Latin America’s missing middle of midsize firms and middle-class spending power

Latin America used to be the world’s most prosperous emerging region, but it is on the verge of being overtaken by other regions that were long considerably poorer. It has lost ground since the 1980s despite reform initiatives because of sluggish growth and the unequal distribution of the gains of that growth.

Latin America’s missing middle: Rebooting inclusive growth (PDF—2MB) focuses mainly on Brazil, Colombia, and Mexico, and examines two “missing middles” holding the region back: a robust cohort of midsize companies and a solid middle class with growing spending power. The two are connected: expanding the pool of modern, competitive firms can create better paying jobs, and lift middle-class demand needed for new investment. The disruption presented by adoption of digital technologies provides an opportunity to reignite this virtuous cycle of growth—but it will require reforms to help companies compete and workers prepare.

Growth in Latin America has been slower and more volatile than in other emerging economies, and the unequally distributed gains from that growth have failed to ignite the domestic economy and sustain growth.

GDP growth in the region averaged 2.8 percent annually between 2000 and 2016, far lower than the 4.8 percent rate in 56 other emerging economies in the same period (not including China, which grew at twice that pace). Among Latin American economies, Peru grew most strongly, averaging 5.2 percent per year. Of the three countries that are the focus of this report, Colombia performed best, with average growth of 4.2 percent. Brazil’s growth rate, at 2.4 percent, was closer to the regional average, while Mexico’s was weaker, at 2.1 percent. In addition to being weak, regional growth has been volatile, buoyed by the commodity boom, which saw prices for resources such as oil, mineral, and some agricultural products surge, only to fall back sharply.

The largest driver of GDP growth in Latin America has been expansion of the labor force, the result of a demographic boom and increased participation of women. The workforce grew by 66 million workers between 2000 and 2016, and this expansion accounted for 72 percent of the region’s overall GDP growth. The demographics are now changing: fertility rates in Latin America declined from 4.2 in 1980 to replacement rate in 2015. This means the region will need to count on productivity growth rather than labor expansion as the main driver of GDP growth. This would bring Latin America into line with other emerging economies, where labor expansion has been a much smaller factor in driving GDP growth—37 percent on average in 56 countries excluding China—and considerably less important than productivity growth.

A sharp reduction in poverty is a notable success, but more than 150 million people in Latin America remain vulnerable

Since 2000, 56 million people—or more than 40 percent of Latin America’s poor in 2000—have crossed the $5 per day absolute poverty threshold, reducing the overall proportion of people in poverty to 13 percent of the total from 27 percent. Both the income boom from rising commodity prices and governments’ pro-poor policies propelled this movement. These include conditional cash transfer programs in Mexico and Brazil, which tie transfers to keeping children in school or making preventive healthcare visits. Bolsa Família, the Brazilian program launched in 2003, now reaches 14 million households, about one-quarter of the population, and helped bring down the proportion of those living in extreme poverty from 10 percent to 4.3 percent over a decade. The poverty-reduction momentum has slowed in the wake of the 2008 global financial crisis and the end of the commodity boom, however. By 2018, poverty and extreme poverty rates had risen above the levels of a decade earlier.

While poverty has been reduced since 2000, the slow expansion of higher-productivity and higher-wage jobs has left many vulnerable. No longer officially poor, they have not yet reached comfortable middle-class spending patterns—meaning they are still unable to afford goods and services that go beyond basic needs and may not have access to credit—and remain at high risk of falling back into poverty during a recession or in the event of a health or employment crisis. More than one-third of the region’s population lives on less than $11 per day, based on purchasing power parity, including 152 million people in the “vulnerable” category of $5 to $11 per day (over 60 percent from Brazil, Colombia, and Mexico). More than 60 percent of the population lives on less than $20 per day (Exhibit 1).

1

The bottom 90 percent in Latin American households consumes just 64 percent of the total, the lowest share in the world

Consumers crossing over to the middle class have more money and more access to credit to spend beyond basic necessities and they tend to spend a higher share of their income than more affluent people. However, social inequality marks consumption patterns in Latin America. The bottom 90 percent of the income distribution in Latin America—the region’s poor and vulnerable, and most of the middle class—accounts for just 64 percent of domestic consumption. This is the lowest share in the world, akin to that in Sub-Saharan Africa but substantially lagging behind other regions, where the bottom 90 percent consumes about 70 percent or more of the total (Exhibit 2).

2

The business landscape in Latin America is polarized. The region has some powerful companies, including some with very high productivity that have successfully expanded from their strong local base to become global companies or “multilatinas”—regional powerhouses operating across Latin America. They include AB InBev, America Movil, Arcor, Bimbo, CEMEX, Embraer, FEMSA, Techint Group, among others. By comparison with large firms in other regions, such companies are fewer in number and less diversified beyond energy, materials, and utilities. At the same time, Latin America has a long tail of small, often informal companies that collectively provide large-scale employment, but whose low productivity and stagnant growth hold back the economy.

Missing is a cohort of vibrant midsize companies that could bring dynamism and competitive pressure to expand the number of productive and well-paying jobs in Latin America, much as these firms do in many high-performing emerging regions. Exhibit 3 shows how much less prevalent large and midsize companies (over $50 million in revenue) are for the size of the Latin American economy compared with peers elsewhere. Argentina, Brazil, Chile, and Mexico have only about half as many of these large firms relative to GDP as a range of ten emerging economies we use as benchmarks.

3

Latin America’s large companies can face less dynamic domestic competition than peers in other regions

The missing middle of companies reduces competitive pressure on large incumbents. In comparing Latin American companies with large firms in a benchmark set of ten countries—China, India, Indonesia, Malaysia, Philippines, Poland, Russia, South Africa, Thailand, and Turkey—we find several indications that the competitive dynamics are less powerful, but with important variation across countries. (We chose these benchmarks based on geographic variation and similar levels of development, but we do not claim they are globally representative.)

First, the number of public companies with revenue rising above $100 million since 2000 is smaller proportionately in Latin America than in high-performing Asian countries. Whereas 87 percent of publicly listed companies of this size in our benchmark countries have crossed the $100 million threshold since 2000, the proportion is less than 60 percent in Brazil and Mexico. Colombia’s much smaller corporate sector has been more vibrant, with 81 percent of firms crossing that threshold since 2000.

Second is the degree of turnover at the top of the rankings. Top quintile of companies with more than $500 million in revenue in the best-performing emerging economies have a hard time getting to the top and staying there: less than half of top-ranked firms in economic profit—that is, returns above the cost of capital—in one cycle were still at the top in the next cycle. In Latin America, we see a distinction between Mexico, where two-thirds of the top quintile firms in terms of economic profit remained in place over the past 15 years, whereas in Brazil, only one-third did, indicating a more vibrant market. Other economies in the region tend toward the average of our benchmarks.

Third, these dynamics have correlated with concentrated gains in some cases. Some recent research shows that Latin American firms have been able to command larger and more stable markups on their products, dating to the 1980s. That was even before the rise of the “superstar” phenomenon among large global firms that are capturing an ever-larger share of profits and pulling away from their peers. We also find that midsize and large firms’ profit margins in our three countries are higher across most sectors since 2000 than in our benchmarks. The variation in economic profit is also higher: the top quintile economic profit is higher and the lowest quintile economic profit is lower (more negative) than in the benchmark economies. Trade protection, a legacy of earlier import substitution policies, also continues to contribute to high consumer prices in a number of industries.

The causes of this firm distribution and dynamics are rooted in common legacies of import substitution that favored a few private licenses or large state-run firms in many sectors. Other reasons are differing ways in which state companies were privatized and, especially in Brazil’s case, tax and compliance-heavy regulation that favors either large scale or informality. Unequal access to finance, weak infrastructure, and high input costs also squeeze the middle. The result is a weak level of innovation and specialization needed for future growth. Brazil, Colombia, and Mexico rank below their development level on innovation indices; they have one-quarter to one-third of the digital diffusion of the United States, for example. Research and development investment in the region is relatively low, at just 0.3 percent and 0.5 percent of GDP for Colombia and Mexico, respectively, although it reaches 1.2 percent for Brazil. This compares with about 1 percent for benchmarks.

Much of Latin America’s labor force is trapped in a long tail of small, unproductive, and often informal firms

The other side of this dual economy is a multitude of small companies, many of which operate outside the formal economy (as do some midsize and even large firms). These small companies tend to have low productivity and absorb a significant proportion of low-skill workers. They are most concentrated in sectors like retail, construction, and agriculture. Brazil has had the most success in reducing the level of informality in the overall labor force: the rate fell below 50 percent in 2015–16. Mexico and Colombia have much higher rates, of 57 percent and 62 percent, respectively, according to household surveys in each country.

Retail highlights how the long tail of small companies drags down overall productivity. In Mexico, for example, small retailers employ two-thirds of the sector’s workforce, or 6.6 million workers, but value added per worker is only about one-seventh that of large and more efficient retailers.

To increase our understanding of the long tail of small businesses and its resistance to change, we conducted a survey of 3,000 workers—informal, formal, and self-employed—in Brazil, Colombia, and Mexico. It suggests that a majority of workers see self-employment as the main fallback option when wage employment is not available: 70 percent of respondents said temporary or informal self-employment was a viable option in a time of financial uncertainty. Such uncertainty is high, with 28 percent of all workers saying they expected to leave their job in the next 12 months.

The survey also highlights the constraints of small business. Some 28 percent of the sample were self-employed, of which about half were formal businesses. Only one in four of respondents indicated that they had access to the credit needed to expand their businesses—but two in three said they would like to expand if they could.

The other missing middle is a cohort of upwardly mobile consumers whose rising disposable income helps drive economic demand and investment on a sustainable basis. As noted, many of those lifted out of poverty remain vulnerable and have not reached a comfortable middle-class existence. For this cohort, the main path to sustained and rising prosperity comes from access to productive, well-paid jobs—which have been too scarce to meet the needs of the rapidly expanding labor force in the region. As the supply-side challenge of a missing middle of companies has constrained the expansion of attractive jobs with higher wages and opportunities to build skills, it has contributed to the demand-side challenge of a missing cohort of middle-class consumers with sufficient income to sustain robust domestic demand. The sluggish demand acts as a serious drag on the economy. It limits the growth of markets for domestic businesses that constitute the majority of jobs in a modern economy. It also holds back the development and provision of more complex goods and services, which in turn could strengthen the economy by encouraging investment.

The growth patterns providing the context for this missing middle differ among our focus countries. In Brazil, the commodity boom led to rising wages and helped draw workers from small-scale agriculture to low-skill jobs in small establishments with below-average productivity. In Mexico, by contrast, wage growth has been strikingly weak and at times negative, even in expanding, high-productivity sectors such as automotive.

Consumption has grown with expanding population, exports are not filling the demand gap, and both investment and savings are relatively low

Household consumption represents the biggest share of GDP expenditure in any economy and 64 percent in Latin America, yet it has expanded only slowly, contributing 1.9 percent to GDP growth compared to 2.7 percent in our emerging-market benchmarks. Furthermore, compared to benchmarks, a higher proportion of the consumption growth that has occurred in Latin America has come from expanding population rather than rising per capita consumption. Since the turn of the millennium, per capita consumption growth in Latin America has been just 1.8 percent, less than half the 4.3 percent in benchmark countries. Colombia is high for the region, with per capita consumption growth since 2000 of 2.8 percent, while Chile and Peru are outliers at 3.7 and 4.2 percent respectively. In Mexico, the reverse is true: per capita consumption growth accounted for less than the population growth’s contribution, consistent with low wage growth. This matters because consumption patterns change as incomes rise, with markets for more complex goods and services growing to provide new growth opportunities for business.

The purchasing power of Latin American consumers is further taxed by the high cost of many consumer goods. Trade barriers continue to raise local prices above global ones, including for basic goods such as food. Similarly, taxes can limit the purchasing power of those most ready to spend. In Brazil, value-added tax represents 40 percent of the total tax burden, compared with 32 percent of the tax burden in Organisation for Economic Co-operation and Development (OECD) countries on average. Taxes on cars approach 50 percent of the cost to consumers. Counting all taxes, the lowest decile ends up paying 32 percent of income in taxes, compared with 23 percent for the middle decile and 21 percent for the top decile.

The difficulty in obtaining consumer credit and underdeveloped digital services in finance and retail compound the plight of the vulnerable. Financial inclusion is limited, with less than 50 percent of people in Mexico and Colombia having access to bank accounts. This restricts the ability of households to make the most of their incomes by borrowing and saving. Only 6 percent of sales in Brazil in 2017 took place online, for example, about the same proportion as in India and Indonesia, but substantially lower than in China (16 percent) and South Korea (13 percent). However, digital disruption is starting to appear. In Colombia, notably, a startup called Rappi is providing on-demand delivery of items such as groceries and medication and has become one of Latin America’s first “unicorns”—a startup with a valuation of at least $1 billion. The Argentine digital marketplace MercadoLibre is also gaining ground and spreading across the region.

Exports could help fill this demand gap, providing the incentives and markets that companies in Latin America need to invest and broaden their production. However, with the notable exception of Mexico, exports play a smaller role than in the other emerging economies we use as benchmarks; exports in the region account for just 22 percent of GDP compared with 36 percent in the benchmark countries. Mexico has achieved a much stronger export performance since the establishment of NAFTA. Its exports of goods and services rose from 25 percent of GDP in 2000 to 38 percent in 2017. However, the benefits of its export growth have not spilled over into the domestic economy more broadly.

Expectations of future demand are a main reason for companies to invest in larger and better production capabilities. With limited access to export markets and slowly growing domestic demand, Latin American economies have maintained rates of investment below those of their peer economies. Latin American and Caribbean investment in 2016 amounted to 19 percent of GDP, below the global average of 24 percent and the investment rate among East Asian emerging countries of 41 percent. Of our three focus countries, only Colombia reached the global average, recording a rapid rise, from 13 to 24 percent between 1999 and 2015, before dipping.

Income is not flowing through to expanded domestic consumption for different reasons: A comparison of Brazil and Mexico

The flow of income to expanding domestic consumption can slow for different reasons. Brazil and Mexico provide two examples. In Brazil, social policies, minimum wage increases, and a credit expansion during the commodity-cycle upswing all gave a boost to the domestic market—but it was not sustained. Employed labor income accounted for 57 percent of gross income growth, due mostly to expansion of low-skill retail and other services. The demand boom was not matched by supply-side reforms to raise productivity, however. Moreover, high consumer prices and expensive credit continued to tax the purchasing power even of those with rising incomes. The benefits faded away through inflation and macroeconomic instability once the commodity boom ended. The demand boom was thus unsustainable because it was not supported by supply-side initiatives or a macro environment conducive to private investment.

Mexico, by contrast, focused on supply reforms and access to external markets, boosting productivity among the large modern segments of many industries by as much as 5.6 percent in the 2000s. Yet domestic demand has lagged. Productivity in the long tail of small industries did not improve and, unlike in Brazil, the gains did not feed through into broad-based wage growth; rather, they accrued mainly to the profit share of income. For example, while manufacturing productivity increased by an average of 1.7 percent annually between 2005 and 2015, average wages were stagnant. Given that a higher share of wage income is spent (rather than saved), this limited the multiplier potential of the productivity gains to translate into a more sustained source of domestic demand expansion. Domestic market growth was thus limited, crimping incentives for investment. The middle lost out amid these changes.

The Mexican auto sector provides one striking example of the divergence between productivity growth and wages. Production has grown at an annual average rate of 7 percent since 2006, and after doubling in a decade, labor productivity now ranks with that of the top producers in the world. Labor income gains have come from added workers at a wage level that is 1.6 times the national average, yet the average wage of Mexican autoworkers declined in the same period. By comparison, South Korea at the same point in its auto industry development saw a 58 percent increase in sector wages over the corresponding ten years of productivity growth.

In Colombia, the patterns of growth since the turn of the century have been more inclusive. Wage growth averaged more than 2 percent annually, exceeding productivity growth and reflecting productivity gains across sectors. Some six million workers joined the workforce between 2000 and 2015, with disproportionate gains for women. However, this rapid expansion slowed wage growth compared with some peers. For example, Thailand grew at about the same rate during this period but nonetheless achieved wage growth of 3.7 percent, almost double that of Colombia, thanks to substantial productivity gains.

If Latin American economies were able to fill in and strengthen the missing middles, thereby establishing a virtuous cycle of inclusive growth, the potential reward would be substantial. Using a macroeconomic simulation that assumes conservative growth in productivity is matched by increased labor shares and consumption in line with more inclusive countries, we find that achieving such a growth cycle could lift GDP in the region in 2030 by 50 percent above a baseline scenario that factors in current trends, including reduced labor force expansion. That would amount to an increase per person of more than $1,000 per year, or a $1 trillion incremental boost to GDP in 2030. Obtaining that prize will be challenging. Governments and business leaders will need to embrace a growth-focused policy agenda that addresses the missing links of past reforms.

Even in a time of social and political tensions, Latin America has a new opportunity to revitalize inclusive growth—and new tools to achieve it: digital technologies that can raise productivity growth and develop the missing middles, if embraced and adopted at scale. Digital entrepreneurship is already on the rise, with new platforms and applications from e-commerce to digital finance springing up across the region; several digital startups have achieved valuations of $1 billion valuation or more.

Digital technologies can help directly address some of Latin America’s challenges around the missing middles. While not a “silver bullet” or cure-all, the technologies can make it easier for companies to open businesses, register property, and file taxes over the internet, reducing the cost of red tape. Digital can facilitate more efficient markets from land and jobs to local services. Digital platforms make it possible for small and mid size companies to become “micromultinationals” able to compete with much larger competitors by offering their goods and services through online marketplaces regionally or globally. Critically, digital can and is already creating new, more productive jobs in Latin America—and the potential productivity boost it could give to the region’s economies as a whole could offset the drag from its changing demographics. At the same time, these technologies present potential risks for inclusive growth in terms of job displacement and further concentration of profits.

Three priorities will be needed to lay the foundations for a pro-growth, inclusive agenda for the region and make it better able to embrace digital disruption.

First is the need to create a competitive business environment in which digital can thrive, innovation is rewarded, and opportunities are created for all, especially squeezed firms in the middle. That will mean using digital tools where available to cut red tape, improve access to finance, and reduce barriers to entry and growth that reduce dynamism and performance pressure. Second, regulation will need to be redesigned to create labor and consumer markets that spread productivity gains to the vulnerable and middle classes. If Latin America is to boost prosperity for all, productivity gains must flow to rising incomes for workers as they increase in skill level and to firms well placed to reinvest in their people, technology, and communities. Third, government itself can leverage digital platforms to raise public-sector efficiency and improve the delivery and cost of public services. Institutions will need to adapt and shift focus from protecting and regulating the status quo to encouraging new investment and experimenting with new delivery solutions.

Such changes will make Latin America better able to harness digital technologies, which could serve as a powerful source of productivity and growth. Many countries in Latin America have already started on this path; if that energy can be maintained and built out, the region may be ripe for a Digital Spring. While the changes may be led by digital companies and technologies, they will extend to non-digital parts of the economy, too. Disruption from digital and in the broader economy offers an opportunity for a new model of inclusive growth in which governments facilitate new sources of productivity for firms and workers and design policies to share benefits with all citizens.