper published in 2020 by Alberto Alesina of Harvard University, and colleagues at the IMF and Georgetown University, measured structural reforms, such as changes to regulations, over time. In the 1980s and 1990s politicians in advanced economies implemented lots of reforms. By the 2010s, however, they had ground to a halt. According to our analysis of data from the Manifesto Project, political-party manifestos in the OECD are about half as focused on growth as in the early 1980s.
Leviathans may not remain lumbering for ever. Running large deficits in order to fund transfer payments will, eventually, become too expensive—as countries such as Greece and Italy discovered in the 2010s. At some point populations, fed up with weak economic growth and poor services, may demand that politicians make some difficult choices. Then again, Lumbering Leviathans are formidable. Interest groups are entrenched, familiar incentives apply and it is easier to live for the short term. The system has a life of its own. ■
Correction (September 25th 2024): This article previously said that the Golden Gate Bridge was built in a year, rather than the Empire State Building. Sorry.
No comments:
Post a Comment