Chapter 3. Then and now: Differences in development trajectories

The process of a shifting economic geography has sped up economic convergence for many developing countries. However, strong economic growth in the South has not solved all problems in countries undergoing rapid economic transformation, and development paths have looked different from one country to the next. That is because development is an inherently more complex and multidimensional concept than gross domestic product (GDP) can summarise by itself. This chapter explores development patterns beyond GDP alone in a long-term historical perspective. It discusses the meaning of development in light of current discussions on “Beyond GDP”, provides evidence on GDP and well-being outcomes since 1820 in a broad range of developing and emerging economies, and compares the experience of early industrialising countries versus more recently emerging economies.

This chapter was prepared jointly by the OECD Development Centre, the OECD Statistics and Data Directorate and researchers from the Clio-Infra team at the University of Utrecht. In particular Rijpma, van Zanden and Mira d'Ercole (2018[1]) provides the basis for the sections on the historical and regional analyses of well-being presented in this chapter.

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.

The world economy is characterised by a radical process of a transforming economic geography due to the strong economic growth experienced by a range of emerging and developing countries. Economic growth in the South has not solved all problems, however. Development is inherently more complex and multidimensional than gross domestic product (GDP) growth alone can summarise. In spite of economic growth, some old problems have persisted, and new ones have emerged. This chapter analyses development outcomes beyond GDP per capita in a long-term historical perspective. In doing so, it explores whether the development paths of more recently emerging economies delivered different results in terms of growth and well-being to those of countries that industrialised earlier.1

It analyses a broad range of outcomes, such as poverty, inequality, health, education, environmental quality and personal security. It also compares the experiences of these countries since the 1950s with that of countries in the “old world” that experienced economic take-off in the 19th and early 20th centuries.

How did the relationship between growth in GDP and other measures of economic, social, political and environmental development evolve over time? Did economic growth and industrialisation in the 19th century have the same impact on people’s well-being as it did in the more recently emerging economies?

Findings suggest that countries in different eras have distinct experiences of growth. Catch-up growth in the People’s Republic of China (hereafter “China”) and India in the late 20th century, for example, had a different impact on well-being than it did during the early industrialisation of countries such as Sweden and Germany in the 19th century. With respect to annual GDP per capita growth,2 the former two countries experienced rates of 5% to 10%, while the latter two had rates of at most 2%. Higher GDP growth provides the means for well-being to grow faster as well. However, the degree to which GDP growth is translated into better well-being outcomes varies substantially. Sometimes it does not translate at all.

This chapter is based on a broad set of well-being measures developed by economic historians and included in the OECD’s How Was Life? report (van Zanden et al., 2014[2]).3

It begins by briefly revisiting the meaning of “development”. It also reflects on initiatives to measure performance “beyond GDP” launched in the aftermath of the Stiglitz-Sen-Fitoussi report in 2009.
It then presents evidence on how the relation between levels of real GDP per capita and well-being measures has changed globally since 1820.
The chapter then examines trends since the 1950s in a range of dimensions of people’s lives for 23 emerging economies,4 based on the Clio-Infra database. It identifies similarities and differences amongst countries across different periods.
The next section identifies key patterns in the experiences of nine countries in the developed world that industrialised earlier (1820-1950).5 It also compares the experience of “early” and “late” industrialisers, showing how gains in well-being lagged behind GDP growth in the early industrialisers over this earlier period.
The final section summarises key findings from the analysis, highlighting the need to rethink development paradigms in light of evolving relationships between economic growth and well-being outcomes.

The three main messages of this chapter are that:

Development is more than growth in per capita GDP; a broad array of indicators is needed to measure development.
GDP per capita and well-being outcomes are not always linked.
The quality of economic growth in recently industrialising countries has not matched that of the early industrialisers: well-being gains could have been even greater given the rapid pace of growth.

“Development” of what?

In 1969, Dudley Seers argued that the nature of the main challenges confronting the developing world in the post-war period had been fundamentally misconceived:

This (challenge) has been seen as achieving an increase in the national incomes of the “developing countries”, formalised in the target of 9% growth rates set for the first development decade. Of course, we have all been aware that development consists of much else besides economic growth. [...] Yet little more than lip service is paid to it [...] [T]he experience of the past decade makes this belief look rather naïve [...] Now that the complexity of development problems is becoming increasingly obvious, this continued addiction to the use of a single aggregative yardstick in the face of the evidence takes on a rather different appearance, it begins to look like a preference for avoiding the real problems of development. (Seers, 1969[3])

Fifty years after these remarks, Seers’ challenge has not yet been met with an adequate response. However, recent developments make it possible to address the challenge more systematically than possible before. In 2009, the Commission on the Measurement of Economic Performance and Social Progress released a seminal report. The commission, convened by former French President Nicolas Sarkozy, stressed the limits of GDP as a metric of welfare. It called for a move from measuring economic production as the sole metric towards consideration of outcomes for people. This approach should stress the importance of combining GDP with broader metrics of household economic well-being, quality of life and inequality, as well as the sustainability of these outcomes over time (Stiglitz, Sen and Fitoussi, 2009[4]). Since then, the OECD has played a central role in moving this agenda forward by regularly monitoring a range of well-being indicators for its member countries.

The notion of well-being is close to that of human development promoted by Sen (1999[5]), amongst others, which underpins the work of many United Nations (UN) agencies. It focuses on outcomes and opportunities that are intrinsically important to people in themselves (an end) rather than only as an instrument to achieve something else (a means); on the diversity of these outcomes; and on their irreducibility to a single aspect (e.g. no amount of income can offset the lack of basic freedom).

Sen’s concept of “capabilities” stresses the importance of understanding development as a process that enlarges one’s choices. However, the OECD’s How’s Life? report recognises that measurements based on outcomes is often the best that can be achieved. Several key principles inform this work. First, it is concerned with people rather than with aggregate economic conditions. Second, it focuses on well-being outcomes – aspects of life that are directly and intrinsically important to people – rather than the inputs and outputs that might be used to deliver those outcomes.6 It does this for two reasons. Inputs may be poorly correlated with the resources devoted to achieve well-being outcomes; and a different combination of inputs and outputs may be equally effective in delivering the same result. Third, it emphasises the importance of inequality in each well-being outcome. Fourth, it considers both objective and subjective aspects of life, as people’s evaluations and feelings matter as much as the objective conditions in which they live. Lastly, it considers the sustainability of such outcomes. This approach does not imply ignoring the importance of GDP and economic growth. Rather, it recognises that these are means to an end rather than ends in themselves.

These principles have informed the framework shown in Figure 3.1 for OECD member countries (OECD, 2017[6]). Current well-being is described through 11 dimensions belonging to the broader domains of quality of life and material conditions. The assessment of future well-being is based on changes in a range of resources. Benefits extend to tomorrow, but are affected by today’s actions; these resources are grouped under the categories of economic, natural, human and social capital. The framework is operationalised through a set of headline indicators pertaining to average well-being outcomes and inequalities, as well as resources to ensure sustainability.

How well does this approach describe the development experience of poorer countries? Analysis of the literature suggests that none of the dimensions in Figure 3.1 can be deemed irrelevant in less developed countries. While there are differences in the relative importance of various aspects of life depending on national circumstances, most dimensions are also common across countries. They differ more in the way they are labelled than in terms of what they regard as most salient. The framework in Figure 3.1 would need to be adapted to better fit the realities and concerns of poorer countries.7 However, differences across countries are more likely to appear in terms of the importance attributed to the different dimensions by people living in the country themselves (Boarini, Kolev and McGregor, 2014[7]). This conclusion is also in line with the Voices of the Poor studies by the World Bank in the late 1990s (Narayan et al., 1999[8]). These studies highlighted the importance of complex needs (i.e. the needs of voice and recognition and of avoiding shame and isolation) as opposed to the simple needs of food and shelter, even amongst the poorest people, in the poorest countries – a finding that runs contrary to the view of a rigid hierarchy of needs shaped by different stages of economic development of countries.

The OECD How Was Life? describes long-term development patterns in a broader range of countries and in longer perspective (starting in the 1820s) (van Zanden et al., 2014[2]); (Box 3.1), based on a similar methodology to the How’s Life report.8 Naturally, this type of historical analysis must contend with a range of practical problems. Historical data are simply not available for some of the dimensions included in Figure 3.1.9 In other cases, available data may refer to concepts that only crudely approximate the variable of interest. Data limitations also make the conceptual distinction between well-being today and well-being tomorrow less applicable to historical analysis. However, decades of historical research have also generated a wealth of measures for various aspects of people’s lives. These can be systematically gathered and, to some extent, compared across countries and time. This undertaking aims to be approximately right rather than exactly wrong, which is what happens when one summarises the development experience of countries through changes in their GDP per capita.

The How Was Life? report presented evidence on the multidimensionality of development in a long-term historical perspective. A sub-set of these variables (Table 3.1) is used in this chapter to shed light on the relationship between GDP per capita and various well-being variables, and to compare the development experiences of countries around the world and in different time periods. This analysis shows that, while there are strong correlations between GDP per capita and most dimensions of people’s life across countries and over time, the correlation is not always linear, with different patterns of leads and lags and shifts in the relationship between GDP and well-being variables.

Box 3.1. The Clio-Infra project and How Was Life? report

Clio-Infra is an international inter-disciplinary effort by a team of economic historians to systematically chart the various dimensions of development between 1500-2010. Clio-Infra builds on a pioneering effort to compile a set of comparable indicators of economic development for the world economy stretching back 1 000 years (Maddison, 2001[9]).

The How Was Life? report (van Zanden et al., 2014[2]) was the culmination of Clio-Infra. The report included data for six (population weighted) world regions. These comprised Western Europe; East Europe and former Soviet Union; Western offshoots (Australia, Canada and the United States); Latin America and Caribbean; sub-Saharan Africa; and the Middle East and North Africa. Analyses are based on all countries in the Clio-Infra database with sufficient data, and a separate series for 25 of the largest countries in the world. These countries are Argentina, Australia, Brazil, Canada, China, Egypt, France, Germany, Indonesia, India, Italy, Japan, Kenya, Mexico, Nigeria, the Netherlands, Poland, the Russian Federation (hereafter “Russia”), Spain, South Africa, Sweden, Thailand, Turkey, the United Kingdom and the United States. Data generally refer to national states based on existing borders. This implies that (when possible) the dataset took the most recent borders as reference and corrected earlier data for changes in borders whenever they occurred. In cases where the same approach could not be adopted in the past, data refer to countries based on their historical borders.

The data in How Was Life? and used for this chapter are state-of-the-art estimates by economic historians for various countries. They are harmonised to the extent possible by project participants. These estimates pertain to GDP and GDP per capita; real wages of unskilled labourers, educational attainment, life expectancy, population height, casualties from homicides, political institutions, emissions of carbon dioxide and nitrates, and biodiversity loss. Data on income inequality and gender inequalities, as well as a composite index of well-being, were also included in the report.

Data in How Was Life? were presented as decadal averages. As data coverage increases for more recent periods, imputations were used for missing countries in earlier periods. For all series, data quality (for individual countries and decades) was assessed based on three criteria. The first is credibility (the degree to which the sources of the data can be confidently relied upon). The second is accuracy (the extent to which the data are deemed to be valid and to reliably represent what they purport to measure). And the third is comparability (the extent to which data from different sources measure the same concept and are collected based on the same methodology).

Based on these criteria, four types of data were distinguished:

high quality, produced by an official statistical agency (national or international) or by researchers using techniques that ensure equivalent credibility
moderate quality, produced using historical sources and methods comparable with (but not necessarily identical to) those applied by official statistical agencies
low quality, resulting from historical research in a data-scarce environment and making use of indirect data and estimates
estimates based on guesses, conjectures and interpolation between benchmark years, where there may be significant inconsistencies between countries or gaps in coverage.

Table 3.1. Well-being variables from the Clio-Infra database
Well-being outcome	Variable name	Max	Min
Health status	Life expectancy at birth	83.1 years (+)	14.5 years (-)
Political institutions	Composite measure of political regimes (Polity2)	10 (+) fully democratic	-10 (-) fully authoritarian
Education	Average years of completed education	13.6 years (+)	0.01 years (-)
Human height	Average height of different birth cohorts	183 cm (+)	152 cm (-)
Income inequality	Gini coefficient	0.74 (+)	0.21 (-)
Earnings	Number of consumption baskets purchased with the real wages of a male unskilled worker in building industry	355 subsistence baskets (+)	0.5 subsistence baskets (-)
Personal security	Homicide rate	82 homicides per 100 000 inhabitants (-)	0 homicide per 100 000 inhabitants (+)
Environmental quality	Sulphur dioxide emissions per capita	425 (-) kg SO₂ per capita	0 (+)
Global well-being	Composite indicator of well-being	3.7 (+)	-1.6 (-)
Note: (+) indicates that higher values of the variable (e.g. education) are increase well-being, while (-) indicates that higher values of the variable (e.g. income inequality) lower well-being. Subsistence baskets are a measure of goods based on a standard amount of caloric and protein intake (van Zanden et al., 2014[2]). The composite indicator of well-being is linear measure consisting of nine variables: GDP per capita, real wages, height, life expectancy, average years of education, income inequality, governance, species abundance and homicide rate.
Source: Clio-Infra (2017[10]), Clio-Infra (database), www.clio-infra.eu (accessed in July 2018).

Development is a complex and multidimensional concept

The story of economic growth since 1990 is relatively positive, as shown in Chapter 2. The transformation of economic geography has raised the prospects of growth for many developing countries, placing them on a converging path with the world’s more developed economies.

However, a more holistic view of development, one that considers the material conditions (e.g. income) as well as quality of life (e.g. health, education) that contributes to well-being, tells a more complex story: in spite of economic growth, in certain countries the number of people living in extreme poverty is rising, the gap between rich and poor is growing, and environmental outcomes are deteriorating. The combination of a transforming economic geography, economic convergence and the dynamic movement of such well-being outcomes has blurred a previously clearer line between a “developed” and a “developing” country.

Since the inception of the Millennium Development Goals (MDGs), there has been tremendous progress on poverty reduction. In fact, the MDG target to reduce extreme poverty by half by 2015 was met five years ahead of schedule (United Nations, 2015). China reduced extreme poverty amongst its population from 67% to 2% from 1990 to 2013, or from 755 million people to 25 million. The number of people living below the extreme poverty threshold outside of China was also reduced by 337 million from 1990 to 2013, despite rapid population growth (World Bank, 2018[11]). The Sustainable Development Goals (SDGs) build on the momentum of the MDGs.

But these great economic leaps over the two previous decades, and the continued growth by some of the world’s poorest countries, is not enough to end extreme poverty. In Africa, for example, although the share of the population living in extreme poverty dropped from 56% in 1990 to 43% in 2012, the absolute number of people living in extreme poverty has grown substantially over this time period due to the region’s rapid population growth (Beegle et al., 2016[12]).

The World Poverty Clock (WPC) provides real-time estimates and monitoring against the first SDG of ending extreme poverty. According to the WPC, an estimated 641 million people were still living below the extreme poverty line of USD 1.90 per day in the world in July 2018. More than one-third of the extreme poor lived in three countries: Democratic Republic of the Congo, India and Nigeria. Despite GDP growth above the world’s average of 3% from 2010 to 2017 in several developing countries, the projected number of poor people will still be higher in 15 countries by 2030, the target year of the SDGs. Moreover, more than half of the population in 12 countries lives in extreme poverty. However, prospects are improving in countries such as Guinea-Bissau, Lesotho, Malawi and Mozambique, where the extreme poverty rate is decreasing (Figure 3.2).

Poverty is not the only thing that matters; how benefits of growth are distributed within countries is equally important (Islam (2006[13]); Khan (2007[14])). Income inequality has been increasing within countries such as China and India despite the GDP growth and convergence experienced by developing countries over the past two decades (Alvaredo et al., 2017[15]). Not all countries are equally affected, however. Unlike the recent experience in developed economies, the rich get richer in developing countries, but the poor get richer as well. Recent increases in inequality in more developed economies are largely because the rich gain, whereas the poor do not (Lang and Mendes Tavares, 2018[16]).

Developments in other well-being outcomes matter just as much. Eroding living conditions in many developing countries in the 1950s and 1960s, amid a wave of enthusiasm in development and a push for industrialisation, sowed the seeds for thinking about development beyond income and its distribution. Dudley Seers, as well as Robert McNamara and Amartya Sen, would bring greater policy importance to poverty reduction and improvements in non-economic outcomes to the development agenda in the 1970s. Inspired by the work of Sen, the United Nations launched the Human Development Report in 1990. More recently, the MDGs in 2001 and their successor, the SDGs in 2015, further cemented the importance of looking beyond GDP for development. They brought it to the mainstream.

The relationship between well-being and GDP per capita is complex. Individual satisfaction with standards of living, for instance, does appear to increase with countries’ GDP per capita. However, the relationship is not linear. Furthermore, the variance in the relationship is not uniform at different levels of GDP per capita. According to a Gallup survey, the share of people dissatisfied with living standards in their country varied widely by country in the first third of the GDP per capita ranking, and less so at higher levels of GDP per capita (Figure 3.3).

Countries share challenges across income thresholds. Indeed, when looking across a series of development outcomes, income groups are not sufficient to characterise the level of the development challenges faced by individual countries.

Income groups do provide a good indication of the prevalence of extreme poverty, however. According to recent analysis by the OECD Development Centre (OECD, 2017[17]), only the Republic of Congo, among all middle-income countries, had a level of extreme poverty that would correspond to that of low-income countries. This is in line with findings in the literature that economic growth plays a major role in the reduction of extreme income poverty (Dollar and Kraay, 2002[18]).

In contrast, income groups provide a poor indication of the level of inequality. Countries’ gross national income (GNI) per capita and the Gini coefficient, a standard measure of income inequality, are not closely correlated. Not surprisingly, 13% of high-income countries have levels of inequality that could well be found in low-income economies. Moreover, almost half of all middle-income countries have high levels of inequality (with a Gini coefficient above 0.4). This is consistent with findings in the literature that several countries transitioning to middle-income status in the past decades have experienced growth with significant inequality increases (Sumner, 2016[19]).

There is consensus on the need for relatively poorer countries to grow faster and that this economic growth is fundamental for their development (Milanovic, 2016[20]). GDP measures domestic production and remains a useful indicator to track this aspect of development. However, when discussing well-being outcomes for individuals in a society, GDP and GDP per capita are less useful concepts. For example, GDP per capita should not be conflated with income, as calculations of GDP include the income that accrues to non-residents, for example multinational companies that may repatriate profits. In this way, GDP per capita does not reveal the average income of individuals, one of the dimensions of well-being under the broader domain of “material conditions”.

A broader concept of development requires a different approach to measurement. Moving beyond GDP metrics as the sole indicator, measuring development requires a range of indicators of well-being outcomes and data on how well-being outcomes are distributed across a population.

Historical GDP per capita and well-being

What is the link between people’s well-being and GDP per capita, and how has it changed since 1820? Globally, well-being indicators have been closely correlated with GDP per capita.10 Countries with higher per capita GDP have experienced higher levels of education, real wages, average height and life expectancy, as well as lower homicides and more democratic institutions.11 Some indicators, such as income inequality and homicides, have had a much weaker relation with GDP per capita, correlations that became negative only in the mid-19th century and in the early 20th century, respectively (i.e. eventually, countries with higher GDP per capita tended to have lower income inequality and homicide rates).

The relation between various aspects of well-being and GDP per capita has changed over time. Two periods can be identified (Figure 3.4).

In the first period, from the middle decades of the 19th century until around 1870, countries with higher GDP per capita did not always have better well-being outcomes. On average, they experienced lower life expectancy and higher homicides, as well as institutions that were no more democratic than in other countries. This suggests that, in this phase, economic growth and industrialisation did not necessarily contribute to the well-being of the population. During the first 50 years of economic growth amongst early industrialisers, gains in well-being were relatively small and sometimes even negative.

In the second stage, which began in about 1870, the correlation between GDP per capita and well-being measures became stronger. This convergence reflected several developments.

First, the import of cheaper American foodstuffs to Europe resulted in a dramatic decline in food prices, which helped raise real wages and consumption levels (O'Rourke, 1997[22]).

Second, while the early stages of industrialisation took place in non-democratic regimes, by the end of the 19th century many industrialising countries had become democratic.

Third, breakthroughs in medical knowledge – such as the germ theory of diseases developed by Pasteur – created the right conditions for much more effective health care. Often, this combined with the growing attention paid to public health issues by governments. As a result, life expectancy started to rise in Europe and its overseas offshoots after 1870, driven by declines in child mortality.

Fourth, the first policy measures to address social concerns, likely driven by the extension of voting rights to working classes, were introduced in Europe. These included bans on child labour and legislation concerning maximum working hours. As a result, a link emerged on a global scale around 1870 between GDP per capita and life expectancy, human height or democratic institutions (Figure 3.4). This was a shift from the mid-19^th century where there was no such correlation. Similar developments can be identified for other well-being indicators.

The 19th century, therefore, first saw a divergence, with well-being lagging behind progress in GDP per capita. This was followed by a certain convergence between per capita GDP and various measures of well-being. This is confirmed through a cross-sectional correlation between GDP per capita and measures of well-being at different periods.

Figure 3.4 shows how these correlation coefficients changed over time at the global level. They were often low, sometimes even negative, in the first half of the 19th century. They increased and became positive from the late 19th century onwards. The panel includes the full set of countries. However, because economic growth was limited to early industrialisers, the experience of these countries drives most of the correlation results at the global level observed in the 19th century.

Cross-country correlations between GDP per capita and various well-being measures are only one part of the story. For many indicators in the 19th century, there was no additional well-being accrued beyond those explained by increases in per capita GDP. This changed, however, in the 20th century, when some indicators began delinking from GDP per capita. Figure 3.5 charts changes in well-being unexplained by GDP per capita to investigate the relationship between per capita GDP and well-being.12 A value of zero implies that levels in well-being outcomes are entirely explained by levels in GDP per capita.

This also signalled a new era in the relationship between GDP per capita and well-being. Well-being now often increased more rapidly than implied by GDP growth alone. This delinking is most pronounced for life expectancy. By the end of the 20th century, life expectancy had increased by 15 years (one standard deviation) more than would be expected from per capita GDP alone. Similar patterns apply to height and education. In the case of real wages, per capita GDP levels explain most of the difference between countries; decoupling does occur in the last two decades. An unexplained effect of higher GDP on democracy (Polity 2) emerged in the second half of the 19th century. Apart from trends in the other direction during the Second World War and in the 1960s and 1970s, this effect has stayed largely in place.13 In the case of sulphur dioxide (SO₂) emissions (hazardous to human health as well as flora and fauna), evidence of a drop relative to what could be expected based on per capita GDP alone is not observed until 2000.

Personal security is the one well-being dimension that deviates from the pattern of delinking from GDP per capita. In the second half of the 20th century, homicide rates were higher than could be expected from countries’ per capita GDP. However, homicide data do not go far back in time for many developing countries, relative to the other variables. Lastly, the unexplained effects for income inequality are somewhat declining. This means that countries become more equal than expected from changes in per capita GDP levels. However, the pattern is erratic, likely due to data quality issues.

What explains this delinking between GDP growth and well-being during the 20^th century? The answer lies in autonomous changes in the regimes, policies and technology that produce well-being outcomes. Delinking is particularly clear in the case of health status. The relationship between life expectancy and GDP per capita shows a constant upward shift starting in 1870, with the realisation of the first great breakthroughs in medical sciences. Shaped by technology and policies, such as public sanitation, the health system constantly improved health outcomes without necessarily requiring increases in GDP per capita – a phenomenon known as the shifting Preston curve (Preston (1975[23]); Bloom and Canning (2007[24])).

A similar evolution also occurred in terms of the average years of education in countries. Since the 1960s, increasingly higher levels of educational attainment were realised without increases in GDP per capita. This result may reflect the impact of government policies, but also structural changes in the economy and individual preferences. This delinking points to a virtuous effect. On the one hand, parents may prefer to invest more in the education of their children, or adults may want to invest more in upgrading their own skills. On the other hand, increasing skills helps achieve better GDP per capita.

Overall, however, across-the-board shifts in the relation between well-being and GDP per capita only partially explain improvements in various measures of well-being. As described above, this effect is important for life expectancy, and for educational attainment since the 1960s. The absence of an autonomous shift in environmental variables (such as sulphur emissions) from their relation with GDP per capita implies that environmental degradation increased with economic production. Technological change seemed to only have a small effect in delinking, and improving, environmental degradation in light of economic growth.

Well-being outcomes in recently emerging and developing countries

How has well-being evolved in more recently emerging and developing countries, along the eight measures discussed above?

This section looks more specifically at development since the 1950s for four key regions, focusing on countries that are either large in terms of population or characterised by contrasting developments within the region. As in Figure 3.5, the analysis looks at both actual developments in different well-being variables and at levels predicted by GDP per capita.14 Unless stated otherwise, regional averages are always taken from How Was Life? (van Zanden et al., 2014[2]). They are population-weighted averages based on all countries in the region for which there is data and imputations for the countries where the data are missing.15

Since the 1950s, late developers, which began to industrialise and grow rapidly only in recent decades, have been distinguished from the early developers by the phenomenon of “catching up”, or GDP per capita convergence. In the 19th century, the differences in GDP per capita between the most advanced countries and the rest of the world were relatively small. Indeed, the rate of economic growth of the fastest growing countries was no higher than 2%. This changed dramatically during the 20th century. The gap between the more productive countries and the rest of the world widened, creating a large potential for catching up. The former Soviet Union during the first stages of central planning, Japan after 1950 and Southeast Asian countries that industrialised more recently achieved rates of economic growth ranging from 5-10% annually. These rates were much higher than the early industrialisers in the 19th century experienced. This catching up also had an impact on well-being outcomes in these countries, which could also increase much faster. Not all countries in the global South were equally successful in this respect, however.

Latin America has seen mixed progress in well-being compared to GDP

Latin America has historically been a fascinating laboratory of experiments, with alternative policy measures adopted by both left- and right-wing governments. The long-term trend of improving well-being outcomes in education and health is quite robust for the region, while inequality has remained high and personal security has decreased sharply. In some dimensions, well-being gains since the 1950s, for example, are stronger than for GDP per capita. Based on Clio-Infra data for Latin America, the region’s performance in terms of well-being was poor prior to 1950. This poor performance occurred even though average GDP per capita in the region was above the global average. In 1930, for example, the global average was USD 1 673, while the average for Latin America was USD 1 795 (van Zanden et al., 2014[2]).

Before the 1950s, however, Latin America was doing worse in terms of all well-being measures than in terms of GDP per capita compared to the rest of the world. The region was particularly characterised by high income inequality, with the highest Gini coefficient worldwide in 1929. Moreover, democracy scores were the lowest in the world. Key metrics were also all below world averages. These included average and educational attainment (2.0 vs. 2.5), the share of population having attained at least basic education (36% vs. 41%) and average life expectancy (37.8 vs. 40) (van Zanden et al., 2014[2]).

This pattern was partially reversed in the second half of the 20th century, despite much slower GDP growth than the global average. In 1950, GDP per capita in Latin America was about 20% higher than that of the rest of the world. This margin remained more or less unchanged until the 1980s. This changed in 2000 and 2010, when GDP per capita was 90% of the global average in Latin America. Strong GDP growth in East Asia, the heart of the transformation of economic geography, combined with the “lost decade” of poor growth in Latin America in the 1980s were the main drivers of this reversal. In 1980, Latin American GDP per capita was more than twice the level of East Asia. It now lags by about 30%.

Since the 1950s, the development of the well-being dimensions of education, health and political stability in Latin America has followed a different pattern to that observed for GDP per capita. First, average years of education grew more rapidly than the global average. In 1980, this dimension was equal to the global average, but by 2010 it was 5% above that level. Life expectancy at birth already exceeded the global average in the 1950s, and since the 1980s the difference widened to 3.5 years (van Zanden et al., 2014[2]). The continent’s democracy scores have risen dramatically since the 1970s. They are now amongst the highest in the global South, far outperforming Africa and Asia. However, personal security (homicide rates), real wages and inequality are the region’s lagging well-being outcomes compared to the rest of the world.

Annex Figures 3.A.1-6 (available on line only) provide well-being outcomes for the six previously discussed measures and their predictions. These are based on what could be expected given their per capita GDP between 1950 and 2010. The figure confirms that economic growth in these six countries (Argentina, Brazil, Chile, Mexico, Peru and Venezuela) has been generally unstable. Declines in predicted well-being are frequent, reflecting episodes of lower GDP per capita. Conversely, the evolution of well-being has been much more stable. Furthermore, its growth curve was hardly affected by large swings in GDP per capita, which illustrates the delinking between GDP and well-being discussed in the previous section. Figure 3.6 below presents a sample of three well-being outcomes for Chile and Peru.

Well-being in terms of life expectancy and education years has been steadily increasing. Democracy has also strongly improved. In 1950, the quality of democratic institutions (the “polity2” variable) was generally lower than would be expected based on GDP per capita. By 2005, it is much ahead. However, notable temporary setbacks can be observed in all six countries. Brazil, Mexico and Venezuela experienced dramatic increases in homicides, whereas one would expect a slow decline linked to higher GDP per capita. Another setback was experienced in income inequality, which is generally higher than would be expected, although less so in Argentina and Venezuela. However, data on income inequality end in 2000, thereby missing much of the more recent decline.

Chile is the most successful country in the region in terms of GDP growth, more than doubling its GDP per capita since the 1990s to 144%. Only Peru comes close to matching this performance (132%). Brazil and Argentina also grew rapidly in the 2000s, but both experienced events in the 2010s that undid some of the earlier gains. Chile’s GDP performance was much more stable. This may be related to the tradition of coalition governments and the absence of the more extreme political tensions experienced by Brazil and Argentina.

Chile’s record in improving several well-being outcomes is even stronger than that recorded by GDP. Average years of education and life expectancy have both increased by more than what would have been expected based on per capita GDP alone. Only for real wages and income inequality does Chile do worse than would be expected based on GDP. Well-being outcomes have performed considerably better in Chile since the 1970s, including education and life expectancy, and for homicide rates throughout the period. For both Chile and Peru, improvements in several well-being dimensions, namely life expectancy and education, preceded stronger GDP.

Venezuela is an outlier in many ways. As an oil-producing country, its per capita GDP levels have been high, but this has not translated into better well-being measures such as higher education and life expectancy or lower homicide rates. Only income inequality has been relatively lower and real wages high relative to its per capita GDP and to other Latin American countries. In recent decades, well-being measures in Venezuela deteriorated strongly. Homicide rates are now very high, both real wages and democracy have declined, and income inequality has risen strongly.

After spiking during the 20^th century, income inequality decreased in the past two decades, although Latin America remains the most unequal region. Since about 2000, policy makers in several countries in the region gave priority to reducing income inequality and poverty, and increasing the well-being of the (poor) population in general. This can be seen as an attempt to correct the high level of income inequality, which has been and remains a dominant feature of Latin America throughout the decades. Amidst the global boom in international commodity prices in the 2000s, including for oil, there is evidence that Latin American governments were successful in reducing income inequality and poverty– particularly in Bolivia, Ecuador, Argentina and, to a lesser extent, Brazil and Chile (Lustig, Pessino and Scott, 2013[25]).

Poverty reduction in the first decades of the 21^st century in Latin America has been remarkable. Today only 3.7% of the population lives under the international extreme poverty threshold of USD 1.90 per day (2011 PPP), compared to 11.5% in 1999. Three countries more than halved extreme poverty during the period. The rate in Brazil dropped from 12.7% to 5.5% over 2003-11. In Bolivia and Ecuador, it dropped from 18% to 7% and from 10% to less than 5%, respectively, between 2003-15. But other countries – Peru, Colombia and Paraguay – did equally well. The average poverty rate of Latin America fell from 12% to 5% over 2002-13 (World Bank, 2018[11]). The only exception to this pattern is Venezuela, which experienced a dramatic increase in extreme poverty up to 2005 (the latest year with data available).

There are two reasons for the decline in income inequality and poverty. First, government spending increased on social programmes. Second, the premium of skilled wages fell due to the expansion of education and the compression of the wage distribution (Lustig, Pessino and Scott, 2013[25]).

Therefore, over the past 50 years Latin America has been much more successful in increasing the well-being of its population in several dimensions than in generating GDP growth. It is perhaps the best (regional) case of the de-linking between GDP and certain well-being outcomes.

Sub-Saharan Africa shows great diversity in development trajectories, as well as in the relation between GDP and well-being outcomes

Until recently, the GDP growth record of sub-Saharan Africa was poor; in 1950, the average GDP per capita of the African countries for which data are available was about 40% of the global average, a level that fell to 20% by 2010. The 1960s and 1970s showed positive GDP growth, and the gap with the global average increased only marginally. However, between 1970 and 2000 no economic growth was achieved (GDP per capita was USD 1 282 in 1970 and USD 1 099 in 2000). Only since 2000, the era of shifting economic geography, did economic growth become positive again (with GDP per capita increasing to USD 1 481 in 2010).16 Overall, the continent achieved average real annual GDP growth of 5.4% between 2000 and 2010. Economic growth, however, slowed down recently, reflecting a sharp drop in commodity prices (AUC/OECD (2018[26]); Leke and Barton (2016[27])).

Improvements in well-being since the 1950s were somewhat better than for GDP per capita. However, they were also characterised by a constant and sometimes growing gap with the rest of the world. Average years of education in sub-Saharan Africa increased strongly – from 0.8 to 4.2 between 1950 and 2010. Still, all other regions did (much) better, and the absolute gap with the global average increased. Similarly, life expectancy in the region also increased from 38 to 52 over the 1950s-2000s. However, it was still lower than elsewhere in the world, and the gap with the global average remained constant at around 25%. Income inequality was and remained relatively high. Only the democracy index shows consistent progress in the region. Democratic rights were poorly protected in the 1970s and 1980s, but have improved considerably since then. Overall, the Clio-Infra composite well-being index suggests a considerable improvement for the region from about 1950 – after a long period of little change between 1850 and 1950 (van Zanden et al., 2014[2]). That said, variation within the region is substantial, both between countries and when comparing actual well-being and the level predicted by per capita GDP.

Developments in eight well-being indicators amongst six countries in the region – Burkina Faso, Ghana, Kenya, Nigeria, Uganda and South Africa – as well as changes predicted based on GDP per capita are depicted in Annex Figures 3.A.7-12 (available on line only). Figure 3.7 below presents a small sample from the full annex figures.

South Africa has always been one of the most prosperous countries of sub-Saharan Africa. Its average levels of GDP per capita are three to four times higher than the sub-Saharan African average. Economic growth has, however, been modest in recent years; since the start of the millennium, GDP per capita has increased by only 1% annually. Income inequality declined somewhat, although South Africa remains one of the most unequal countries in the world. It has a Gini coefficient far higher than would be expected based on per capita GDP alone (Annex Figure 3.A.8, available on line only). Personal safety in South Africa has also improved, though homicide rates are still more than five times higher than would be expected based on its economic record. Since the end of apartheid, South Africa has exceeded its expected performance in education and democracy. However, these positive developments were overshadowed by the decline in life expectancy; the spread of HIV/AIDS reduced life expectancy from 62 to 52 over 1990-2005. A recent study concluded that in 2015 life expectancy for women in South Africa was the lowest in the world: 48.7, compared to 50.7 for men– a notable gender difference given that typically women live longer than men elsewhere in the world (He, Goodkind and Kowal, 2016[28]).

South Africa’s record in terms of reducing extreme poverty is also rather poor, although the share of the population living below the poverty line fell from 29% in 1993, to 26% in 2006 and to 19% in 2014 (Sulla and Zikhali, 2018[29]). However, the Moatsos (2017[30]) estimates, which refer to the costs of a “bare bone” consumption basket and are available on an annual basis, show an even smaller decline: from 46% in 1994, to 50% in 2004 and to 36% in 2014. Social transfers went up since the end of apartheid, but structural problems with the labour market continue to plague the South African economy.

In most of the six sub-Saharan African countries included in Annex Figures 3.A.7-12 (available on line only), the trend in well-being has been upward, usually due to improvements in years of education and life expectancy. In this respect, developments in sub-Saharan Africa within the global context of strong increases in educational attainment and health status appear to be independent of economic growth. The development of democratic institutions is much less consistent. In fact, it is characterised by huge swings due to alternations between dictatorship and more democratic phases. However, the trend is upward in most cases and certainly since the 1990s (Annex Figure 3.A.11). Real wages, when data are available, have largely been flat and differences between the six countries are small. For Ghana, Kenya and Nigeria, real wages are what would be expected given per capita GDP in these countries. Real wages are higher than expected in Burkina Faso and Uganda, but much lower than expected in South Africa, a reflection of the country’s high inequality.

The HIV/AIDS crisis has strongly affected the well-being of large parts of the continent beyond just South Africa. Civil wars (in Uganda, for example, in the 1970s) and political instability in general may have contributed to stagnation in life expectancy observed since the 1970s. Instead of profiting from the autonomous shifts suggested by the Preston curve, countries such as Kenya, Uganda and Nigeria showed a strong stagnation or even decline in life expectancy, though it was followed by a recovery in the 2000s.

Looking at the record of the six sub-Saharan African countries in reducing extreme poverty, a mixed picture also emerges. Some countries were quite successful. Botswana, Burkina Faso and Uganda combined moderate to fast GDP growth with a strong reduction of extreme poverty. Angola, on the other hand, was dynamic in terms of GDP growth, thanks mainly to growing revenues from oil production (and the end of a civil war). Yet extreme poverty remained high in Angola – more than 80% of the population was living below the USD 1.90 per day poverty line. Nigeria, an even more important oil exporter, has seen its GDP per capita double since 2000. However, its level of extreme poverty remained almost unchanged (at about 70%). Kenya and Tanzania, which are often compared to each other, experienced a convergence in their poverty rates. Kenya – the more successful market-oriented economy – did not register much progress with poverty reduction. Conversely, Tanzania’s level of poverty fell to the much lower level prevailing in Kenya.

The overall picture in sub-Saharan Africa is one of many different development trajectories. These depend on economic starting points, rates of GDP growth and types of growth achieved, particularly the importance of strategic exports like oil. Extreme poverty is declining substantially in only a few countries. Poverty rates are, in most African countries, still very high. They do not show the same systematic decline that was observed in Latin America.

Asia has high but declining well-being benefits from economic growth

Asia, especially the area stretching from the Eastern Mediterranean Sea to South China, was the core of the world economy until it was overtaken by Western Europe in the early modern period (late 15th and early 16th century) (Maddison, 2001[9]). At the start of the 19th century, GDP per capita in Western Europe was two to three times the level of China, India or Indonesia. This divergence rapidly increased in the 19th century. When Europe industrialised, large parts of Asia lagged behind and saw their share in manufacturing output decline rapidly due to European competition. In the second half of the 19th century, only Japan successfully emulated the European model of labour-intensive and export-oriented industrialisation. After 1950, other Asian economies – Chinese Taipei, Korea and Singapore – developed similar outward-oriented development strategies to profit from opportunities in international markets. This strategy then spread to other parts of the region: Thailand, Malaysia, Indonesia and, more recently, Viet Nam. Since the 1980s, India and China developed their own versions of such open door policies, with great success, and account for much of the process of a transforming economic geography and the rise of the BRICS (Brazil, Russia, India, China and South Africa).

In 1900, GDP per capita in East and Southeast Asia was about USD 600 on average. This was only 20% of the level of the most advanced countries in Western Europe, and 50% of the global average. In 1950, levels of real GDP per capita of the region increased only marginally to about USD 660, whereas the global average had gone up by 70%. Until the 1970s, the gap between Asia and Europe did not show any signs of closing. The exceptions to this rule were Japan and a few of the other “flying geese” – countries that adopted an economic strategy based on technological leadership, regional hierarchy and international trade – which took off in the slipstream of Japan’s success. The market-oriented reforms of the 1980s and 1990s, and their impact on China and India, triggered GDP convergence. The most dramatic change was the switch to markets and international openness that occurred during the years of Deng Xiaoping’s leadership (1978-1989). For India, the turning point was 1991, when the economic liberalisation of the country really started. In Indonesia, the third largest country in the region, fast GDP growth began about 1970. It was associated with the Suharto regime’s New Order economic policies. The economic success of all these changes is well-known. Since 1990, GDP per capita increased by more than 5 times in China, and by 3.5 times in India – a pace unmatched in history. However, India and Indonesia still display large gaps in GDP per capita with Europe, while China is making much faster progress. China’s GDP per capita is now attaining a level of about half the Western European average.

Gains in some dimensions of well-being, although not all, have been spectacular as well. Annex Figure 3.A.13-18 (available on line only) show various well-being measures for the six largest Asian countries (China, Indonesia, India, the Philippines, Thailand and Viet Nam) (Figure 3.8 below shows a small sample from the complete annex figures). For China there was hardly any increase in well-being before 1940, but this changed after the communist takeover of 1949. In 1958, China introduced the Great Leap Forward, a large economic and social programme to rapidly propel the country towards socialism. After this catastrophic programme ended in 1962, life expectancy began a spectacular growth (from 33.7 years in the 1930s to 65.4 years in the 1970s). Since then, the increase has been slower, up to 73.9 years in 2000. Educational attainment has been the second source of rapid increases in well-being in China. This was partly because the Chinese state invested a lot in education, but primarily because Chinese parents invested heavily in the education of their children. The one-child policy introduced in 1979 was perhaps the most effective instrument to enhance investment in education. Following the introduction of this policy, average years of education began a strong rise, growing faster than per capita GDP in the 1950s and 1960s. Average years of schooling increased from 1.7 years in 1950 to 6.9 years in 2000.

There were no gains in terms of political rights, however, according to the polity2 measure in Annex Figure 3.A.17. Income inequality in China was already low in the 1950s. Given the country’s per capita GDP, it dropped further during the first decades of communism (1950s-1960s). However, from the 1970s onwards income inequality has increased dramatically. The Gini coefficient increased from 0.28 in 1970 to 0.44 in 2000 – roughly the same level observed in other countries with a similar level of GDP per capita.

Improvements in India’s well-being were much more gradual (Figure 3.8 and Annex Figures 3.A.13-18). There was hardly any GDP growth before 1948, but since then, the trend has been upward, with a decisive acceleration in the 1980s. In the colonial period, life expectancy already began rising from 23.7 years in the 1900s to 32.6 years in the 1940s. At independence, life expectancy was still lower than what would be expected given India’s (low) per capita GDP. After independence, a steady rise took place, especially from the 1950s to the 1970s. India today has a higher life expectancy than could be expected based on its per capita GDP.

India, however, has not done as well as other Asian countries in other well-being indicators. Average years of education stood at a very low level of 2 years in 1890 and this further declined to 1.2 years in 1950. The colonial state failed to enhance the general education of India’s growing population, and the demand for human capital remained low. After 1950, average years of education increased steadily, but it is never higher than expected based on per capita GDP. By 2000, gains in average years of education had not kept pace with India’s recent increase of GDP growth. Furthermore, the overall trend in income inequality is upward, and unskilled labourers’ real wages were essentially flat. Both developments contradict what could be expected from India’s per capita GDP growth. Conversely, India’s strong record as a democracy stands out compared to both its Asian peers and to what is expected based on per capita GDP levels. This diversity of well-being developments, some of them countering GDP growth, show again the two are not always correlated.

Generally, however, well-being in Asia has shown tremendous progress, particularly with respect to extreme poverty. China achieved the most spectacular increases in well-being in the last 50 years or so (see Figure 3.8). Extreme poverty declined in equally dramatic fashion, despite a strong increase in income inequality. India’s poor population as a share of total population declined according to World Bank data from 54% in 1983 to 46% in 1993, 38% in 2004 and 21% in 2011.

The continued high level of extreme poverty in India has been the subject of debate (Dréze and Sen, 2013[31]), but seems beyond dispute. Bangladesh and Pakistan, which grew much less strongly in terms of GDP per capita, managed to lower extreme poverty in a much more significant way. Pakistan, for which data seem to be best, lowered its poverty rate from 62% in 1983 to 6% in 2013.17 Indonesia, the other giant in this region, experienced a spectacular decline of extreme poverty based on World Bank estimates, from 70% in 1987 to 7.5% in 2015. However, Moatsos only finds a halving of poverty rates in the country between 1983 (57%) and 2014 (29%). Other countries in the region – Thailand, Malaysia, Philippines and Viet Nam – managed to lower poverty levels into single digits (less than 10% of the population). No data are available for Cambodia and Myanmar.

Sen (2011[32]) stressed the relative success of the Chinese state-led model in terms of well-being.18 He concluded that “those who are fearful that India’s growth performance would suffer if it paid more attention to social objectives such as education and health care should seriously consider that notwithstanding these social activities and achievements, China’s rate of gross national product (GNP) growth is still clearly higher than India’s”. Sen argued that China outperformed India in well-being terms because investments in health care and education were the engine of China’s economic growth. India was also lagging behind other countries in the region in this respect. Indonesia, for example, has now overtaken India in terms of educational attainment. It was already ahead of India in the 1950s in terms of life expectancy.

Evidence from the well-being measures in this section supports Sen’s argument. First, Figure 3.8 shows that China’s progress in life expectancy began in earnest in the 1960s, and even earlier for other measures. Only later did strong GDP growth begin; as a result, Figure 3.8 shows that actual well-being was higher than what would be predicted by per capita GDP alone through most of the period.19 Only in the last decade has this gap diminished, as well-being gains slowed while GDP growth continued. Second, well-being measures in India are consistently lower than those in China, except for democracy, and the gap has been widening since 1950. Furthermore, the last decades have experienced higher GDP growth combined with comparatively sluggish progress in well-being measures. This means that India’s well-being predicted from per capita GDP is now higher than its actual score. Such slowing down of well-being gains compared to per capita GDP growth can also be observed in several Asian countries. This includes both countries where predicted well-being is now higher than actual well-being (India, Indonesia) and those where GDP growth is catching up to earlier progress in well-being (China, Viet Nam, Bangladesh).

Eastern Europe and the former Soviet Union experienced high GDP and well-being performance until the 1980s, followed by collapse and recovery

The final group of countries discussed are those from Eastern Europe or belonging to the former Soviet Union. This section presents evidence for Russia, Bulgaria, Estonia, Hungary, Poland and Romania.

Between 1820 and 1930, GDP growth in Eastern Europe and the former Soviet Union closely followed the global average. GDP per capita doubled between 1820 and 1910, with most of this increase realised after 1870. The First World War and the Russian Revolution of 1917 resulted in a sharp decline in GDP. However, Russia recovered dramatically in the 1930s due to central planning and forced industrialisation (at the expense of agriculture).

As an instrument to modernise the economy and increase GDP per capita, central planning proved successful. In the 1930s, the Soviet Union was the only region that grew rapidly (from USD 575 in 1920 to USD 1 448 in 1930 and USD 2 144 in 1940). During the 1950s and 1960s, the model still seemed to work well in generating GDP growth: USD 3 945 in 1960 and USD 5 575 in 1970 (van Zanden et al., 2014[2]). GDP per capita in the Soviet Union had been below the global average until 1930. From the 1960s onwards, it exceeded the global average by a sizable margin (by 37% in 1950 and by 55% in 1970). GDP growth in Eastern European economies accelerated as the same central planning model was introduced in these countries in the late 1940s. It is, however, less clear if this acceleration can be attributed to the forced industrialisation that central planning induced, or to the generally favourable conditions in the world economy in the post-war economic boom.

The well-being consequences of central planning and forced industrialisation can only be sketched here. The core idea was to transfer surplus production from agriculture and consumption to large-scale investment in capital-intensive industry. A sharp reduction of living standards was therefore inevitable. This is confirmed by the massive famine of 1931-32, which was especially severe in Ukraine. Data on population height also show declines in the 1920s and 1930s. However, there were also countervailing forces (Allen, 2003[33]). Large numbers of farmers migrated to cities, where incomes were much higher than in the countryside. Levels of education rose rapidly (from 2.5 years of education on average in 1930 to 5 years in 1950) and income inequality fell to extremely low levels. Life expectancy rose sharply (after a severe dip in 1931-32), as the quality of public health services improved. Still, political rights were at an extremely low level during the Stalin years.

The assessment of the long-term consequences of these policies is even more complicated. The development of life expectancy in the former Soviet Union is at odds with developments in other countries. After a strong increase between the 1920s (when it was 32.6 years) and the 1960s (when it was 69 years), life expectancy began to stagnate at around that same level (Annex Figure 3.A.22, available online only). Infant mortality may even have increased in the 1970s. These developments led to some debate amongst Western specialists about a possible crisis in the Soviet health care system in the 1970s and 1980s (Kingkade and Arriaga, 1997[34]). Estimates of population height, however, contradict this pessimistic picture. From the 1950s onwards, Russians continued to become taller (from 169 cm in the 1940s to 177 cm in the 1990s). Conversely, after the disintegration of the Soviet Union in the early 1990s, there are clear signs of an acute health crisis. Life expectancy fell to 66 years in the 2000s. This was one of the largest reductions of life expectancy on record that cannot be attributed to war or infectious disease. Life expectancy also declined in other countries belonging to the former Soviet Union (OECD, 2008[35]). Higher mortality rates have been linked to acute psychosocial stress and excessive alcohol consumption during this period of political instability and social upheaval (Cornia, 2016[36]). Homicides also increased sharply after the fall of communism, further contributing to the decline of well-being in these years (Annex Figure 3.A.21, available online only).

Stagnation in life expectancy also occurred in other Eastern European countries during the 1970s and 1980s (see Figure 3.9 and Annex Figure 3.A.22). However, other dimensions of well-being increased rapidly during the years of central planning. Human capital was higher than predicted based on GDP per capita. Income inequality was low – until the 1990s, when it started to increase rapidly – while political rights were extremely low during communist rule. In many dimensions (except for democracy), Eastern Europe remained throughout the 20th century the region with the third highest level of well-being, on average, following Western Europe and Western offshoots (Australia, Canada and the United States) (van Zanden et al., 2014[2]).

The development experience of the “old world”

Well-being lagged behind GDP growth among the early industrialisers

The How Was Life? report analysed the long-term trends of GDP growth and many well-being dimensions in the world economy since the start of industrialisation in the early 19th century. It is therefore possible to compare economic growth and well-being in the early stages of industrialisation for a large sample of industrialised countries. These include the United Kingdom, the United States, Belgium, Sweden, Italy, France, Germany, the Netherlands and Japan. In Western Europe, the break with the pre-industrial period occurred around 1820. In that period, economic growth was either slow, as in the United Kingdom and the Netherlands, or absent, as in the other countries.

During the first 50 years of industrialisation, between the 1820s and 1870s, the rate of GDP growth was still relatively low for these industrialised countries (Western Europe and Western offshoots in the classification of (Maddison, 2001[9])). This is certainly the case compared to the high growth rates today in China. GDP per capita in Western Europe increased by about 1% annually between 1820 and 1910. The Western offshoots did somewhat better, with an average growth rate of 1.5%. This implied that real GDP increased by 160% over this period of 90 years. When compared with the near stagnation prior to 1820, it was a remarkable achievement.

Although relatively slow, GDP growth was underway, but initially had almost no positive impact on well-being. This is especially evident from data on real wages and body height, which closely reflect the income position and consumption patterns of the population. The average height in the United States declined by more than 4 cm between the 1830s (174 cm) and the 1890s (169 cm). Western Europeans, at the start of the century already much shorter than Americans, also shrank from 166 cm to 165 cm between the 1820s and 1850s. The inhabitants of Great Britain in the 1890s were still shorter than those living in the 1820s.

Similarly, real wages in Western Europe in the 1870s were at the same level as in the 1820s – and in the years between they were often below these levels. Health data – life expectancy, infant mortality – tell a similar story. Life expectancy in England was 41 in the 1820s and 41.1 in the 1860s. French and Swedish data show a more positive trend. The real break in the trend occurred in 1870, when life expectancy started to rise. Until 1870, the process of democratisation stagnated in large parts of Europe. The average measure for democracy (polity2 score) in Western Europe is -4.2 in the 1820s and -3.3 in the 1860s before reaching to -0.4 in the 1870s. Given the growth of GDP and the stagnation of the standard of living of the population, income inequality likely also increased rapidly. However, the data are too fragmentary to be confident about this trend.

Educational attainment, which increased in almost all early industrialising countries between the 1820s and 1870s, is perhaps the most important exception to this pattern of stagnant well-being. Great Britain, however, did relatively poorly. Educational attainment was relatively high at the start of industrialisation, and increased slowly (from 1.8 years in 1820 to 3.6 years in 1870). Early industrialisation in England was not based on large-scale demand for skilled labour; large-scale women’s and children’s labour may have crowded out schooling. As a result, the Dutch (5.1 years), Germans (5.4 years), French (4.1 years) and Swedes (4.2 years) had a higher level of educational attainment in 1870 than the British (3.6 years). Continental countries (except possibly for Belgium) followed a development path that was more based on skilled labour than in the birthplace of the industrial revolution.

The early growth paradox describes growth without improvements in well-being

These findings confirm the idea in the literature of an “early growth paradox”. Economic growth did occur during this period, but did not translate into improvements in well-being. This paradox is related to several developments (Komlos, 1998[37]).20 First, it was probably the price that early industrialisers paid for rapid urbanisation and proletarisation. For the working class in England, life in the rapidly growing cities was harsh, the cost of living was much higher than in the countryside and the commodification of labour increased uncertainty of work and income (Engels, 1845[38]). The supply of social services by the state and by the urban communities also lagged behind. The rise of liberal economic ideas also led to cuts in social expenditure, reform of poor laws and probably a reduction in social transfers (Lindert (2004[39]); van Bavel and Rijpma (2015[40])). The period lasting from 1840 to 1870 also saw a general liberalisation in economic matters, including internationally. In 1846, the Corn Laws, a series of tariffs and trade restrictions on imported food and grain to England, were abolished and free trade became the dominant ideology. The very strong growth of international trade, capital markets and migration flows generated the first wave of globalisation. Different social classes profited differently from these changes (O’Rourke and Williamson, 1999[41]), a notable parallel with globalisation since the 1980s.

Levels of education were low in the industrialising districts, and child labour was the norm, strongly competing with education. Health care services also did not keep up with the growth of urban populations. This resulted in poor living conditions, bad sanitation, high health risks and stagnant life expectancy (Szreter (1988[42]); Szreter and Mooney (1998[43])). The wealth of the new class of industrial and commercial entrepreneurs provoked growing tension with the expanding proletariat. This, in turn, gave rise to new ideologies (socialism, anarchism) and new social movements (trade unions, workers and consumer co-operatives, movements for the extension of voting rights). It is no coincidence that the Communist Manifesto was published in the middle of this period (Marx and Engels, 1848[44]). It helped put the social question high on the political agenda. In the long run, it contributed to the rise of social spending and social transfers that would alleviate the most urgent social problems.

The growing divergence between GDP and various dimensions of well-being reversed after 1870. The growing efficiency of transport and trade combined with lower tariffs resulted in the rapid growth of exports of American foodstuffs to Europe. Dubbed an agricultural invasion, it resulted in a dramatic decline in food prices that helped to raise real wages and consumption levels in Europe (O'Rourke, 1997[45]). Breakthroughs in medical knowledge also created conditions for much more effective health care, often in combination with growing attention to health issues by public authorities. Life expectancy started to rise after 1870, and child mortality declined equally dramatically. At the same time, the first policy measures to address social questions were taken and the first experiments with social transfers and social insurance – such as Otto von Bismarck’s legislation of the 1880s – began. Whereas in the mid-19th century there was no correlation between GDP per capita and life expectancy or human height on a global scale, these changes and policy reforms led to the emergence of such a link around 1870 (Figure 3.4 and Figure 3.5).

Annex Figures 3.A.25-32 (available on line only) plot levels of the various indicators of well-being for nine early industrialisers. These are more or less representative of the various patterns in Western Europe and Western offshoots. Figure 3.10 below presents a sample of the complete annex figures. As in previous figures, the data show the actual value of these well-being dimensions, and the expected value based on the level of GDP per capita of the country involved.

Differences within this group are large. The figure shows that economic take-off in the 19th century, especially in Great Britain and Italy, was linked to much lower levels of educational attainment than GDP levels would predict. Sweden and the United States, however, were different, and exhibited a relatively high level of human capital from the start. Similarly, income inequality was extremely high in the first 50 years or so of early industrialisation. Gini coefficients were sometimes 10-15 points higher than what could be expected based on GDP per capita alone.

Most early industrialised countries show a trend towards lower income inequality over 1820-2000. However, there are upswings as well, notably in the second half of the 19^th century and the late 20th century. Homicide rates were low compared to what GDP would predict in Western Europe – but not in Italy, and certainly not in the United States, where they remained high during the entire period. The differences between the values predicted by per capita GDP and actual homicides are large. This suggests a relatively secure environment in Germany, France and Britain given their GDP per capita levels (eight per 100 000 inhabitants, as much as one standard deviation in the entire dataset).

The panel for life expectancy in Figure 3.10 and Annex Figure 3.A.28 shows stagnation during the first half-century, and sharp increases after about 1870. The shortfall of life expectancy compared to values predicted from GDP levels is especially large, as much as 20 years. By the second half of the 20th century, and especially since the 1970s, the situation reversed. Most industrialised countries performed better on life expectancy than predicted by their GDP level. The United States was the exception, where the increase in life expectancy stagnated relative to per capita GDP levels from 1965 to 2010. High income inequality and high homicides rates also contributed to convergence of well-being between both sides of the Atlantic observed since the 1960s, as measured by a composite indicator of well-being outcomes (van Zanden et al., 2014[2]).

Historical trends in well-being

What lessons can be deduced from the well-being experience of early industrialising countries? A comparison between GDP growth and a range of well-being indicators between the early industrialisers in the 19th century and emerging economies in recent decades provides several insights.21

First, economic growth amongst the early industrialisers was much slower than that achieved by many emerging economies in recent years. Per capita GDP of Western European countries grew on average by 1% annually in the 19th century. The Western offshoots increased their GDP per capita somewhat faster. However, the gap with the rate of economic growth of China and India in recent years, and of emerging economies in general, remains large. As one important reason for that difference, emerging economies are much further away from the productivity frontier. Therefore, they can profit from catching up. Conversely, the early industrialisers were at or close to the productivity frontier and could not profit to the same extent from the advantages of backwardness.

Second, such differences in the development process had important consequences for the effect that “early growth” and “catch-up” growth had on well-being. During the first 50 years of early industrialisation, the increase in income inequality meant that the mass of the population did not profit much from higher GDP per capita. The welfare backlash – “dark satanic mills”22, extremely polluted cities, high food costs – of early industrialisation cancelled out the potential effects of higher GDP per capita in raising well-being. High income inequality led to growing socio-political tensions and to the emergence of socialist ideologies and movements. In contemporary emerging economies, income inequality has also increased rapidly. This was often driven by the process of globalisation, although endogenous forces (such as increased scarcities of certain skills) may have played a role as well. But GDP growth in emerging economies is so strong, and autonomous changes in the health system, for example, so effective, that – despite growing income inequality – the well-being of the population has nonetheless increased across the board. In another important difference, urbanisation in the 19th century had negative welfare effects (lower stature, lower life expectancy). Today, despite worse air quality, its overall effects are probably positive due to better nutrition and living conditions in urban centres. In fact, it has taken emerging economies much less time to reach a given level of life expectancy and education than was the case for developed economies (Figure 3.11).

Third, the relationship at the global level between GDP growth and well-being shows signs of change over the years. The correlation of various well-being measures with per capita GDP were low in the first decades of the 19th century. However, they increased considerably during the second half of the 19th century and the first half of the 20^th century. In this period, increases in per capita GDP translated into higher well-being. In the second half of the 20th century, per capita GDP could not account for a large part of the gains in well-being. Relatively high well-being outcomes in countries with low GDP per capita in this period are one of the reasons for this pattern, and they partly reflect autonomous changes in the health system – the shift of the Preston curve. Latin America illustrates these changes most clearly: whereas after 1950 growth in GDP per capita was below the global average, the rise of most dimensions of well-being was clearly faster than the global average. Since about 2000, these positive developments were further enhanced by lower income inequality and poverty, higher education attainment and life expectancy. Latin America’s record in improving well-being is therefore positive, in spite of its sometimes erratic GDP performance. At the same time, the experience of Chile demonstrates there is no necessary trade-off between the two. It grew fastest since the 1980s, and also increased well-being most strongly. The polar case is Venezuela, which recently experienced an economic collapse due to lower oil prices and misguided policies, resulting in a dramatic decline in well-being. Dramatic declines in well-being can be largely independent of economic growth. This could be seen in the effects of the uncontrolled HIV/AIDS epidemic on life expectancy of men and especially women in parts of Africa, which inappropriate policies potentially worsened.

Last, parallels and differences exist between GDP growth and well-being experiences of early industrialisers and today’s emerging economies. In both cases, GDP growth was accompanied by rising income inequality and rapid globalisation. The first wave in 1840-70 and the second wave in 1980-2010 both increased income inequality. In an important difference, the middle decades of the 19th century were not accompanied by higher political rights in the early industrialisers, while a shift towards democracy started after 1870. Conversely, such a change has occurred in emerging economies in Latin America, sub-Saharan Africa and in parts of Asia (Indonesia, Korea, Thailand and, most recently, Myanmar) since the 1980s.

The evidence shows that development, when defined more broadly to encompass well-being and environmental sustainability, does not always follow economic growth. This raises questions on how we think about development and what type of strategy countries should follow to reach better and sustainable levels of economic, social and environmental well-being – questions which are examined in Chapters 4 and 5 of this report.

References

[33] Allen, R. (2003), Farm to Factory: A Reinterpretation of the Soviet Industrial Revolution, Princeton University Press, Princeton.

[15] Alvaredo, F. et al. (2017), World inequality report 2018, World Inequality Lab, https://wir2018.wid.world/files/download/wir2018-full-report-english.pdf.

[26] AUC/OECD (2018), Africa's Development Dynamics 2018: Growth, Jobs and Inequalities, OECD Publishing, Paris/AUC, Addis Ababa, https://doi.org/10.1787/9789264302501-en.

[12] Beegle, K. et al. (2016), Poverty in a Rising Africa, The World Bank, https://doi.org/10.1596/978-1-4648-0723-7.

[24] Bloom, D. and D. Canning (2007), “Commentary: The Preston Curve 30 years on: still sparking fires”, International Journal of Epidemiology, Vol. 36/3, pp. 498-499, https://doi.org/10.1093/ije/dym079.

[7] Boarini, R., A. Kolev and A. McGregor (2014), “Measuring Well-being and Progress in Countries at Different Stages of Development: Towards a More Universal Conceptual Framework”, OECD Development Centre Working Papers, No. 325, OECD Publishing, Paris, https://doi.org/10.1787/5jxss4hv2d8n-en.

[10] Clio-Infra (2017), Clio Infra, International Institute of Social History, http://www.clio-infra.eu (accessed on 01 July 2018).

[36] Cornia, G. (2016), “The mortality crisis in transition economies”, IZA World of Labor, Vol. 298/October, https://wol.iza.org/uploads/articles/298/pdfs/mortality-crisis-in-transition-economies.pdf.

[18] Dollar, D. and A. Kraay (2002), “Growth is good for the poor”, Journal of Economic Growth, Vol. 7/3, pp. 195-225.

[31] Dréze, J. and A. Sen (2013), An Uncertain Glory: India and Contradictions, Princeton University Press, Princeton.

[38] Engels, F. (1845), The Conditions of the Working Class in England, Stanford University Press, Stanford.

[46] Feinstein, C. (1998), “Pessimism Perpetuated: Real Wages and the Standard of Living in Britain during and after the Industrial Revolution”, The Journal of Economic History, Vol. 58/03, pp. 625-658, https://doi.org/10.1017/s0022050700021100.

[28] He, W., D. Goodkind and P. Kowal (2016), “An aging world: 2015”, Aging Population Reports, , United States Census Bureau, Washington, DC, http://www.census.gov/content/dam/Census/library/publications/2016/demo/p95-16-1.pdf.

[13] Islam, R. (2006), Fighting Poverty: The Development-Employment Link, Lynne Rienner.

[14] Khan, A. (2007), Asian Experience on Growth, Employment and Poverty: An Overview with Special Reference to the Findings of Some Recent Case Studies, ILO, Geneva and UNDP, Colombo.

[34] Kingkade, W. and E. Arriaga (1997), Mortality in the new independent states: Patterns and impacts, National Academy Press, Washington, DC.

[37] Komlos, J. (1998), “Shrinking in a Growing Economy? The Mystery of Physical Stature during the Industrial Revolution”, The Journal of Economic History, Vol. 58/03, pp. 779-802, https://doi.org/10.1017/s0022050700021161.

[16] Lang, V. and M. Mendes Tavares (2018), “The Distribution of Gains from Globalization”, IMF Working Paper, International Monetary Fund, Washington, DC., Vol. 18/54.

[27] Leke, A. and D. Barton (2016), “3 reasons things are looking up for African economies”, World Economic Forum, 5 May 2016, http://www.weforum.org/agenda/2016/05/what-s-the-future-of-economic-growth-in-africa/.

[39] Lindert, P. (2004), Growing Public: Social Spending and Economic Growth since the Eighteenth Century, Cambridge University Press, Cambridge.

[25] Lustig, N., C. Pessino and J. Scott (2013), “The impact of taxes and social spending on inequality and poverty in Argentina, Bolivia, Brazil, Mexico, Peru and Uruguay: An overview”, CEQ Working Paper, Tulane University, New Orleans 1313.

[9] Maddison, A. (2001), The World Economy: A Millennial Perspective, Development Centre Studies, OECD Publishing, Paris, https://doi.org/10.1787/9789264189980-en.

[44] Marx, K. and F. Engels (1848), Communist Manifesto, Progress Publishers, Moscow.

[20] Milanovic, B. (2016), Global Inequality. A New Approach for the Age of Globalization, Harvard University Press, Cambridge.

[30] Moatsos, M. (2017), “Global absolute poverty: Behind the veil of dollars”, Journal of Globalization and Development (forthcoming).

[8] Narayan, D. et al. (1999), Can Anyone Hear Us? Voices From 47 Countries, World Bank.

[41] O’Rourke, K. and J. Williamson (1999), Globalization and History: The Evolution of a Nineteenth-Century Atlantic Economy, MIT Press, Cambridge.

[6] OECD (2017), How's Life? 2017: Measuring Well-being, OECD Publishing, Paris, https://doi.org/10.1787/how_life-2017-en.

[17] OECD (2017), Next Steps for Development in Transition: A Background Paper, presented during the 18 May 2017 meeting in Brussels, Belgium, co-hosted by The Directorate-General for International Cooperation and Development of the European Commission, the OECD Developmen, http://www.oecd.org/dev/BackgroundPaper_DiT.pdf.

[35] OECD (2008), Black Sea and Central Asia: Promoting Work and Well-Being, OECD Publishing, Paris, https://doi.org/10.1787/9789264047303-en.

[22] O'Rourke, K. (1997), “The European Grain Invasion, 1870–1913”, The Journal of Economic History, Vol. 57/04, pp. 775-801, https://doi.org/10.1017/s0022050700019537.

[45] O'Rourke, K. (1997), “The European Grain Invasion, 1870–1913”, The Journal of Economic History, Vol. 57/04, pp. 775-801, https://doi.org/10.1017/s0022050700019537.

[23] Preston, S. (1975), “The Changing Relation between Mortality and Level of Economic Development”, Population Studies, Vol. 29/2, p. 231, https://doi.org/10.2307/2173509.

[1] Rijpma, A., J. van Zanden and M. Mira d'Ercole (2018), “A long-term perspective on the development experience of emerging and industrialised economies”, SDD Statistics and Data Working Paper, forthcoming, OECD Publishing, Paris..

[21] Roser, M. and E. Ortiz-Ospina (2018), "Global Extreme Poverty", OurWorldInData.org, https://ourworldindata.org/extreme-poverty.

[3] Seers, D. (1969), “The Meaning of Development”, IDS Communication, Vol. 44.

[32] Sen, A. (2011), Quality of Life: India vs. China, https://www.nybooks.com/articles/2011/05/12/quality-life-india-vs-china/.

[5] Sen, A. (1999), Development as Freedom, Oxford University Press.

[4] Stiglitz, J., A. Sen and J. Fitoussi (2009), Report by the Commission on the Measurement of Economic Performance and Social Progress, http://ec.europa.eu/eurostat/documents/118025/118123/Fitoussi+Commission+report.

[29] Sulla, V. and P. Zikhali (2018), Overcoming Poverty and Inequality in South Africa : An Assessment of Drivers, Constraints and Opportunities (English). Washington, D.C. : World Bank Group, http://documents.worldbank.org/curated/en/53048152173590653.

[19] Sumner, A. (2016), Global Poverty: Deprivation, Distribution and Development Since the Cold War, Oxford University Press, Oxford and New York.

[42] Szreter, S. (1988), “The importance of social intervention in Britain’s mortality decline c.1850-1914: A re-interpretation of the role of public health”, Social History of Medicine, Vol. 1/1, pp. 1-38.

[43] Szreter, S. and G. Mooney (1998), “Urbanization, Mortality, and the Standard of Living Debate: New Estimates of the Expectation of Life at Birth in Nineteenth-century British Cities”, The Economic History Review, Vol. 51/1, pp. 84-112, https://doi.org/10.1111/1468-0289.00084.

[40] van Bavel, B. and A. Rijpma (2015), “How important were formalized charity and social spending before the rise of the welfare state? A long-run analysis of selected western European cases, 1400-1850”, The Economic History Review, Vol. 69/1, pp. 159-187, https://doi.org/10.1111/ehr.12111.

[2] van Zanden, J. et al. (eds.) (2014), How Was Life?: Global Well-being since 1820, OECD Publishing, https://doi.org/10.1787/9789264214262-en.

[11] World Bank (2018), PovcalNet (database), World Bank, http://iresearch.worldbank.org/PovcalNet/home.aspx (accessed on 01 May 2018).

Annex 3.A. Additional figures

This annex is available on line at https://doi.org/10.1787/persp_glob_dev-2019-10-en

Figure 3.A.1. Average education in Latin America

Figure 3.A.2. Income inequality in Latin America

Figure 3.A.3. Homicide rates in Latin America

Figure 3.A.4. Life expectancy in Latin America

Figure 3.A.5. Quality of democratic institutions (Polity 2) in Latin America

Figure 3.A.6. Labour real wages in Latin America

Figure 3.A.7. Average education in Africa

Figure 3.A.8. Income inequality in Africa

Figure 3.A.9. Homicide rates in Africa

Figure 3.A.10. Life expectancy in Africa

Figure 3.A.11. Quality of democratic institutions (Polity 2) in Africa

Figure 3.A.12. Labour real wages in Africa

Figure 3.A.13. Average education in Asia

Figure 3.A.14. Income inequality in Asia

Figure 3.A.15. Homicide rates in Asia

Figure 3.A.16. Life expectancy in Asia

Figure 3.A.17. Quality of democratic institutions (Polity 2) in Asia

Figure 3.A.18. Labour real wages in Asia

Figure 3.A.19. Average education in Eastern Europe and the former Soviet Union

Figure 3.A.20. Income inequality in Eastern Europe and the former Soviet Union

Figure 3.A.21. Homicide rates in Eastern Europe and the former Soviet Union

Figure 3.A.22. Life expectancy in Eastern Europe and the former Soviet Union

Figure 3.A.23. Quality of democratic institutions (Polity 2) in Eastern Europe and former Soviet Union

Figure 3.A.24. Labour real wages in in Eastern Europe and the former Soviet Union

Figure 3.A.25. Average education in the early industrialising countries

Figure 3.A.26. Income inequality in the early industrialising countries

Figure 3.A.27. Homicide rates in the early industrialising countries

Figure 3.A.28. Life expectancy in the early industrialising countries

Figure 3.A.29. Quality of democratic institutions (Polity 2) in the early industrialising countries

Figure 3.A.30. Labour real wages in the early industrialising countries

Figure 3.A.31. SO₂ emissions per capita in the early industrialising countries

Figure 3.A.32. Stature in the early industrialising countries

Notes

← 1. The chapter presents an analysis of historical well-being for four key regions: Latin America, sub-Saharan Africa, Asia and Eastern Europe and the former Soviet Union. Regional averages are always taken from the How was life? report. Within each of these regions, the chapter takes a closer look at a sub-set of countries chosen either because they are large in terms of population or characterised by contrasting developments within the region: Argentina, Brazil, Chile, Mexico, Peru and Venezuela in Latin America; Burkina Faso, Ghana, Kenya, Nigeria, Uganda and South Africa in sub-Saharan Africa; China, Indonesia, India, the Philippines, Thailand and Viet Nam in Asia; and Bulgaria, Estonia, Hungary, Poland, Romania and Russia in Eastern Europe and the former Soviet Union.

← 2. Unless stated otherwise, “GDP growth” refers to growth in GDP per capita, based on 1990 US dollars.

← 3. How’s Life is the bi-annual report released by the OECD Statistics and Data Directorate since 2011 to monitor, benchmark and analyse well-being in OECD members and selected partner countries. The report relies on headline indicators of current well-being, resources for the future (since 2013) and inequalities (since 2017) selected in consultation with statistical offices of OECD member countries.

← 4. Argentina, Chile, Brazil, Venezuela, Peru, China, Indonesia, India, Viet Nam, the Philippines, Thailand, South Africa, Kenya, Nigeria, Ghana, Uganda, Burkina Faso, Russia, Poland, Hungary, Romania, Bulgaria and Estonia.

← 5. The United Kingdom, the United States, France, Germany, Belgium, Sweden, Japan, Italy and the Netherlands.

← 6. For example, in the education dimension, measures focus on the skills and competencies achieved, rather than on the money spent on schools or the number of teachers trained.

← 7. For instance, Boarini, Kolev and McGregor (2014[7]) recommended to refer to “consumption possibilities”, rather than “income and wealth”, to recognise the prevalence of consumption as a metric of economic well-being in developing countries. They also argued for broadening the concept of “personal security” to “vulnerability”, to reflect the broader range of risks faced by people in developing countries; and for referring to “empowerment and participation”, rather than “civic engagement and governance”, to stress the importance of giving political voice and means of expression to individuals, local communities and indigenous populations. Similarly, on the indicators side, they suggested including measures of vulnerable employment (in addition to the standard unemployment rate used in OECD countries) to account for high labour market informality in these countries.

← 8. How Was Life? omitted some of the dimensions included in the How’s Life? framework because certain dimensions of well-being (such as subjective well-being) cannot be measured in the distant past, while for other dimensions (such as work-life balance) no sufficient historical data are available.

← 9. Self-reporting on life satisfaction and of day-to-day experience did not exist back in time, for instance.

← 10. GDP per capita is expressed in purchasing power parity based on 1990 US dollars.

← 11. Life expectancy and homicide data are only available from 1850.

← 12. This is done by regressing the well-being measures (standardised to have zero mean and unit standard deviation for comparability) on the logarithm of per capita GDP and a set of time dummies. Time dummies capture the additional well-being compared to 1820 (or the earliest year of observation) that is not explained by the level of per capita GDP in that period.

← 13. The polity2 variable is expressed on a 21-point scale, scoring countries between autocracy (low) and democracy (high). This means there is a ceiling beyond which, according to this indicator, political institutions cannot be improved anymore. Further developments in countries already categorised as full democracies or full autocracies can only compensate for developments elsewhere in the world to a limited extent.

← 14. The approach is similar in spirit to that used in Figure 3.5, although without separate time dummies. The model is wb = b0 + b1*log gdppc, where wb is again one of the well-being measures. The relationship is estimated using the full global sample of countries from 1820–2010 (or the earliest/latest date for which data is available). This provides us with predicted values for the well-being measures that can be used to evaluate the actual developments in well-being relative to the volume of economic production of a country.

← 15. If less than 40% of the regional population is covered by the data, it was set to missing.

← 16. The growth spurt in the 2000s did not go unnoticed. Sub-Saharan Africa was identified as one of the rising stars of the world economy in 2010 by the McKinsey Global Institute, which described the potential and progress of African economies as “lions on the move”; see www.mckinsey.com/featured-insights/middle-east-and-africa/lions-on-the-move.

← 17. According to Moatsos (2017[30]), however, the decline was less steep, from 65% in 1983 to 22% in 2014, which is more similar to the experience of India.

← 18. As noted by Sen (2011[32]), “Life expectancy at birth in China is 73.5 years; in India it is 64.4 years. The infant mortality rate is 50 per thousand in India, compared with just 17 in China; the mortality rate for children under five is 66 per thousand for Indians and 19 per thousand for the Chinese; and the maternal mortality rate is 230 per 100 000 live births in India and 38 in China. The mean years of schooling in India were estimated to be 4.4 years, compared with 7.5 years in China. China’s adult literacy rate is 94%, compared with India’s 74% according to the preliminary tables of the 2011 census”.

← 19. The governance indicator is the one exception to this pattern.

← 20. In the 1960s and 1970s, the “standard of living debate” focused mainly on the development of real wages during industrialisation in England (cf. Feinstein (1998[46]); since then, this debate has broadened via the systematic analysis of new data made available by the study of population height (see van Zanden et al., (2014[2]), Ch. 7).

← 21. Two limits in analysis included in this chapter should be mentioned. First, the evidence presented is limited to data on the average level of education, life expectancy, security and political rights of the countries studied. The chapter assumes that distribution of these dimensions of well-being among the population has not changed dramatically over time. It also assumes it is possible to assess, for example, the level of education of the population based on average measures alone.

Second, interpreting the link between GDP and well-being outcomes needs to recognise the limits of the analysis. A positive relationship between GDP per capita and well-being outcomes does not necessarily mean that GDP growth drives improvement in well-being, or that it is the only important driver at work. Good health, long life expectancy, a high level of education, secure political rights, and high personal security, besides implying higher well-being per se, also contribute to economic growth. In other words, the causality runs in both directions. This implies that investments in education or health, as well as in other well-being dimensions, may be a better way to further GDP growth and well-being than stimulating GDP growth through other means.

← 22. “Dark satanic mills” alludes to William Blake’s poem “And did those feet in ancient time”, which in turn refers to the industrial revolution and its destruction of nature and human relationships.

End of the section – Back to iLibrary publication page