How big is the state?

Today, there is a sense that the state is striking back. And that was true even before the arrival of Covid-19. In advanced economies, more firms were nationalised than privatised in the early years of the 21st century, while economies where state ownership is widespread, such as China and Singapore, have experienced exceptional rates of economic growth.1 Household surveys reveal significant and rising support for the expansion of state ownership, perhaps as a reflection of rising inequality and the scars of the global financial crisis of 2008-09.

Against that background, this chapter looks at a novel dataset measuring the size of public-sector employment across economies and over time and tracks the size of the state, both on the demand side of the economy (where the state pays for certain goods and services and redistributes income) and on the supply side (where the state provides certain services directly and employs workers in government agencies or state-owned enterprises). This chapter also builds on household surveys such as the Life in Transition Survey (LiTS) and the World Values Survey, as well as various country-level economic and social indicators.

In a well-functioning market economy, the size of the state may vary in response to citizens’ preferences. Since the middle of the 19th century, government spending has risen as a share of gross domestic product (GDP), reflecting the increasing importance of education, rising life expectancy, the growing cost of providing education and healthcare, and demand for stronger social safety nets and redistribution on account of technological change.

State employment has also grown over the longer term, peaking in the 1980s. It has since declined somewhat in advanced economies and emerging markets alike on account of privatisation and automation, despite government spending continuing to rise or remaining high. In post-communist economies, the public sector’s share of employment declined from around 45 per cent in the mid 1990s to 24 per cent in the mid-2010s. However, state employment in those economies remains around 7 percentage points higher than the levels seen in similar economies with no legacy of central planning. Their government spending, meanwhile, is in line with that of their peers at around 35 per cent of GDP. The state’s footprint tends to be larger in older societies, reflecting higher levels of public healthcare, long-term care and state pensions. Public spending also tends to be higher in economies with higher-quality economic institutions.

Even as the public sector’s share of employment has declined in recent decades, public support for state ownership has grown. Surveys in post-communist economies suggest that 45 per cent of people favour an increase in public ownership, with views on public ownership tending to be more favourable among individuals with lower levels of education and income. Analysis also shows that women, older people, highly educated individuals and people who are less willing or able to take risks are all more likely to work in the public sector.

As the size of the state increases, it becomes ever more important to ensure that the state represents the broader interests of all citizens. At the same time, many groups (notably the young) appear to be becoming increasingly disillusioned with the way that democracy works, while at the same time being absent at the ballot box – a vicious circle that needs to be broken.

This chapter then looks at the role of the state in the context of the response to the Covid-19 crisis. The response to Covid-19 has been different from the pandemic responses seen in 1918 and 1957, highlighting high levels of demand for the socialisation of the risks faced by individuals (from health risks to the risk of becoming unemployed), even if that entails a significant fiscal cost. It has also revealed citizens’ growing expectations with regard to the services that the state should provide. That increased demand for the socialisation of risks can, in part, be seen as a response to the fact that economic risks have increasingly been shifted onto those least able to tolerate them – particularly individuals with lower levels of education and income.

The ability of the state to deliver on citizens’ expectations, both in response to Covid-19 and in the longer term, will depend on its fiscal space and administrative capacity. Most governments have seen increases in their fiscal space on account of higher revenues and lower interest rates, while administrative capacity varies considerably across countries.

Will public ownership increase? The answer to that will depend on people’s preferences (with support for public ownership likely to rise further on the back of Covid-19), the objectives of state ownership and whether the private sector could potentially achieve those objectives more efficiently. This discussion is then picked up in subsequent chapters, which look at state-owned enterprises, state-owned banks and the use of industrial policy to foster a green economy.

A larger state footprint in the economy may manifest itself on the demand side, on the supply side, or both. When governments decide to pay new benefits (such as wage subsidies during the Covid-19 crisis), finance the construction of a new bridge or increase payments to defence contractors, they act on the demand side of the economy – to the extent that the goods and services that are purchased using the transferred funds are supplied predominantly by the private sector.

The state may also become an increasingly important direct provider of services and employment – for instance, in education, healthcare, finance or transport, or as a result of state bailouts and partial nationalisations in other sectors (see, for example, Box 1.1 on flag carriers in the airline industry).

Historically, patterns in terms of state expansion have differed across economies. In the United States of America, for example, the state significantly increased spending on healthcare and education in the aftermath of the Second World War, but the private sector remained the primary provider of those services. In Europe, on the other hand, education and healthcare are largely provided by the public sector.

This reflects differing preferences as regards the public or private provision of services such as healthcare and education. Preferences in respect of the magnitude of defence spending or international aid or the desired degree of income redistribution also vary from country to country. Meanwhile, population ageing may lead to an increase in the size of the public sector where the state has primary responsibility for healthcare, pensions, mid-career retraining and other aspects of countries’ social safety nets.5 In addition, the public sector tends to be a more stable source of employment (as discussed in Chapter 2), and preferences regarding the trade-off between stability of employment and income may vary over time, affecting the preferred size of the state.

Furthermore, the boundaries between the state and the private sector can sometimes become blurred, reflecting mixed ownership of enterprises and active use of industrial policy tools (as discussed later in this chapter, as well as in Chapter 4). The following sections examine changes in the size of the state over time, starting with the demand side of the economy and moving on to state employment.

In the United Kingdom and the United States of America, both the First and Second World Wars led to major increases in public spending that were only partially reversed subsequently. In Sweden, meanwhile, the welfare state gradually expanded in the 1970s and 1980s, and a major increase in public spending followed the banking and economic crisis of the early 1990s.

To some extent, that secular increase in public spending reflects the rising importance of the accumulation of skills – and thus education. Increases in life expectancy have also led to higher levels of spending on healthcare and pensions. Moreover, healthcare spending is likely to rise further following the outbreak of Covid-19. At the same time, services such as education and healthcare have become relatively more expensive, owing to the fact that productivity tends to rise more slowly in the “non tradeable” service sectors than it does in tradeable sectors (such as manufacturing and agriculture), while wages are largely determined by productivity in goods that are traded across borders (manufacturing). In fact, in non-tradeable sectors, differences between advanced economies and low-income economies in terms of productivity levels tend to be small relative to the corresponding differences in manufacturing, mining or agriculture.6 Thus, as manufactured goods have become cheaper, services typically provided by governments (such as education and healthcare) have become relatively more expensive.

In addition, the range of market failures and externalities that government policies seek to address has become wider. Last year’s Transition Report, for instance, found that in the absence of active government policies, managers of firms were unlikely to pay attention to green issues.7

With modern technology polarising employment (as evidenced by the rising numbers of high-skilled and low-skilled jobs) and medium-skilled jobs being particularly vulnerable to automation, the need for income redistribution and stronger social safety nets has risen.8 Meanwhile, those same types of technology require more complex regulations in order to underpin modern markets, with governments footing the bill for creating and enforcing such regulations.

The next section looks at the other side of the coin, examining the state’s role in the production of added value on the supply side of the economy.

Data on state employment have been obtained from national authorities, the International Labour Organization (ILO), labour force surveys, reports published by the IMF and the World Bank, and various other sources. Inevitably, however, sources and definitions vary. Efforts have been made to account for employment in small enterprises and rural employment in economies such as China and Russia. As a result, estimates of the public sector’s share of employment in Russia (around a quarter of total employment) are lower than alternative estimates derived from official data on employment in large and medium-sized enterprises and entities.10

Reassuringly, those data are fairly closely aligned with estimates derived from the three rounds of the Life in Transition Survey, a representative household survey that was conducted by the EBRD and the World Bank in 37 economies in 2006, 2010 and 2016 (see Chart 1.3). In each economy, at least 1,000 individuals were randomly selected to participate in the survey. Among other things, survey respondents indicated whether they were employed by the private sector, worked for a state-owned enterprise, had some other kind of government-paid job (in education, healthcare or public administration, for instance) or were not in employment.11 In the case of Russia, for example, the Life in Transition Survey suggests that the public sector accounts for 28 per cent of total employment.

In the second half of the 19th century, the state employed around 4 per cent of people in Sweden, the United Kingdom and the United States of America. Those shares peaked at an average of around 25 per cent in the early 1980s, before privatisation reduced the public sector’s share of employment somewhat. Similar trends were observed in many other advanced economies.

State employment has tended to rise somewhat after major upheavals (notably the First and Second World Wars) and major recessions, partially reflecting state bailouts of private companies and banks, as well as lay-offs in the private sector. The 2008-09 financial crisis was no exception in that regard. Most of those increases have since been partially reversed.

In relative terms, state employment remains higher in the EBRD regions than it is in other emerging markets. (Analysis later in this chapter looks at the extent to which this could be explained by various country-level characteristics.) Indeed, the decline in the public sector’s share of employment weakened in the EBRD regions in the mid-2010s. What is more, in around a third of all economies in the EBRD regions, the public sector’s share of employment was actually higher in 2018 (the latest available reading) than it had been three years earlier, with notable increases being observed in countries such as Armenia, Georgia, Kazakhstan and Mongolia.

In part, this is also a product of the growing economic heft of sovereign wealth funds – particularly in commodity-rich economies, where such funds provide a cushion against the volatility of commodity prices and save wealth for future generations.17 Related to that is the fact that commodity prices have risen strongly since the late 1990s, leading to much higher valuations for national oil companies and other state-owned commodity exporters (see Chapter 2). At the same time, many large state owned firms operate in capital-intensive sectors that are also exposed to rapid automation.18 As a result, the fact that state-owned multinationals are increasingly featuring among the world’s largest firms is not inconsistent with the public sector’s share of employment remaining stable or declining.

In Finland, Norway and other Nordic economies, both government spending and state employment are relatively high. However, in other advanced economies in Europe, including Italy, Germany and the Netherlands, the state plays a major role as a source of demand and a mechanism for redistributing income, but a more limited role on the supply side of the economy. A similar pattern can be observed in some low-income economies, including Mozambique and Liberia, where high levels of government spending have been facilitated by large-scale external borrowing and inflows of aid.

In contrast, there are other economies (such as Armenia, Brunei, China, Kazakhstan and Suriname) where the state plays a major role on the supply side, but the ratio of government spending to GDP is relatively modest in comparison. In post-communist economies, the public sector’s share of employment generally tends to be relatively large, while government spending tends to be broadly in line with that of their peers, something that is corroborated by regression analysis taking account of countries’ income per capita and other characteristics. To some extent, the combination of high state employment and relatively modest government spending may reflect the limitations of the data collection exercise. While employment by state-owned enterprises is included in total public employment, their spending on goods and services (procurement) can be sizeable, but is not included in the indicator of the state’s role on the demand side of the economy owing to the limited data available.19

Lastly, in much of Latin America and sub-Saharan Africa the role of the state is limited on both the supply side and the demand side.

The state tends to be larger, in terms of both public spending and state employment, in ageing economies (those where the ratio of people aged 65 and over to people aged 15 to 64 is higher). In older societies, the provision of public services such as healthcare, disability care and long-term care tends to be more labour-intensive and more expensive (see Chart 1.9). Doubling the old-age dependency ratio (going, for example, from the level seen in Moldova to that observed in Bulgaria) is associated with a 5 percentage point increase in the public sector’s share of employment. This holds when other characteristics of the economy (such as income per capita) are taken into account.

The relationship between the size of the state and ageing can also be explored within countries over time. This analysis explains the average level of government spending or state employment over a four-year period using the country’s average values for the preceding period (given a high degree of persistence in state employment and government expenditure), as well as various country-level characteristics. In this dynamic panel setting, income per capita, economic and political institutions, openness to trade and the ratio of natural resource rents to output can all be instrumented using their values in previous periods using a version of the Arellano-Bond generalised method of moments (GMM) estimator.20 This helps to account for the possibility that government spending or state employment could itself affect income per capita or the quality of economic institutions.

The results of this analysis suggest that the public sector’s share of employment does indeed tend to rise as the population ages (see Table 1.1). Moreover, it also tends to rise as population growth accelerates, since that drives up demand for education. That second correlation may also reflect the difficulty of creating jobs in the private sector in economies where the labour force expands rapidly.

In contrast, it appears that there are no longer any systematic differences between post-communist economies and their peers in terms of government spending. In the period 1995-2004, post-communist economies spent, on average, 9 percentage points of GDP more than their peers. Since then, government spending has risen in many emerging markets. Average government spending in post-communist economies (which stood at 35 per cent of GDP in 2019 and is projected to rise towards 40 per cent in 2020) is now in line with the levels seen in other economies once various relevant characteristics are taken into account – notably the rapid population ageing seen in many economies in emerging Europe.

On average, 33 per cent of the respondents who were surveyed in advanced economies between 2017 and 2020 favoured the expansion of public ownership, up from 27 per cent two decades earlier. In post-communist economies, meanwhile, 45 per cent of respondents were in favour of increasing public ownership (see Chart 1.10; all quoted differences are statistically significant at the 1 per cent level). Support for public ownership in post-communist economies has been broadly stable over time – rising slightly, if anything (having stood at 43 per cent in the mid-1990s).

These findings are derived from the World Values Survey, various rounds of which have been conducted worldwide since 199523 (with the Life in Transition Surveys conducted in the EBRD regions and a number of comparator economies giving a similar result). In both surveys, respondents are asked to express their views on the ownership of business and industry on a scale of 1 (“completely agree with the statement that private ownership should be increased”) to 10 (“completely agree with the statement that government ownership should be increased”). People who give a response of 5 or lower are deemed, on balance, to support private ownership, and those who give a response of 6 or higher are deemed to support public ownership.

Among individuals who reached adulthood after the start of the transition from central planning to market economics (that is to say, those born in the mid-1970s or later), support for state ownership does not depend strongly on the individual’s age, a pattern similar to that observed in advanced economies and other emerging markets. Among older individuals, however, support for public ownership is stronger among those who were older at the time of market reforms, possibly reflecting the strong increase in inequality that was seen in the early years of the transition process. Given this pattern, support for public ownership could be expected to decline over time as people born after the mid-1970s gradually account for a growing percentage of the population. At the same time, however, average support for state ownership among people born in a given year has risen by an average of 6 percentage points, resulting in a slight increase in overall support for the expansion of public ownership.

In comparator economies, support for public ownership as expressed in World Values Surveys increased from 48 per cent in the 1990s to 53 per cent in the late 2010s.

More recently, the Asian flu epidemic that claimed an estimated 2 million lives in 1957 (a share of the world’s population that would be equivalent to 6 million people today) may have reduced the growth rates of major economies by around 3 percentage points in that year.26 In the United Kingdom, some factories and mines closed, but those closures were fairly limited.27 On the basis of those events, pre-2020 studies looking at the likely economic impact of a future pandemic mostly restricted themselves to the impact on tourism and trade, concluding, for example, that “although a pandemic would take a huge toll in human suffering, it would most likely not be a severe threat to the European macroeconomy”.28 Pandemic scenarios with a death toll of close to 15 million were assumed to be compatible with positive economic growth in Europe and the United States of America.29 Those studies did not factor in widespread social distancing.

When Covid-19 struck, the world had great expectations in terms of the state’s ability to minimise the risks posed to individuals’ lives, despite the economic costs. Hundreds of thousands of restaurants, retail shops, beauty salons and other businesses, small and large, were ordered to close. The resulting disruption to global economic activity in the medium term is projected to be the largest since the Great Depression and the Second World War.

The contrast with earlier pandemics underscores the extent to which views about the state’s role in society have changed. That change has been observed in virtually every corner of the globe, regardless of the political and economic systems in place, and reflects increased demand for the socialisation of risks, even if that may entail weaker average growth.

Indeed, younger individuals and low-income households are likely to be disproportionately affected by the Covid-19 crisis. Recent studies show that the self-employed and workers whose hours vary at their employers’ discretion under zero-hours contracts are more likely to have been negatively affected by the downturn.31 This is because poorer workers are likely to be concentrated in the occupations and sectors that have been most affected by closures (such as retail services) and are least likely to be able to work from home. A recent study estimates that only around a third of US jobs can be performed from home and that those jobs pay an average of around 55 per cent more than others.32 Thus, the Covid-19 pandemic risks further exacerbating inequality.

Even before the Covid-19 crisis, economic risks had already been pushed onto the very people who tend to dislike them most – those with lower levels of education and income. That trend comes across strongly in the results of the third round of the Life in Transition Survey, where respondents indicated their willingness to take risks on a scale of 1 (maximum risk aversion) to 10 (maximum tolerance of risk).

The results of that survey indicate that individuals on lower incomes and those with fewer years of education are significantly less willing or able to tolerate risks (see Chart 1.17). That may reflect low levels of savings among individuals on lower incomes or a multitude of other factors (since higher incomes may, for example, come as a result of risky choices in the past). The differences in risk aversion across income deciles and by level of education are even more pronounced in the EBRD regions than they are in the advanced economies covered by the survey (Germany and Italy). One way or another, people who are less able to tolerate risks have seen a significant amount of economic risk being shifted onto them.

What is more, periods of major economic upheaval and conflict have, historically, tended to reduce people’s appetite for risk, often leading to greater demand for state intervention. For example, after the Great Depression of the 1930s the state emerged as a major investor (with the New Deal in the United States of America providing for major public investment in transport infrastructure, for instance). Similarly, the Second World War gave rise to the welfare state and a significant expansion in public education and healthcare (with the United Kingdom’s National Health Service being established in 1948). Early survey evidence suggests that the Covid-19 pandemic is having a similar effect.33

The fiscal space used in this report takes account of the level of government debt and net government lending/borrowing as a percentage of GDP, the cost of borrowing, and governments’ ability to raise revenue as measured by the ratio of government revenue to GDP. The administrative capacity index takes account of a measure of e-government (which looks at the scope and quality of online services, the development of telecommunication infrastructure and inherent human capital),34 a Worldwide Governance Indicator measuring the effectiveness of government, a Doing Business indicator assessing the distance to the frontier and an indicator measuring the routine use of bank accounts by the country’s population.

In many countries, the amount of fiscal space has increased over the last two decades (albeit there are a number of notable exceptions, such as Lebanon). This is particularly true of countries where it used to be very limited, with many countries seeing increases in revenue and declines in the cost of servicing public debt, despite higher debt levels.

In contrast with many previous economic crises, the cost of financing has remained low for many economies in 2020. In early June 2020, the yields on the debt of many middle-income economies (including most countries in the EBRD regions) were, if anything, below the average cost of servicing those economies’ debt over the period 2014-19 (as obtained by dividing government interest expenditure by the stock of debt; see Chart 1.19).

Large-scale government assistance programmes can, in turn, significantly raise functional financial inclusion, albeit with a lag of a few months or years. Mongolia, for instance, has introduced universal cash handouts based on future copper royalties, with payments being made into individuals’ bank accounts. As a result, almost 90 per cent of the Mongolian population now use bank accounts, broadly on a par with the levels seen in Latvia and Estonia.

Countries with ample fiscal space but more limited administrative capacity (as is the case, for instance, in parts of the Western Balkans and Central Asia) may need to rely more on one-off universal transfers (as seen, for example, in Kazakhstan, the Kyrgyz Republic, Mongolia, Serbia and Uzbekistan).

Countries with more limited fiscal space but a relatively strong administrative capacity could expand coverage and increase benefit levels under existing targeted support programmes. For instance, Cyprus, Greece and Montenegro have all increased the coverage of existing unemployment benefits and/or enhanced sick leave. Pension increases were one of the most common measures in the early weeks of the Covid-19 crisis in the EBRD regions (being seen in around one-third of those economies), not because pensioners were particularly badly affected by the crisis, but because pension increases could easily be administered at speed.38

Lastly, countries with more limited fiscal space and a relatively weak administrative capacity, such as Lebanon or Tajikistan, may need to rely on policies targeting specific sectors or locations and ensure the direct provision of goods and services to satisfy the basic needs of their populations.

As Box 1.6 shows, individuals who reach adulthood during a pandemic are also around 2 to 4 percentage points more likely to support the expansion of state ownership. A shift of this magnitude in average support for state ownership could result in support for the expansion of private ownership changing from a majority view to a minority view in many economies. Indeed, in a quarter of the economies that participated in the most recent round of the World Values Survey, support for the expansion of state ownership averaged between 45 and 55 per cent of survey respondents.

For instance, state ownership may seek to facilitate the redistribution of income from natural resources or other sources of economic rents, both between individuals and across generations. In that case, the state could be limited to a minority stake. It could also be aimed at securing foreign assets in order to achieve greater diversification. Limiting state investment to minority stakes could also be an effective way of encouraging risk-taking and innovation in specific industries or accumulating state assets in order to fund future liabilities (such as liabilities relating to pension benefits or healthcare in a rapidly ageing economy).

When it comes to addressing job displacement in specific regions or industries as a result of technological change, alternatives to state ownership may involve subsidising employment in the private sector, possibly through income tax credits. Such subsidies could also be directed towards specific groups, such as older individuals. The cost of such assistance could be weighed against the cost of inefficiencies at state-owned enterprises or state agencies. If state bailouts are deployed to see major employers through temporary difficulties, structures could be put in place to facilitate the unwinding of state ownership in the future (see Box 1.7).

Private-sector solutions often require support to be provided through targeted policies and state interventions – referred to as “industrial policy” (see Box 1.8). Those measures could, for example, take the form of regulation and monitoring (in the area of the green economy, for instance, as discussed in Chapter 4), the provision of finance to riskier borrowers (as discussed in Chapter 3), the upgrading of infrastructure, efforts to foster exports and investment, or the establishment of vocational training programmes.

If, after weighing up the various policy options, the state opts for majority state ownership, arrangements need to be put in place to strengthen governance at state-owned enterprises, as discussed further in Chapter 2.

A large percentage of the major airlines in the EBRD regions remain at least partly state-owned. Formal ownership structures vary, with airlines being owned by government ministries (such as the Romanian Ministry of Transport or the Croatian Ministry of State Property), a sovereign wealth fund (Kazakhstan) and a central bank (Lebanon), with some minority stakes being held by airports (Croatia and Romania).

A number of flag carriers have gone bankrupt as a result of the reduction of state support, sometimes after rounds of failed privatisations and re-nationalisations. Examples include Malev in Hungary (which is now largely served by the privately owned Wizz Air) and Cyprus Airways. In both of those cases, bankruptcy was preceded by the European Commission issuing a ruling against the use of state aid. Other examples include Air Armenia, B&H Airlines, Estonian Air, FlyLAL in Lithuania, Slovak Airlines and, as recently as 2019, Adria Airways in Slovenia.

Other carriers, including Air Moldova and Ukraine International Airlines, have been successfully privatised. In Greece, Aegean Airlines bought the previously state-owned Olympic Airlines. Several flag carriers have also turned to foreign partners to help sustain their operations. For instance, Turkish Airlines now owns 49 per cent of Air Albania, while Etihad owns 49 per cent of Air Serbia. Meanwhile, Air Baltic, which has its main hub in Riga, has established secondary hubs in Vilnius and Tallinn, leveraging economies of scale across several relatively small markets.

The airline industry is one of the sectors that have been hit hardest by the pandemic, with demand for air travel falling by around 60 per cent in the first half of 2020 relative to the first half of 2019, according to estimates by the International Air Transport Association (IATA). As a result, many airlines (including privately owned carriers) have sought bailouts totalling between 0.1 and 1 per cent of GDP (see Chart 1.1.1), often accompanied by an increase in state ownership.

State aid has taken various different forms, such as loans with favourable terms, the purchase of minority or majority stakes by the state, and the provision of state guarantees. For instance, the German government has taken a 20 per cent stake in Lufthansa (complete with two seats on the airline’s supervisory board), the Italian government has decided to acquire full ownership of Alitalia, and the Latvian government has decided to increase its stake in Air Baltic from 80 to 91 per cent. The Romanian government, meanwhile, has promised state aid to both the state-owned Tarom and the privately owned low-cost carrier Blue Air. More bailouts may be on the way, given the highly uncertain future of air travel as of mid-2020.

In a couple of cases, bailouts have been accompanied by environmental conditions. Air France-KLM, for example, is now committed to increasing its use of alternative fuels from 0 per cent to 2 per cent by 2025. Similarly, Austrian Airlines is required to reduce its total emissions to less than 70 per cent of the 2005 level and end all flights that are competing with a train journey of under three hours.40

The data come from (i) national sources (as in the case of Albania (1995-2018), Armenia (1998-2018) and Jordan (2000-17)), (ii) an ILO database (as with Belarus (1997-99), North Macedonia (2000-05) and Lithuania (1995-2018)), (iii) labour force surveys compiled by the ILO (as in the case of Bosnia and Herzegovina (2006-19), Tunisia (2005-15) and Egypt (2005-18)), and (iv) country reports produced by the ILO (as with Russia (1995) and Serbia (2001-10)), the IMF (as in the case of Uzbekistan (1992-99), Ukraine (1994-98) and Tajikistan (1990-96)), the Organisation for Economic Co-operation and Development (OECD) (as with Slovenia (1992-96), Turkey (2008) and the Slovak Republic (2008)) and the World Bank (as in the case of Poland (1994)). Those estimates have been cross-checked against the results of representative international household surveys, notably the Life in Transition Surveys conducted by the World Bank and the EBRD in 2006, 2010 and 2016 and the OECD’s Programme for the International Assessment of Adult Competencies (PIAAC).

The estimates in those sources often vary owing to differences in the way that state employment is defined and the way that data are collected. Nonetheless, the correlation between the various available estimates of the public sector’s share of employment tends to be high – between 0.7 and 0.9 across economies and over time. Where different sources have been used for different time periods for the same economy, those estimates have been spliced together using official estimates from national authorities where available and applying changes in levels of state employment derived from other sources. Decisions on the use of individual sources were guided by LiTS and PIAAC data.

In addition to IMF data and national sources such as the US Census Bureau, the long-term data on public employment and government expenditure that are used in this chapter also draw on Edvinsson (2005) for Sweden, Thomas and Dimsdale (2017) and Mitchell (2011) for the United Kingdom, Carter et al. (2006) for the United States of America, and Tansel (2001) for Turkey.

Several experiments have shown that such incentives can change voting behaviour. In one such experiment in California, for example, voters were chosen at random and given either a reminder to vote or the chance to receive a financial reward for voting. An incentive payment of US$ 25 raised turnout by 5 per cent in municipal elections.45 Moreover, it has been shown that people who vote in a single election are substantially more likely to vote again.46 Another option would be to lower the voting age, for instance to 16, as Austria did in 2007.

Within Kazakhstan’s male-dominated energy sector, state-owned enterprises appear to employ fewer women than private firms – not only overall, but also at management level and at board level (see Chart 1.4.1). Indeed, among workers with a technical or vocational education, the largest group of employees, women’s share of employment is around 20 percentage points lower in state-owned enterprises than it is in private firms.

Among engineers and specialists, on the other hand, women’s employment shares are higher in state-owned firms, averaging more than 40 per cent for specialists. When it comes to policies on maternity and paternity entitlements, flexible working arrangements or support for care-related responsibilities, no significant differences are observed between the survey responses of public and private-sector firms. State-owned enterprises are, however, less likely to have human resources policies on sexual harassment and gender-based violence.

Overall, the findings of the survey suggest that there is scope for further cooperation between state owned enterprises and vocational institutes with a view to changing perceptions about the types of job that are suitable for women. Change is happening, though. Kazakhstan’s Gender Action Plan for 2020-22 aims to continue removing regulatory restrictions on women’s employment in specific occupations, including in the energy sector. And building on EBRD assistance, Samruk Energy, the state-owned national power company, has signed up to the UN Women’s Empowerment Principles, including the tracking of sex-disaggregated data and the achievement of the GRI-G4 international standard on gender reporting.

This regression analysis compares attitudes to private and public ownership across individuals with differing degrees of exposure to epidemics during their formative years while taking into account various individual characteristics (X), such as age, year of birth, gender, employment and income decile, for a given country and year. In particular, the following regression is estimated using a linear probability model:

Y_{i c t b} = β₁Exposure(age 18-25)_icb + β₂X_i + β₃Exposure (time of survey)_ct-1 + C_c + T_t + ε_ictb (1)

where Y is a dummy variable capturing whether respondent i in country c, born in year b and interviewed in year t, favours the expansion of state ownership of business over the expansion of private ownership (responses of 6 or higher on a scale of 1 to 10). Exposure to an epidemic while aged between 18 and 25 is measured by the number of individuals affected by an epidemic as a share of the country’s population, averaged over the eight-year window. Regressions also control for any exposure to an epidemic in the year preceding the year of the survey. Specifications also control for country of respondents C and year of survey T.

This analysis reveals that an individual with the highest level of exposure to an epidemic during their formative years (as measured by the affected share of a country’s population) is, on average, 1.7 percentage points more likely to favour the expansion of public ownership than an individual with no exposure to epidemics (see Chart 1.6.1). This effect is larger in high-income countries and economies with stronger democratic institutions, where respondents may expect to have a greater say in how the state manages its assets.

Thus, if history is any guide, the Covid-19 pandemic will lead to a further increase in support for public ownership among people aged between 18 and 25 today – members of what is termed “Generation Z”.

While they address an immediate problem, nationalisations also need to take account of longer term considerations, ensuring that the enterprises in question can be run efficiently and that control is able to revert to the private sector in a transparent manner. The ultimate objective is to ensure that state involvement delivers value for the taxpayer.

In this regard, where bailouts target large listed companies, they could involve instruments such as preferred stock with warrants.51 Unlike common stock, preferred stock does not confer voting rights on the state. In this scenario, therefore, the bailout does not interfere with the running of the company. This feature may be particularly valuable in normally competitive sectors (such as hospitality and transport), where governments may need to bail out multiple players. On the other hand, preferred stock gives the holder a preferential claim on dividends (which could potentially be higher than the dividends on common stock), thereby protecting taxpayers. Warrants – which grant the right to buy common stock at a specified price before a specified date – could provide a further upside for the taxpayer if the bailed out company and its stock price recover. It is also useful to define up front the exit strategy that will be implemented by the state if the industry recovers.

Where nationalisation targets smaller firms, structures similar to private equity funds could be considered, perhaps with private-sector equity funds providing investment in tandem. Such structures may help to ensure that portfolios of smaller companies are run efficiently. The involvement of private-sector co-investors also introduces a market test allowing the implicit or explicit valuation of non-listed firms receiving state aid.

Where nationalisation pursues long-term objectives associated with state ownership, common stock can be used, with a particular focus on the way that state ownership is structured and the enhancement of corporate governance. State asset holdings could benefit from a high degree of operational independence (as enjoyed, for instance, by many sovereign wealth funds). Conditions relating to environmental or social policy objectives need not necessarily be imposed on specific nationalised enterprises that receive assistance, but bailouts may present an opportunity to review regulations and standards in the relevant sectors (as in the case of the air transport industry).

For example, an economy may benefit from people being able to work remotely and pupils being able to study online. However, such solutions only work if most individuals and businesses have reliable broadband access – a good example of a network effect. If broadband providers charge high fees for access in remote areas, it may be that few individuals are willing to pay for that service. That combination of a high price and low demand represents a coordination failure, with social costs far exceeding the cost of providing a reliable internet connection. Likewise, the electric car industry will only take off if consumers can easily charge their cars wherever they go. But at the same time, a private network of charging points can only be established if there are enough consumers. In that case, of course, the coordination failure entails not only significant social costs, but also considerable environmental costs.

Solutions may vary depending on the circumstances. Governments may, for example, compensate service providers directly for any public service obligations that are imposed on them, or they may ask service providers to average the cost of provision across all consumers, thus cross-subsidising some users at the expense of others. Some countries opt for state ownership as a means of delivering on such public service obligations (see Chapter 2).

Another increasingly important area is investment in education and basic research. A person’s private returns to education (which are reflected in a worker’s productivity) are dependent on market opportunities, and they, in turn, are dependent on other people’s educations – a coordination failure that governments are well placed to address. In a virtuous circle, educated workforces help to attract a diversified universe of productive firms. In addition, close partnerships between the scientific community and the private sector foster innovation (with basic research being publicly funded for the most part).

Furthermore, unlike investment in physical capital, knowledge that is developed by one private firm can easily be copied by another for a fraction of the cost of developing it. As a result, the private sector may supply too little knowledge in the absence of government intervention.

Traditionally, a distinction has been drawn between horizontal and vertical industrial policies. In some areas, it is possible, as an alternative to focusing “vertically” on specific firms and industries, to implement a “horizontal” package of measures which seeks to facilitate access to finance for high risk ventures, provide small grants to entrepreneurs on the basis of the competition of ideas, reduce the fixed cost of entry into markets (such as the cost of licensing and permissions), lower information barriers, or leverage companies’ efforts to find new export markets.

In other cases, the lines between horizontal and vertical policies are increasingly becoming blurred. Indeed, any company could, in principle, benefit from the public provision of infrastructure (such as a government-supported network of electric charging points). In practice, however, such measures often benefit specific investors (in this instance, an incumbent developer of electric vehicles). Meanwhile, in the case of vocational training, successful initiatives often involve partnerships with specific private sector investors (as seen, for instance, with efforts to establish a large automotive cluster in Morocco). If the policy package is successful, its benefits may be enjoyed indirectly by the wider region through a pick-up in economic activity, but if it fails, taxpayers will have to pick up the bill.

As such policies target specific technologies or firms (be it explicitly or implicitly), the question of how to ensure that taxpayers get value for money is a matter of constant debate. In some instances, governments may be better than the markets when it comes to predicting future winners, but there is no evidence that they are able to do so consistently. Indeed, market failures may actually be exacerbated by government failures.53

In broad terms, policies targeting specific industries need to be based on a careful assessment of local skill-sets and the quality of economic institutions. For instance, attempts to deepen local supply chains are often pursued by requiring a certain percentage of the inputs used in the manufacturing of, say, a car or a wind turbine to be supplied locally. Such requirements may incentivise companies to reach out to existing or new local suppliers, and they, in turn, may be in a position to adopt the latest technologies, leveraging the scale of the new market open to them and benefiting from training provided by large off-takers. That was the case in Norway, for example, following the discovery of offshore oil and gas. However, if the right skills and incentives are not present, such requirements may also create excess profits for firms supplying substandard products at inflated prices and limit imports of the latest technologies, thus undermining the development of the very industries that the local content requirements were intended to support.54

Calibrating such requirements and gradually phasing them out as local producers become internationally competitive relies on industry regulators being highly independent and highly professional. Thus, governments with significant administrative capacity have far more policy options than those with more limited capacity, as discussed earlier in this chapter. Where administrative capacity is more limited, policy solutions involving state ownership tend to be more common, as discussed in Chapter 2.