Chapter 2. A view on taxes, social protection and informality in Latin America using Taxing Wages modelling1

1. Introduction

This special feature explores some of the relationships existing in Latin American (LAC) countries between taxes on wages (specifically social security contributions), characteristics of labour markets and levels of informality and identifies some of the outcomes of the interaction between tax policy and labour markets. Section 2 considers the structure of tax revenues in LAC countries and in particular the weight that personal income tax (PIT), social security contributions (SSCs) and payroll taxes have in the region’s total tax revenues. Section 3 describes the main labour market outcomes in the region during the last decade. Section 4 reviews one the most important characteristics of labour markets in LAC countries: informality. The Taxing Wages models developed for the LAC countries have been used to estimate a theoretical cost for workers to move from the informal to the formal economy (a proxy of the formalisation cost). This is defined as the amount of SSCs workers would have to pay, given the labour income they declare to earn, if they decided to register and comply with tax and labour regulations. Section 5 sets out some conclusions.

The gains attained during the last decade from increased fiscal revenues and improved labour market outcomes are being jeopardized by the current lacklustre economic growth in LAC countries. The commodity super cycle that propelled much of the economic growth over the period has come to an end. Low commodity prices coupled with weak global growth prospects and tight financing conditions are undermining the region’s growth prospects. On average, gross domestic product (GDP) levels are expected to fall in 2016 for the second year in a row, before recovering slightly in the following year (OECD/CAF/ECLAC, forthcoming).

The downturn in economic activity is having a negative impact on the labour markets in the region, enhancing the risks of long term detrimental effects on workers in the lower halves of the labour income distributions. On average, unemployment is increasing and job quality is diminishing. In addition, wages are stagnant and informal sectors continue to represent a large portion of economic activity (OECD/ECLAC/CAF, forthcoming; Alaimo et al., 2015). Within this framework, the role of taxes and their implications for labour markets are becoming increasingly relevant.

Tax systems in LAC countries rely heavily on consumption taxes while taxes on labour income have a more limited share. In 2013, consumption taxes accounted for almost half of tax revenues in LAC countries whereas the share of taxes on labour income was about one quarter. Of the latter the major share is accounted for by SSCs at around 16% of the total with less than 10% from personal income tax (PITs). However, despite their relatively low share of total revenues, PITs and SSCs do impact on the levels of total labour costs and the gaps between gross and net wages for workers.

There is evidence of a correlation between the cost of becoming formal and the levels of informality particularly at the lower end and middle range of earnings distribution. High theoretical formalisation costs are correlated with high levels of informality. In Latin America, some authors have highlighted the negative effects that high tax burdens might have on the incentives for transition between informality and formality (Loayza, et al., 2005; De Soto, 1989; Melguizo and González-Páramo, 2012). Similarly, others have illustrated the positive impact that reductions in tax rates and labour costs could have on economic formality (Levy, 2008; Pagés, 2010). This chapter uses the new data on taxes on labour income for the countries in this Report to explore in further detail the relationship between informality levels and formalisation costs, as proxied by the SSCs workers would have to pay to become or remain formal.

2. Taxation in Latin America: A work in progress

LAC economies are characterised by relatively low tax revenues, with heterogeneity across countries. Tax revenues as a share of GDP in LAC countries averaged 21.7% in 2014, considerably lower than the OECD average of 34.4% of GDP. Tax revenues diverge considerably across the region. In 2014, the tax to GDP ratios ranged from 33.4% in Brazil and 32.2% in Argentina to 12.6% in Guatemala and 14.1% in the Dominican Republic (Figure 2.1).

The gap between the LAC region and the OECD economies is mostly explained by the region’s relatively low direct-tax take (especially PIT and SSCs). OECD PIT revenues are higher than in the LAC region (8.8% compared with an estimated 2% of GDP, respectively in 2013). SSCs, on average across the OECD, account for 9.1% of GDP while the LAC average is only 3.7%. Payroll tax yields are relatively small for both groups (0.1% of GDP in LAC and 0.4% in OECD economies).

The relatively low levels of revenues from PIT can be explained by a range of factors, from differing economic characteristics in LAC countries compared with the OECD, to tax design, and even tax avoidance and evasion. First, the share of labour income in GDP is substantially lower in LAC countries (Daude, Melguizo and Neut, 2010). Secondly, PIT minimum thresholds are relatively high compared with average wages resulting in the vast majority of the working population having taxable incomes below these thresholds. Thirdly, tax expenditures tend to be relatively high in the region, as many economies have generous allowances for a range of items such as mortgage interest deductions, education and health expenses in addition to other justified allowances such as employee SSCs. Fourthly, the PIT base is eroded by high levels of tax avoidance and evasion that combined with high levels of informality decimate the ability of the tax to raise revenue. Some authors argue that about half of the potential revenue from PIT is lost through evasion (IDB, 2013). Finally, high PIT privileges are granted on capital income to encourage investment.

The relatively low levels of SSCs stem both from low levels of social insurance coverage and the more significant role of the private sector as a social protection provider. In many LAC countries, social security systems did not aim to cover all households but were only targeted at employed salaried workers (Bosch, Melguizo and Pagés, 2013). From the 1980s onwards, many LAC countries reformed their social security systems and introduced fully capitalized accounts for pension plans and privatised healthcare. Chile, El Salvador and Mexico for example have shifted their public pension programmes to fully privately funded pension systems. In Colombia and Peru, the private and public programmes co-exist and employees opt for either one or the other (OECD/ECLAC/CIAT/IDB, 2015). Argentina, Brazil and Uruguay also follow a model where the public and private systems are complementary and substantial reforms have been undertaken in these countries to extend the coverage of social security systems to those previously excluded (ECLAC, 2014). Across the OECD, SSCs to privately managed funds exist in several countries, but the main emphasis is on the public provision of these services which requires higher revenue receipts, particularly in Europe.

In summary, as shown in Chapter 3 of this report, while PIT is only paid at very high levels of income, the fact that SSCs are only levied on salaried workers results in a difference between gross and net wages that might provide an incentive towards informality for these workers, thus affecting the outcomes on labour markets.

3. Labour markets in Latin America: Informal is normal

The last ten years have seen several positive developments in Latin American labour markets, although these could be challenged or even reversed in the current macro-economic context. Between 2003 and 2013 there have been substantial job creation, real wage increases and reductions in unemployment and informality. In this period, the region’s economies were able to employ a larger number of workers, with the employed population increasing by 4 percentage points on average. At the same time, real wages increased by an average of 2% per year, unemployment fell from 9% to 6.1% and informality decreased to 55%. Nevertheless, the current economic slowdown could reverse some of these achievements. For example, in 2015, the rate of unemployment increased for the first time since 2009 (ECLAC/ILO, 2016).

LAC averages hide wide disparities across countries, and also within countries by income level. In 2013, unemployment rates for Bolivia, Ecuador, El Salvador and Guatemala were below 4%. At the other end of the distribution, Barbados, Costa Rica, Colombia and Jamaica faced unemployment rates above 8%, with the Caribbean countries experiencing the highest rates, at 11.3% and 14.7%, respectively of the total labour force. Unemployment rates are higher among workers in the first quintile of the income distribution, except in Honduras, where the first quintile has an unemployment rate of 4% compared with 5% for the fifth quintile. In Argentina, Brazil, Colombia, the Dominican Republic, Uruguay and Venezuela, the unemployment rate among the lowest income groups varied between 14% and 19%, while it exceeded 20% in Chile, Costa Rica and Jamaica (Alaimo et al., 2015).

Despite the gains in formality in the past decade, the bulk of employment in Latin America remains informal. In 2013, on average, approximately 55% of the workers in the region were informal. Although, this represented a 7 percentage point decrease from the 2003 level (Alaimo et al., 2015), it still meant that more than half of the working population did not contribute to social security programmes, a common feature in many emerging economies (Jütting and de Laiglesia, 2009).

Informality rates also vary significantly across countries and by income within countries, affecting the poor and low-middle income workers the most. Figure 2.3 describes formality rates across the region. Compared with the whole labour force, three Central American countries (Honduras, Guatemala and Nicaragua) have the lowest formality rates in 2013 (implying informality rates of over 80%). At the other end of the distribution, Chile, Costa Rica and Uruguay have formality rates of more than 70% and informality rates below 30%.

There is a strong variation in informality rates by income within countries. They are highest in all countries for workers in the poorest quintile, and tend to fall as income rises. In Guatemala and Honduras, the informality rate for the first quintile is close to 100% falling to below 60% in the fifth quintile. On the other hand, Uruguay, Costa Rica and Chile present the lowest informality rates among the first quintile workers (56%, 52% and 41%, respectively). The results show that informality is high among middle-income workers (middle-class proxy, according to Easterly, 2001) at 64% over quintiles 2-4, and still prevalent for the highest income workers in the fifth quintile (42%) (Figure 2.4).

Informality tends to more prevalent among the self-employed than employees in Latin American countries. Figure 2.5 shows a breakdown of individuals by category of labour relationship in 2013. On average, 60.8% of the workforce are employees, 27.9% self-employed, 4.86% employers and 6.48% unpaid family workers. There is a substantial rate of informality among the self-employed (88.3 %). Informality is also important among paid employees, accounting for 46.5 % of this group.

There are relatively high turnover rates of workers in jobs in the region and many jobs are only temporary. At least one in four workers in the major part of their working life (i.e. aged between 24 and 54) has been in their current job for one year or less. And many of the transitions between jobs or from unemployment to a job imply becoming an informal worker (Figure 2.6). So even for the 45% of workers with formal jobs at a point in time, many of them have experienced a period of informality in the past and will do so in the future.

The region is trapped in a ‘vicious cycle’ of high employment turnover, low productivity, high labour costs and high informality. High employment turnover discourages workers from seeking more on-job education and training, which could otherwise result in employment relationships that generate increases in productivity. In turn, low labour productivity relative to the costs of formality, including taxes and SSCs, contributes to a high level of informality. Alongside the high levels of job turnover and high informality, most workers lack safety nets (e.g. unemployment insurance) to withstand a period of unemployment while looking for a good job that is suited to their skills and training, leading to inefficient job matches. These inefficient matches in turn tend to break up quickly, continuing high turnover and completing the vicious cycle (Alaimo et al.,2015).

4. Social security contributions and informality: In it together?

The high levels of informality in the region have a very diverse set of roots. There are several factors that can influence an individual’s decision between formality and informality, such as: tax burdens, wage levels, job security, excessive labour regulations, the valuation a person places on the programme or services, expectations of receiving future benefits, and a component of myopic behaviour from the individual, not to mention the role of institutions in performing inspections and setting up enforcement mechanisms. However, there are cases in which the decision of the individual is purely based on the price of social security programmes and taxes. In such scenarios, the cost of becoming part of the formal sector is judged to be too large when compared with the level of earned income, thus effectively preventing any possibility of participating in these social schemes (Bosch, Melguizo and Pagés, 2013; and Melguizo, 2015 for a focus on middle class).

This section identifies the informality rates among dependent workers over the whole of the income distribution and makes corresponding estimates of the theoretical costs of becoming formal. Figure 2.7 shows the rates of informality (measured on the left axis) by income decile for 18 countries drawing on data from the household income surveys in the region.2 Harmonised data from these sources is used to estimate the annual labour earnings distributions of all dependent workers3 and to calculate informality rates by deciles of income. The analyses identify the approximate location of the minimum wage in each country with a white bar. The figure also shows estimates of the theoretical costs of becoming formal (measured on the right axis) expressed as the proportion of their wages that workers should pay in social security contributions to become or remain formal. This measure can be taken as a lower bound, given that formalisation generally entails other monetary and non-monetary costs stemming from various legislations. For workers earning the minimum wage or above, this cost is defined as the amount of employee SSCs payable on the wages. However, many workers are excluded from social security programmes because their earnings are below the minimum wage which commonly acts as a lower income threshold for these schemes. For these workers, the cost of becoming formal is the amount of SSCs payable at the minimum wage or at the lower earnings threshold if that is different. As a result, the larger the shortfall between a worker’s income and the established minimum threshold/wage, the higher is the theoretical cost of formalisation that this individual faces. These costs are expressed as a percentage of the workers actual wages in the charts.

The theoretical cost of becoming formal, even when only considering social security contributions, is relatively high in some countries of the region, especially in the lower and middle income deciles although there are significant variations among countries. On average, most countries show relatively high formalisation costs, as a percentage of their labour income, in the first two income deciles. In Mexico, the theoretical formalisation costs for the first decile exceeds the workers’ income, implying that the earnings are less than the minimum payments of social security contributions, while in Panama and Venezuela, it is around 20% of the workers income. As incomes rise, the costs of formalisation steadily fall till they become flat or proportional. This is always the case at earnings deciles above the minimum wage except where there are declines due to the existence of upper earnings ceilings to the SSCs.

High costs of adhering to social security programs as a percentage of earned income are associated with higher levels of informality. Figure 2.7 shows the relationship between theoretical formalisation costs and informality rates and how these costs and rates vary over the income distribution. In most countries the informal workers at the lower end of the income distribution (deciles 1-2) have higher formalisation costs that the remainder. Similarly, in most countries, informality tends to be largest in the first two deciles of the income distribution, although in some economies informality is also substantial in higher income groups. In Argentina, Brazil, Chile, Costa Rica, Dominican Republic, Panama, Uruguay, Venezuela, informality impacts mainly those on the lowest incomes (deciles 1-2). In Ecuador and Colombia, it additionally impacts on a so called ‘vulnerable middle class’ (deciles 1-5) (as proxied by Easterly, 2001). Finally, relatively high informality is present even at middle income levels in Bolivia, El Salvador, Guatemala, Honduras, Mexico, Nicaragua and Peru.

For the workers at the lower end of the income distribution (deciles 1-2), there are relatively high levels of both the costs of formalisation and informality rates in the majority of countries, implying that higher costs are associated with higher levels of informality (Figure 2.8, Panel A).

As levels of income rise, (deciles 3-5 in Figure 2.8, Panel B), two groups of countries can be identified, one with relatively high levels of informality (above 60%) and the other with lower levels (around 40% or less). In both groups of countries, the positive relationship between formalisation costs and rates is still evident. But in each group, both the level of dispersion and the formalisation costs are lower than in the corresponding analysis for deciles 1 and 2.

In income deciles 6-8 (Figure 2.8, Panel C), the distinction between the two groups of countries fades. However, in contrast with the two analyses covering the lower half of the income distribution, they show a negative relationship between formalisation costs and informality rates. This means the informality rates are falling while the costs of formalisation are rising. This implies that at these levels of income, formalisation costs may not be the main determinant in worker decisions with respect to formality. There are two other notable features of this group. First, countries with the larger proportion of informally employed workers also have lower formalisation costs relative to the other group of countries, and secondly the range of the levels of informality covered by the two groups is smaller than in the income deciles 3-5.

At the upper level of the income distribution (deciles 9-10 in Figure 2.8, Panel D), informality rates also tend to fall as the cost of formalisation rises. The correlation between formalisation costs and informality rates remains negative as in Figure 2.8 Panel C. As in the latter case, this points to the existence of other factors, apart from the cost of formalisation, explaining levels of informality.

5. Conclusions

The analysis shows that taxation, specifically social security contributions, has impacts on labour market outcomes. High rates of informality are correlated with high theoretical formalisation costs, as proxied by the SSCs workers would have to pay to remain or become formal, especially for those at the lower end and in the low-middle range of the income distribution. The interaction of lower earnings thresholds of SSCs (usually minimum wages) and social security program rates increases the costs of adhering to social programmes for those with earnings in the lower and middle income deciles, while the costs of participation in these programmes as a percentage of earnings decreases for those at the higher levels of the income distribution. From the employer’s perspective, high non-labour costs (i.e. holiday pay, severance costs and other labour regulations) might affect also play a detrimental role in the supply of formal jobs. This adds to the many other non-tax factors that explain individual decisions to become or remain informal.

This situation is evidence of a need for new policies on taxation, labour markets and social insurance programmes. Ensuring a proper environment for the creation of quality jobs in the formal sector should be at the centre of the agenda for inclusive growth. Improving the range of opportunities for formal employment so that workers and companies can fully enjoy, not only the benefits of social insurance, but also the capacity for personal growth and added value associated with formal economies is critical to LAC countries. Focusing on taxation and social insurance, an appraisal of the levels of lower earnings thresholds of social security programs is required. Also, progressive social security schedules, and pension schemes where governments or employers match the individual’s contributions would increase the returns from becoming formal would be good avenues to achieve these aims. Promoting productive job stability to lay the foundation for longer lasting and productive employment relationships is key to ensuring growth in long-term and consolidating a stronger and more vibrant middle class.

References

Alaimo, V., M. Bosch, D. Kaplan, C. Pages and L. Ripani (2015), Jobs for Growth, Inter-American Development Bank, Washington, D.C.

Bosch, M., A. Melguizo and C. Pages (2013), Better pensions, better jobs. Towards universal coverage in Latin America and the Caribbean, Inter-American Development Bank, Washington DC.

De Soto, H. (1989), The other path: the invisible revolution in the third world, HarperCollins, New York.

Daude, C., A. Melguizo and A. Neut (2010), “Fiscal Policy in Latin America: Countercyclical and Sustainable at Last?”, OECD Development Centre Working Paper No. 291, OECD, Paris.

Easterly, W. (2001), “The Middle Class Consensus and Economic Development”, Journal of Economic Growth 6(4), 317-335.

ECLAC (2014), Tax policy in Latin America. Assessment and guidelines for a second generation of reforms, ECLAC, Santiago de Chile.

ECLAC/ILO (2016), Employment Situation in Latin America and the Caribbean. ECLAC-ILO Bulletin No.14, Santiago de Chile.

Ferreira, F. H. G., J. Messina, J. Rigolini, L. F. López-Calva, M. A. Lugo, and R. Vakis (2013), Economic Mobility and the Rise of the Latin American Middle Class, World Bank, Washington, D.C.

IDB (2015), Base de datos Sociómetro-BID: Sistema de Indicadores Sociales de América Latina y el Caribe, Inter-American Development Bank, Washington, D.C., http://www.iadb.org/en/research-and-data//sociometro-bid,6981.html.

IDB (2013), More than revenues, Taxation as a development tool, Inter-American Development Bank, Washington, DC.

Jütting, J. and J. de Laiglesia (2009), Is Informal Normal ?: Towards More and Better Jobs in Developing Countries, OECD Publishing, Paris, https://doi.org/10.1787/9789264059245-en

Loayza, N., A.M. Oviedo and L. Serven (2005), “The Impact of Regulation on Growth and Informality – Cross-Country Evidence”, World Bank Policy Research Working Paper 3263, World Bank, Washington, DC.

Melguizo, A. (2015), “Pensions, informality, and the emerging middle class”. IZA World of Labor, pp. 1-10, Institute for the Study of Labor.

Melguizo, A. and J. González-Páramo (2012), “Who bears labour taxes and social contributions? A meta-analysis approach”, Journal of the Spanish Economic Association, 4(3), 247-271.

OECD/CAF/ECLAC (forthcoming), Latin American Economic Outlook, 2017: Youth, Entrepreneurship and Skills for Development, OECD publishing, Paris.

OECD/CAF/ECLAC (2015), Latin American Economic Outlook 2016: Towards a New Partnership with China, OECD Publishing, Paris, https://doi.org/10.1787/9789264246218-en.

OECD/ECLAC/CIAT/IDB (2016), Revenue Statistics in Latin America and the Caribbean 1990-2014, OECD Publishing, Paris, https://doi.org/10.1787/rev_lat-2016-en-fr.

OECD (2015), Revenue Statistics 2015, OECD Publishing, Paris, https://doi.org/10.1787/rev_stats-2015-en-fr.

OECD/ECLAC/CIAT/IDB (2015), Revenue Statistics in Latin America and the Caribbean 1990-2013, OECD Publishing, Paris, https://doi.org/10.1787/rev_lat-2015-en-fr.

OECD/IDB/WB (2014), Pensions at a glance in Latin American and the Caribbean 2014, OECD Publishing, Paris, https://doi.org/10.1787/pension_glance-2014-en.

World Bank (2015), “Working to End Poverty in Latin America and the Caribbean: Workers, Jobs, and Wages”, LAC Poverty and Labor Brief, World Bank, Washington, D.C.

Notes

← 1. The authors of the special feature are Veronica Alaimo, Juan Carlos Benítez, Angel Melguizo and René Orozco. The authors thank Maurice Nettley, Bert Brys and Dominique Paturot for their contributions to this special feature and Melany Gualavisi and Maria Laura Oliveri for providing the data that enabled the analysis. The opinions expressed and arguments employed herein are the views of the authors and do not necessarily represent the official views of the Inter-American Development Bank (IDB), the Inter-American Centre of Tax Administrations (CIAT), the Organisation for Economic Co-operation and Development (OECD) and the OECD Development Centre or the governments of its member countries.

← 2. Details on the household income surveys are presented in the methodological appendix.

← 3. Self-employed and independent workers are not included in this simulation exercise as many special regimes and provisions may apply to these cases.