Chapter 1. Introduction: Tax, skills, and inclusive growth1

This chapter places this study in the context of OECD work on productivity and inclusive growth, as well as the broader literature on the public finance of education. The importance of skills for growth and productivity, as well as for equality and inclusive growth are all discussed. The impact of the tax system on skills is briefly summarised, and an outline of the study is also provided.

  

1.1. Skills, growth, and productivity

Skills are the cornerstone of building productive economies and inclusive societies. In a world of increasing globalisation and rising inequality, increasing the quality of and access to education has never been a higher priority for policy makers. This study considers how the tax system can affect skills by building indicators that measure the impact of income tax and spending policy on individuals’ incentives to invest in skills.

The nexus between tax, productivity, growth and equity has been the subject of significant study at the OECD in recent years (OECD, 2015a, 2016). Recent work has investigated how tax policy can be used to raise growth levels in the OECD, by shifting the tax mix towards growth-friendly taxes (OECD, 2010b). Other research has also focused on how the tax system can do more to encourage equity and inclusiveness, by examining the whole tax system from a distributional perspective, by improving tax administration, and by removing tax expenditures that mainly benefit those on higher incomes (Brys et. al. 2016; OECD, 2014b). Often, however, tax policies that improve efficiency of the economy run counter to equity considerations, and policies that increase the equity of the tax system may reduce growth. Optimising the tax system for skills investments offers tax policy makers the opportunity to increase both equity and efficiency, to foster growth that offers benefits for all.

Raising skill levels is crucial for increasing economic growth rates and building economies that can provide employment and prosperity. Economic growth will increasingly depend on improvements in productivity (OECD, 2015a). Scare resources, slow population growth, and low levels of investment in physical capital have led to concerns about the future sources of growth across the OECD. Increasing skill levels and boosting productivity is an important response to these concerns: higher productivity means that even in the context of slowing rates of growth of the capital or labour stock in the economy, growth can continue to improve well-being and raise living standards in the OECD (OECD, 2016).

Over recent decades, productivity growth has been slowing. This is a key concern given the importance of productivity growth for improving well-being. Figure 1.1 shows the decline in factor productivity growth across selected OECD countries over the last decade compared to previous decades. Ninety per cent of OECD countries experienced a decline in the trend of labour productivity growth after the turn of the millennium (OECD, 2016). The ability for technological developments to continue to provide strong productivity growth across the OECD is increasingly being called into question, and concerns about a period of secular stagnation across the developed world have been expressed by some policy makers (Summers, 2014). The decline in the growth in productivity raises questions about whether the countries can continue to raise living standards in years to come. There are also questions as to whether future increases in living standards will accrue to a broad spectrum of workers or whether only certain groups will benefit.

Figure 1.1. Multifactor productivity in long run comparative perspective
Annual average growth
picture

Notes: Multifactor productivity growth rates for the period ranges are the annual averages.

Source: OECD Dataset on Growth in GDP per capita and productivity

Raising skill levels can help policy makers meet these challenges. Adequate investment in skills can ensure that all individuals can both contribute to and benefit from productivity growth. While this is true for all workers it is especially vital among those demographics and communities that currently have lowest skills levels. Workers with higher skills are more likely to help firms innovate, to participate in global value chains, and increase the knowledge spillovers from more productive sectors to less productive sectors.

The link between skills and productivity is strengthened by the continuing integration of the global economy. Those who are left without skills are less likely to work in the kinds of industries and companies that participate in global value chains (OECD, 2016). This in turn hampers prospects for future skills development and productivity gains for these individuals; participation in global value chains is a means by which productivity gains in the form of innovations in work practices are passed from firm to firm and from worker to worker. Without adequate skills investment, certain demographics, sectors, and even countries may be increasingly left out of global value chains and may fall further away from the productivity frontier.

Though improving workers’ skills is important for growth, raising the amount of human capital in the economy is about more than just increasing participation in education and lifelong learning. Individuals must develop the kinds of skills that are in demand in the labour market, reducing mismatches between those fields of study chosen by students and those that will yield benefits in the labour market. Individuals must also develop soft skills such as communication and teamwork that are necessary in the modern workplace. Skills that are developed must be activated in the labour market by raising labour market participation. This is especially true among marginalised groups where participation rates are comparatively low, including women, migrants, the elderly, and the disabled. Finally, workers’ skills must be used effectively in the workplace. The right skills must be paired with the right jobs so skills are not under-utilised. There is thus a strong need for better alignment between workers’ skills and those skills demanded in the economy.

1.2. Skills, tax and inclusive growth

The centrality of skills in the current policy environment does not just stem from their important role in boosting productivity and growth; it also stems from the increasing importance of reducing inequality for policy makers (OECD, 2015b). Raising skill levels can make growth fairer, more inclusive, and more durable. Increasing skills across the workforce allows more individuals to participate in the economy and to transition into higher-quality jobs with higher wages.

Figure 1.2 shows that inequality in disposable income has increased in most OECD countries over the last three decades. The Gini coefficient of income inequality stood at 0.29 on average across OECD countries in the mid-1980s. By 2013, it had increased by about 10% or 3 points to 0.32, rising in 17 of the 22 OECD countries for which long-time series are available (Brys et al., 2016).

Figure 1.2. Income inequality increased in most OECD countries
Gini coefficients of disposable income inequality, mid-1980s and 2013, or latest date available
picture

«Little change» in inequality refers to changes of less than 1.5 percentage points. Data year for 2013 (or latest year).

Source: OECD Income Distribution Database (IDD) www.oecd.org/social/income-distribution-database.htm.

In It Together - Why Less Inequality Benefits All (OECD, 2015b).

Previous trends in inequality have been exacerbated by the economic crisis. In many OECD countries, the crisis most affected those who had low savings levels, less secure employment, and low skills: those people who were most vulnerable to economic shocks. Widespread job losses and wage stagnation over this period compounded modest wage growth over previous decades.

The crisis and its aftermath also resulted in straitened public finances across the OECD. Addressing budget deficits in many OECD countries resulted in reductions in the generosity of transfer payments to those on low incomes. In these and other ways, the negative impacts of the crisis were visited most heavily on those with low incomes. Figure 1.3 shows the decline in household real disposable income in the post-crisis period. Across 33 OECD countries, disposable income fell for those with low incomes, median incomes, and for those with high incomes. But on average those with low incomes saw their incomes fall most. By comparison, top earners’ incomes fell on average across the OECD, but by a smaller amount. Indeed, in 15 countries, the real disposable income of the top 10% rose during the crisis period. So the crisis has exacerbated decades-long trends in inequality.

Figure 1.3. Changes in household disposable income by income groups
Annual percentage changes between 2007 and 2011 by income groups, total population
picture

2007 refers to 2006 for Chile and Japan: 2008 for Australia, Germany, Finland, France, Israel, Mexico, Norway and Sweden. 2011 refers to 2009 for Japan; 2010 for Austria, Belgium, Ireland and the United Kingdom and 2012 for Australia, Hungary, Korea and United States; Switzerland is not available. OECD33 refers to the unweighted average.

Source: OECD Income Distribution Database (IDD), www.oecd.org/social/income-distribution-database.htm.

Recent policy debates have also focused on increasing inequalities between capital and labour income. For the vast majority of individuals, wages are by far the largest component of income. However, those on higher incomes earn more of their income from capital: from dividends, capital gains and other forms of business income.

Capital’s share of income has been rising. Figure 1.4 shows changes in labour’s share of total income across OECD countries from 1990 to 2009. While significant heterogeneity across countries exists, on average labour’s share of income has fallen. This means that inequalities have been driven not just by increasing differences in wage levels, but also by divergence in the returns to different factors of production. Those who earn their income from their human capital have seen their share of total income fall relative to those who earn their income from physical capital. Many individuals receive income from both their human capital (through wages) and physical capital (through income from savings). Those with higher incomes have a higher share of physical capital and so have benefited more from the rise in capital’s share of total income. This means that shifts in the returns to different factors of production have exacerbated trends in inequality that are present with respect to wage income.

Figure 1.4. The decline of the labour share in OECD countries, 1990 – 2009
picture

Notes: a) Germany and Iceland: 1991; Estonia: 1993; Poland: 1994; Czech Republic, Greece, Hungary, Slovak Republic and Slovenia: 1995; Israel: 2000. b) Portugal: 2005; Canada and New Zealand: 2006; Australia, Belgium, Ireland, Norway and Sweden: 2007; France, Iceland, Israel, Poland and the United Kingdom: 2008.

Source: OECD calculations based on OECD STAN and EUKLEMS.

OECD Employment Outlook 2012 (OECD, 2012).

These increases in inequality present a combination of challenges for policy makers. Increasing inequality generates more pressure on governments to engage in redistributive spending, to reduce market inequalities using the tax and transfer system. While this can reduce disposable income inequalities, redistribution using the tax and transfer system can have efficiency costs (Brys et al., 2016). High taxes on labour income can reduce work effort and reduce labour market participation. High degrees of welfare spending put pressures on limited government resources at a time of high debt levels. Moreover, increased spending on poverty alleviation and social benefits can create poverty traps. These factors demonstrate that shrinking the gap between market income inequality and disposable income inequality can be costly. The larger the amount of market inequality, the larger those costs can be. This means that inequalities in wages and between capital and labour income are not only problematic in their own right; they are also concerning because of the higher efficiency and growth costs of policy efforts to address them.

One factor that has contributed to the decrease in labour share of income is the favourability with which capital income is taxed relative to labour income across the OECD. Part of the reason for this is the highly mobile nature of capital income. This makes capital income harder to detect, and capital taxes harder to enforce. High taxes on capital income can also negatively impact on savings and investment in physical capital.

In part due to these policy challenges, the increase in inequality across the OECD has taken place at the same time as an overall reduction in the amount of redistribution being undertaken by OECD governments. Figure 1.5 shows the percentage difference between pre- and post-tax Gini coefficient for a selection of OECD countries, as well as the OECD average. This functions as a proxy for the total amount of redistribution occurring in OECD countries. Overall, the reduction in the estimated Gini coefficient caused by the tax and transfer system fell from 29.6% to 26.3% in 2008, before rising slightly to 27.6% in 2012. For a variety of reasons, OECD member states are redistributing less using their tax and transfer systems than was the case in 1999.

Figure 1.5. Redistribution became weaker in most countries until the onset of the crisis
Percentage difference between inequality (measured by the Gini coefficient) of gross market income and inequality of disposable income, working age population
picture

Note: OECD average: un-weighted and based on 10 countries for which data are available at all points (Canada, Denmark, Germany, Israel, Italy, Netherlands, New Zealand, Sweden, United Kingdom and United States).

Source: OECD Income Distribution Database, www.oecd.org/social/income-distribution-database.htm.

Increasing skills can potentially address inequality while at the same time raising growth rates. Recent research has suggested that gaps in human capital may be seen as the most important worldwide determinant of inequality (Blöndal et al., 2002; Sequeira et al., 2014).

Raising skill levels can reduce inequality through a number of channels. Those with high skills are more likely to earn higher wages and to participate in the labour market. Raising wages and employment through higher skills is a key inclusive growth oriented policy goal; it can raise efficiency with modest efficiency costs. Reduced market inequality will also reduce the pressure on governments to undertake redistributive spending, and mean that existing redistributive spending can go further in reducing inequality than might otherwise be the case.

Increased skill levels can also reduce the extent to which inequality is passed down through the generations. Reducing inequality can raise the incomes of low-income families who are most likely to be credit constrained with respect to skills investments, which may make them more likely to invest in skills and in the skills of their children (OECD, 2015b). OECD research suggests that inequality may be associated with greater variation in educational outcomes: an increase in inequality of around six Gini points lowers the probability of poorer people graduating from university by around four points (OECD, 2015b). Similarly, as inequality rises, people from poorer families face much weaker job prospects while there is little change for those from better-off families (OECD, 2015b). More equal societies may by themselves increase education prospects of future generations.

Better skills policies can also make productivity gains more inclusive through the diffusion of innovation. OECD studies on productivity have highlighted gaps between the developments at the productivity frontier and behind the frontier: the gap between those workers, firms and sectors that are highly innovative and have high productivity and those that do not (OECD, 2015a). A key means by which high-performance work practices pass from the frontier to the rest of the economy is through movement of workers through churn in the labour market. However, existing skills policies may hamper this movement. As will be discussed in Chapter 5, many OECD countries currently provide tax support for skills investments that are related to current work, but do not provide similar support for skills investments for workers seeking to change career. In doing so, they may reduce the amount of churn of workers through the labour market. This in turn may reduce the diffusion of skills, raising mismatch levels and reducing spread of innovation.

The positive impacts of higher skills on inequality are part of the reason why access to education and training has undergone a dramatic expansion over the last sixty years. More recent years have seen access to education continue to expand, especially in countries where education rates are low (see Figure 1.6). Some research suggests, however, that expansion in education has resulted in a decline in educational quality (OECD, 2016). In spite of the expansion of educational opportunities, skills gaps remain even among younger cohorts. Even amongst those with ready access to education, the pace of technological change raises concerns of a ‘digital divide’ between those who have the skills to participate in a digitalised knowledge economy and those who do not.

Figure 1.6. Trends in enrolment rates of 15-19 and 20-29 year-olds (2005-2013)
Students in full-time and part-time programmes, in both public and private institutions
picture

Source: (OECD, 2015c), Education at a Glance, Data for Germany are for 2006 instead of 2005. Data for Luxembourg are underestimated because many resident students go to school in the neighbouring countries.

In addition, education outcomes for children are still strongly associated with the education levels of their parents: those with more educated parents are more likely to be educated themselves. Educational advantage and disadvantage are propagated throughout the lifecycle – the education systems in many countries are not reducing intergenerational replication of inequality as much as they could be. Figure 1.7 shows the shares of students in OECD countries who match, exceed, or do not exceed the education level of their parents. In most countries, the levels of upward mobility are low: more than half of students achieve the same or a lower level of education than their parents.

Figure 1.7. Intergenerational mobility in education (2012)
Survey of Adult Skills, educational attainment of 25-34 year-old non-students compared with their parents
picture

Source: (OECD, 2015c), Education at a Glance. Countries are ranked in descending order of upward mobility to tertiary education among tertiary-educated 25-34 year-old non-students.

1.3. The returns to skills

Increasing skills are a vital mechanism to address important policy challenges in OECD countries: lower productivity and higher inequality. However there remains much debate in the academic community and in the policy literature regarding what constitutes the right tax and spending policy mix when it comes to education and training. It is not clear what the optimal amount of total spending on skills should be, or how this spending should vary with existing skill levels and with economic development. In addition, the mix of spending between the public and private sectors is the subject of much debate. Not all skills investments are equal: the extent to which skills spending should focus on soft skills is debated, as is the extent to which spending should focus on early childhood education, lifelong learning, or tertiary, secondary or primary education. Debates also exist about the extent to which government spending on skills should encourage skills investments in certain areas such as STEM skills, and how the risks of skills investments should be shared between individuals, governments and firms.

Choosing the right skills policies requires a comprehensive assessment of the costs and benefits of skills investments. A key return to skills is higher wages. Those with better skills are more productive in the workplace and can demand higher wages from their employers. Recent OECD work based on the Survey of Adult Skills has shown that not only are wages higher for those who have spent more years in education, but they are also higher for those with better literacy, numeracy and problem solving skills. Figure 1.8 shows the distribution of wages by literacy proficiency level. There is a seven USD hourly wage gap between the wage levels of those with a literacy level in the lowest of five literacy categories compared to those with a literacy level in the highest category.

Figure 1.8. Distribution of wages, by literacy proficiency level
Hourly wages at 25th, 50th and 75th percentiles of the wage distribution, USD PPP
picture

Notes: Employees only. Hourly wages, including bonuses, are expressed in purchasing-power-parity-adjusted USD

Source: (OECD, 2013) OECD Skills Outlook – First Results from the Survey of Adult Skills (PIAAC), Table A6.4 (L).

While the data on the current wage premium earned by those with higher skills is clear, assessment of the future path of wages is more challenging. Some research suggests that the returns to skills investments may be lower than what is suggested by current wage levels. For example, in some OECD countries, those with tertiary degrees earn a significant earnings premium over those without tertiary degrees, pointing to a significant skills shortage in these countries. This suggests that the returns to skills investments may be high in these countries. However, as more and more people become educated and skills shortages are reduced, this wage premium may fall. Indeed, some of the OECD countries with the lowest wage premiums for tertiary education are countries where tertiary education is most widely available. So a dynamic approach to assessing the returns to skills investments may yield lower estimates than a static approach such as the one taken in this study (Heckman and Jacobs, 2010).

At the same time, there is a large economic literature debating the extent of skill-biased technological change across the OECD. This literature suggests that technological advances have raised the wage premium for certain kinds of skills while reducing the wages available in the labour market for many low-skilled workers. Jobs requiring routinised activities that formerly would have been undertaken by low-skilled workers have been automated, leading to a decline in wages for these workers. Figure 1.9 shows the evolution of employment across occupations of various skill levels (OECD, 2013). These results suggest that availability of low and medium-skilled jobs may continue to decline, suggesting in turn that the returns to high-skilled work relative to low-skilled work may continue to rise. This highlights the complexity of assessing the future returns to skills in the form of wages.

Figure 1.9. Evolution of employment in occupational groups defined by level of education
Percentage change in the share of employment relative to 1998, by occupational groups defined by workers’ average level of education
picture

Notes: Only the 24 OECD countries available in the 1998 LFS database are included in the analysis. High level of education refers to tertiary level or more than 15 years of schooling; medium level of education refers to no tertiary but at least upper secondary education or around 12 years of schooling; low level of education refers to less than upper secondary education or 11 years of schooling. Occupations with high-educated workers: legislators and senior officials; corporate managers; physical, mathematical and engineering science professionals; life science and health professionals; teaching professionals; other professionals; physical and engineering science associate professionals; life science and health associate professionals; teaching associate professionals; and other associate professionals. Occupations with medium-educated workers: managers of small enterprises; office clerks; customer services clerks; personal and protective services workers; models, salespersons and demonstrators; extraction and building trades workers; metal, machinery and related trades workers; precision, handicraft, craft printing and related trades workers; stationary plant and related operators; and drivers and mobile plant operators. Occupations with low-educated workers: other craft and related trades workers; machine operators and assemblers; sales and services elementary occupations; and labourers in mining, construction, manufacturing and transport.

Source: (OECD, 2013) OECD Skills Outlook – First Results from the Survey of Adult Skills (PIAAC).

The returns to skills may change over time due to the depreciation of skills assets. OECD research has suggested that skills level rise over time before beginning to sharply depreciate as workers approach their mid-40s (OECD, 2013). This depreciation of skills that are not kept up-to-date, especially in technology rich-environments, may see wage levels fall as workers grow older. On the other hand, there is evidence that more highly skilled workers not only earn higher wages immediately after upskilling, but also see their wages increase at a faster pace compared to their low-skilled counterparts (Hanushek et al., 2013). This means that assessing the impact of skills on the future path of wages is subject to a significant degree of uncertainty.

There is also significant heterogeneity in the returns to education in the form of future wages. The returns to education may be higher for individuals with higher natural abilities. Returns may be higher for individuals who are already well-educated, or for those who are poorly-educated. The returns to education in subjects such as science and engineering may be higher than the returns to other fields. The returns to education may be higher at a younger age than at older ages. The literature suggests that the returns to early childhood education in the form of future incomes are very high, in part due to higher abilities to learn new skills at very young ages (Heckman and Jacobs, 2010). At older ages, abilities to learn new skills may be diminished. In addition, skills investments in later life have fewer years to earn returns compared to skills investments earlier in life. For example, those who are near retirement may not see sufficient returns to their skills investments to make these investments worthwhile.

The returns to skills investments are also not confined to wages and employment prospects. This study focuses on higher wages as a measure of the returns to skills investments. However, many other forms of financial and non-financial benefits also result. Those with higher skills are likely to work longer, raising their lifetime income and reducing the demographic pressures on pension systems. Those with higher skills are also less likely to leave the labour market or become unemployed, further increasing the returns to skills investments. There are also other non-financial benefits to investing in skills. Those with higher skills are more likely to report being in good health, potentially reducing the pressure on public health systems. Skills can have positive impacts on other aspects of well-being as well (Hanushek et al., 2013; OECD, 2013).

The value of these various benefits from skills on government finances, on the economy, and for individuals is challenging to quantify. Estimates of the positive impact of skills investments that rely on wages alone such as those in this study likely underestimate the true benefits of these investments (de la Fuente and Jimeno, 2008). This study takes a step towards such an assessment by examining the financial costs and benefits of certain stylised skills investments in the OECD. In assessing the returns to skills investments, the study focuses on tertiary education and lifelong learning. The key return to tertiary education considered is higher wages. Consideration of the impacts on employment and labour market participation is not factored into the analysis, nor are the broader impacts on trust, health, crime and well-being. From a government’s perspective, estimates of the future revenue impacts are confined to increases in income tax revenue; other positive economic impacts are not analysed. Much more work in this area is needed, and so the results presented in this study should be carefully interpreted.

1.4. Public finance of education

The previous discussion has outlined that the returns to governments, firms and individuals from skills investments are substantial. However, a cost-benefit analysis of skills policies also requires a discussion of the costs of investment in skills, and who bears these costs. Figure 1.10 outlines how the direct costs of education from primary to tertiary level (i.e. costs excluding lost earnings) are apportioned between governments and individuals in OECD countries. Education spending as a share of GDP varies substantially. So too does the share of spending accounted for by governments and by private actors such as firms and individuals. On average, direct education spending comprised 5.3% of GDP in the OECD in 2012. On average spending comprising 4.6% of GDP was carried out by governments, and 0.7% by private actors. The largest amount of private spending is in Chile, where 37% of all education spending is carried out by private actors. In some European countries, however, very little direct education spending at primary to tertiary level is undertaken by private actors.

Figure 1.10. Spending on primary-tertiary educational institutions as a share of GDP
picture

Source: (OECD, 2015c), Education at a Glance. Public expenditure only (for Switzerland, in tertiary education only; for Norway, in primary, secondary and post-secondary non-tertiary education only). Countries are ranked in descending order of expenditure from both public and private sources on educational institutions.

Given the large private benefits from skills investments, government subsidies for skills investments are substantial. There are two principal motivating factors behind government intervention into the provision of public capital goods such as education. The first is the presence of market externalities. As discussed in Section 1.3, the returns to skills investments go far beyond the narrow returns to individuals in the form of wage increases or even broader individual returns such as increased well-being and life expectancy. There are social returns as well, in the form of increased growth, increased innovation, and reduced crime. Given that these benefits are not internalised by individuals making skills investment decisions, they may underinvest in skills relative to the socially optimal level. Extra government spending resolves this externality raising the incentives to invest in skills through subsidies (Dur and Teulings, 2003; Heckman and Jacobs, 2010).

The second principal motivating factor behind government intervention into the provision of education relates to imperfections in the capital market for education. Human capital is different from physical capital in that human capital is not transferable from person to person. Physical capital assets can be sold by their owners, the costs of transferring them from person to person is low. This means that physical capital assets can be offered as collateral on debt to finance such investments. The same is not true for human capital investments; if a debt is incurred to finance a skills investment, the skills cannot be recouped in the event of non-repayment of the debt. This may make lenders reluctant to provide financing for human capital investment, in turn meaning that profitable human capital investments do not proceed for lack of finance (Dur and Teulings, 2003; Lochner and Monge-Naranjo, 2002).

These particular features of skills investments mean that financing skills investments presents challenges for individuals, firms and governments. For individuals, skills investments may be difficult to finance because it may be difficult or impossible to use skills as collateral. Capital markets do not function as effectively for human capital investment as they do for physical capital investment. This means that profitable skills investments are more likely to not occur compared to physical capital investments.

Firms face financing pressures as well. Skills investments by business are a key form of spending on lifelong learning. However firms may underinvest in the skills of their employees because of concerns that newly-skilled employees may leave or be poached by rival firms. It is difficult for firms to account for the depreciation of their human capital assets in the same way that they usually account for the depreciation of their physical capital assets. Many tax systems attempt to counteract these difficulties by allowing skills expenses to be immediately deducted from the personal and corporate tax bases, but such provisions are of benefit mainly to firms that are highly profitable: firms with low profits may not benefit from such provisions. In addition, investments in skills may require significant sunk costs for firms, which may be challenging for credit constrained firms, especially SMEs.

In addition, governments face financing challenges in providing support for skills. Demographic pressures, low growth rates, and increased difficulties in taxing mobile factors of production including labour and capital all have increased the financial pressures facing governments in recent years. Many OECD countries have high debt and deficit levels. At the same time, there are increasing pressures for governments to finance new forms of educational investments in early childhood education and in lifelong learning.

Governments may also face challenges due to increased taxpayer mobility. Countries that attempt to keep private education costs low through extensive government support may try to recoup the costs of this support through higher taxes. However this may result in well-educated workers emigrating to lower tax countries. Individuals may also immigrate to countries where private education costs are low to take advantage of generous education support. In such instances, governments can face similar poaching dilemmas, reducing skills investments relative to a social optimum.

This study analyses the ways in which the costs and benefits of skills investments are shared across society. A key conclusion in the study is that in many cases the market for financing skills investments does not work. This may mean that risks and returns for skills investments are not shared in proportion to costs. Risky skills investments may not be undertaken due to lack of access to finance for skills, or lack of insurance against skills outcomes. This in turn reduces the positive impacts that skills can have on inequality, on productivity, and on growth.

1.5. Tax, skills, and financial incentives

A key policy lever by which the government intervenes in skills financing decisions is through the tax system. The tax system is widely regarded as a key lever in affecting many important objectives of OECD governments: raising physical capital investment, reducing inequality, raising employment, increasing R&D activity. However the impact of the tax system on human capital investment has not been at the centre of tax policy making.

Tax and education spending policies need to be examined in a holistic way. Education policy makers cannot rely on education spending levels alone to assess the skills-friendliness of their system from a financial perspective. Nor can tax policy makers examine their tax system solely from a revenue-raising perspective, or even simply from a combination of revenue-raising, labour market activation, and redistribution perspectives. Instead, the efficiency and equity consequences of tax policies from a skills perspective are a key aspect of tax policy and tax design.

Optimising the tax system from a skills perspective could positively affect economic and social outcomes through a number of channels. It may reduce the need for redistribution through the tax system by reducing the inequality of market income. In doing so, raising skill levels can alleviate the distortions that come with trying to reduce inequality through the tax and transfer system (Bovenberg and Jacobs, 2005). Better tax and skills policies can also potentially reduce the need for costly education spending by governments. In addition, raising the stock of human capital may increase the growth-friendliness of the tax system overall.

The impact of the tax system on skills investments is complex. The first and most obvious impact is that progressive income taxation reduces the returns to skills investments from the perspective of an individual because their wages – their returns to skills investments – are taxed away. As taxpayers earn higher and higher returns from their skills investments, progressive tax systems reduce the returns to skills investments at an increasing rate (Cameron and Heckman, 1999).

The tax system also reduces the costs of skills investments for the individual. A key input into a skills investment is the individual’s time: while studying individuals’ earning capacity is diminished. Foregone earnings are for many individuals a larger cost component of a skills investment than direct costs such as tuition fees. But as individuals earn less their tax liability falls in progressive income tax systems. This reduction in tax liability offsets lost earnings: in this way the tax system reduces the cost of skills investments. The tax system can also reduce the costs of skills investments through tax expenditures that reduce an individual’s tax liability in proportion to the direct costs of education, as is discussed further below. Taxes also impact the financial incentives to invest in skills by influencing the costs of the financing of skills investments. Many OECD countries allow financing costs to be deducted from the tax base, and so the tax system can provide added support to skills investments financed with debt. All of these channels through which the tax system impacts the costs of skills investments are modelled in this study.

This study focuses on the ways in which the tax system affects the financial decision to invest in skills through the personal income tax (PIT) and social security contribution (SSC) system. But other parts of the tax system may matter for skills investments as well. For example, countries may provide VAT relief to education providers or educational institutions. Skills expenditures by firms are generally deductible from the corporate income tax base. Taxes on savings can reduce the opportunity costs of skills investments by making physical capital investment less attractive (D’Andria and Mastromatteo, 2012; Jacobs and Bovenberg, 2010). These and other non-PIT and SSC provisions to encourage skills investments are discussed further in Torres (2012).

Taxes also affect the supply of skills through their impact on work effort and labour market participation and the demand for skills, which will impact the level of unemployment. The impact of the PIT and SSC systems modelled in this study focuses on the ways in which increased taxation can reduce incentives to invest in skills by taxing away higher wages. But high taxes on labour may result in reduced labour market participation. This in turn reduces the incentives to invest in skills and may reduce skills investment.

All taxes distort economic activity. The challenge for tax policy makers is to design the tax system in a way that minimises these distortions as much as possible. The literature on optimal taxation has argued that reduced taxation of physical capital encourages investment and raises growth (Chamley, 1986; Judd, 1985). The effective tax rates discussed in this tax policy study are based on the effective tax rate methodology for physical capital developed by Devereux and Griffith (2003). Concerns about the negative impacts of physical capital investment have seen statutory tax rates on physical capital fall over recent decades (Brys et al., 2016).

In recent years however, attention in the theoretical literature on optimal taxation has increasingly turned to how optimal taxes should be considered in light of investment in human capital (Brys and Torres, 2013; Gottardi et al., 2014; Krueger and Ludwig, 2013; Stantcheva, 2015; Torres, 2012). This literature has argued that the taxation of physical and human capital should be more closely aligned. Many of these recent studies recognise the differences between physical and human capital; including that human capital cannot be offered as collateral, and that it may be difficult to design efficient contracts that insure against the risks of human capital investment. Given the centrality of skills for policy outcomes such as productivity and inclusive growth, it is important that the impact of taxes on skills is taken into account by policy makers.

This study provides empirical grounding for many of the theoretical insights from this literature, and highlights how the details of tax systems in different countries can result in variation in the incentives to invest in skills depending on the taxpayers’ age, income, and family status, as well as on the duration, nature and costs of a skills investment. In doing so, the study aims to nuance the discussion of the impact of taxes and skills relative to the theoretical academic literature, and place consideration of skills investments at the centre of the design of tax systems for policy makers.

1.6. Plan of the study

The study proceeds as follows. Chapter 2 outlines tax expenditures used to support skills investment (skills tax expenditures, or STEs) in the 29 countries examined in this study. The year of analysis is 2011. Incorporating these STEs into a broader analysis of the overall impact of the tax and spending system on skills is a key innovation of this study. There is a substantial body of research on the impact of education spending on skills outcomes. However, government spending on skills that occurs through the tax system has not been studied in a comparative way.

There are several kinds of STEs. Some countries provide tax credits and tax allowances that allow tax liability to be reduced in proportion to skills expenditures. Some countries reduce the amount of labour taxation levied on student wage income, or on student scholarship income. Student wage income or scholarship income can also be subject to relief regarding social security contributions. Some tax systems can also offer tax relief in proportion to the amount of debt individuals incur as part of their skills investment. These tax expenditures are all discussed in Chapter 2.

Chapter 3 outlines the core methodology for designing the tax and skills statistics developed in this study. Three main indicators are developed:

  • A Breakeven Earnings Increment (BEI); which measures how much earnings need to increase for an individual after their skills investment so that they earn back the costs of that investment over their remaining years in the workforce.

  • An Effective Tax Rate on Skills, which measures how much taxes increase or reduce the net returns to skills investments for an individual. Effective tax rates are developed for different estimates of the returns to education. The Marginal Effective Tax Rate on Skills (METR) examines the case of an individual just breaking even on a skills investment. It measures the extent to which the BEI is increased by the tax system: how much it rises compared to a world without taxes. The Average Effective Tax Rate on Skills (AETR) examines the case of an individual who earns a higher-than-breakeven return. This indicator measures the difference in the net present value of education between a world with and without taxes.

  • The third indicator measures the returns to skills investments for governments, comparing the government’s costs of educating an individual to the government’s expected returns in the form of higher future tax revenues. This indicator is referred to as a Returns to Costs Ratio (RCR). As with the Effective Tax Rates on Skills, break-even and higher-return scenarios are considered, so a Marginal Returns to Costs Ratio (MRCR) and Average Returns to Costs Ratio (ARCR) are both developed.

The chapter also describes how to interpret the indicators, how they relate to each other, and how they change in response to changes in educational spending, student income, and tax rates before education, during education and after education.

Chapter 4 outlines the main results for the BEI, METR, AETR, MRCR, and ARCR for four stylised skills investment scenarios: a 17-year-old student undertaking a four-year degree, a 27 year-old undertaking a one-year degree, a 32-year old undertaking a short course of job-related training, and a 50-year old undertaking a one-year degree. These examples are chosen to be representative of the different kinds of post-secondary skills investments commonly undertaken in the OECD. This chapter also discusses how these different indicators depend on the returns to education, on the countries concerned, on the individual’s income level, on the individual’s age, and on how the skills investment is financed.

Chapter 5 features an analysis of the specific STEs outlined in Chapter 2. The overall value of these STEs in some of the example cases in Chapter 4 is considered. Chapter 5 also examines the impact of these STEs on the results presented in Chapter 4. The chapter also discusses the relative impacts of tax-based means of encouraging skills investment compared to non-tax means of encouraging skills investment such as reducing tuition fees, expanding loan support to students, and increasing scholarships and grants. The chapter concludes with a discussion of the overall impact of the tax system on skills investment.

Four annexes are provided. Annex A outlines in greater detail the methodology behind the indicators: a formalised version of Chapter 3. Previous work by Brys & Torres (2013) outlined a formal methodology for defining the BEI and the METR on Skills. This Annex expands on their work, extending the discussion of the METR and also showing how the same approach can be used to define the AETR on Skills as well as the MRCR and ARCR. The Annex also outlines how the Brys and Torres methodology can be expanded to include consideration of student debt.

The remaining three annexes provide country-specific tables and information. Annex B provides cross-country comparisons of the key indicators, the BEI, the METR, the AETR and the MRCR for the stylised skills investment examples outlined in Chapter 4. Annex C outlines key details of the STEs modelled for each country. This section is based on the discussion in Torres (2012). Annex D provides tables of the key tax and skills results for the stylised education scenarios in Chapter 4, as well as selected other results on a country-by-country basis. Extensive details of the various factors making up the indicators are provided, corresponding to the equations outlined in Chapter 3 and Annex A.

References

Blöndal, S., S. Field, and N. Girouard, (2002), “Investment in human capital through upper-secondary and tertiary education”, OECD Economic Studies, Vol. 2002/1, OECD Publishing, Paris, http://dx.doi.org/10.1787/eco_studies-v2002-art3-en.

Bovenberg, A.L. and B. Jacobs, (2005), “Redistribution and education subsidies are Siamese twins”, Journal of Public Economics, Vol. 89, pp. 2005–2035.

Brys, B., S. Perret, A. Thomas, and P. O’Reilly, (2016), “Tax Design for Inclusive Economic Growth”, OECD Tax Policy Working Papers, No. 26, OECD Publishing, Paris, http://dx.doi.org/10.1787/5jlv74ggk0g7-en.

Brys, B. and C. Torres, (2013), “Effective Personal Tax Rates on Marginal Skills Investments in OECD Countries: A New Methodology”, OECD Taxation Working Papers, No. 16, OECD Publishing, Paris, https://doi.org/10.1787/22235558.

Cameron, S. V. and J.J. Heckman, (1999), “Can Tuition Policy Combat Rising Wage Inequality”, in Kosters, M.H. (Ed.), Financing College Tuition: Government Policies and Educational Priorities, The AEI Press, Washington, D.C.

Chamley, C. (1986), “Optimal taxation of capital income in general equilibrium with infinite lives.”, Econometrica, Vol. 54, pp. 607–622.

D’Andria, D. and G. Mastromatteo, (2012), “Fiscal Policy Impacts on Growth: an OECD Cross-Country Study with an Emphasis on Human Capital Accumulation”, CEIS Tor Vergata Research Paper, Vol. 10 No. 217.

Devereux, M.P. and R. Griffith, (2003), “Evaluating Tax Policy Decisions for Location Decisions”, International Tax and Public Finance, Vol. 10, pp. 107–126.

Dur, R.A.J. and C.N. Teulings, (2003), “Are Education Subsidies an Efficient Redistributive Device?”, in Agell, J., Keen, M. and Weichenrieder, A. (Eds.), Labor Market, Institutions and Public Regulation, M.I.T. Press, Cambridge (MA).

Gottardi, P., A. Kajii, and T. Nakajima, (2014), “Optimal Taxation and Debt with Uninsurable Risks to Human Capital Accumulation”, Federal Reserve Bank of Atlanta Working Paper Series, No. 2014–24.

Hanushek, E.A., G. Schwerdt, S. Wiederhold, and L. Woessmann, (2013), “Returns to Skills around the World - Evidence from PIAAC”, NBER Working Paper, No. 19762.

Heckman, J.J. and B. Jacobs, (2010), “Policies to Create and Destroy Human Capital in Europe”, NBER Working Paper, No. 15742.

Jacobs, B. and A.L. Bovenberg, (2010), “Human capital and optimal positive taxation of capital income”, International Tax and Public Finance, Vol. 17, pp. 451–478.

Judd, K. (1985), “Redistributive taxation in a simple perfect foresight model”, Journal of Public Economics, Vol. 29, pp. 59–83.

Krueger, D. and A. Ludwig, (2013), “Optimal Progressive Labor Income Taxation and Education Subsidies When Education Decisions and Intergenerational Transfers are Endogenous”, American Economic Review: Papers & Proceedings, Vol. 103 No. 3, pp. 496–501.

de la Fuente, A. and J.F. Jimeno, (2008), “The Private and Fiscal Returns to Schooling and the Effect of Public Policies on Private Incentives to Invest in Education: a General Framework and Some Results for the EU”, Barcelona Graduate School of Economics Working Paper, No. 342.

Lochner, L. and A. Monge-Naranjo, (2002), “Human Capital Formation with Endogenous Credit Constraints”, NBER Working Paper, No. 8815.

OECD (2010), “Tax Policy Reform and Economic Growth”, OECD Tax Policy Studies, No. 20, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264091085-en.

OECD (2012), Employment Outlook 2012, OECD Publishing, Paris, http://dx.doi.org/10.1787/empl_outlook-2012-en.

OECD (2013), OECD Skills Outlook 2013 - First Results from the Survey of Adult Skills, OECD Publishing, Paris, https://doi.org/10.1787/9789264204256-en.

OECD (2014), The Distributional Effects of Consumption Taxes in OECD Countries, OECD Tax Policy Studies, No. 22, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264224520-en.

OECD (2015a), The Future of Productivity, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264248533-en.

OECD (2015b), In It Together: Why Less Inequality Benefits All, OECD Publishing, OECD Publishing, Paris, https://doi.org/10.1787/9789264235120-en.

OECD (2015c), Education at a Glance 2015: OECD Indicators, OECD Publishing, Paris, http://dx.doi.org/10.1787/eag-2015-en.

OECD (2016), The Productivity-Inclusiveness Nexus, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264258303-en.

Sequeira, T., M. Santos, and A. Ferreira-Lopes, (2014), “Income Inequality, TFP, and Human Capital”, Unpublished Working Paper.

Stantcheva, S. (2015), “Optimal Taxation and Human Capital Policies over the Life Cycle”, NBER Working Paper, No. 21207.

Summers, L.H. (2014), “U.S. Economic Prospects: Secular Stagnation, Hysteresis, and the Zero Lower Bound”, Business Economics, Nature Publishing Group, Vol. 49 No. 2, pp. 65–73.

Torres, C. (2012), “Taxes and Investment in Skills”, OECD Taxation Working Papers, No. 13, OECD Publishing, Paris, http://dx.doi.org/10.1787/5k92sn0qv5mp-en.

Note

← 1. The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.