Inequality and Poverty

Abstract

This chapter assesses how the rewards of sustained economic growth were distributed over time. The evolution of income inequality resembles a wide inverted U with a peak in 1916, and a Kuznets curve results when the Gini coefficient is plotted against real per capita income. The functional distribution of income led personal income distribution until the early 1950s, while the dispersion of labour incomes took over from then onwards. Although growth and inequality do not move together over time, in the last century, the main phases of economic growth went hand-in-hand with inequality decline. The substantial fall in absolute poverty resulted from growth but also from inequality reduction during the interwar period and the late 1950s, and was eradicated by 1975.

An earlier version was published as L. Prados de la Escosura (2008), “Inequality, Poverty, and the Kuznets Curve in Spain, 1850–2000”, European Review of Economic History 12(3): 287–324. This chapter draws on it but includes a deep revision and extension of the estimates covering the post-2000 period and a full re-working of the text.

[S]peculation is an effective way of presenting a broad view of the field; and so long as it is recognized as a collection of hunches calling for further investigation rather than a set of fully tested conclusions, little harm and much good may result (Simon Kuznets, 1955: 26)

1 Introduction

This chapter aims to assess long-run inequality and the joint impact of growth and inequality on absolute poverty. Modern Spain provides a good case study, as this is a mid-size country that experienced a long and painful transition from the Ancien Régime to a liberal society during the nineteenth century, broken by revolutions and civil strife; a short and convulsive democratic experience, followed by a bloody civil war (1936–1939); and long-lasting autocracy under General Franco (1939–1975) until the emergence of a liberal-democratic society.

Since the mid-nineteenth century, Spain has seen irreversible modern economic growth. Real Net National Disposable Income per person multiplied by 13.5 over 170 years, which represents an average growth rate of 1.5% per year (Fig. 5.1). But how much of this growth percolated through to reach the lower deciles of the income distribution and had an impact on absolute poverty reduction? This is the question addressed in this chapter, which consists of five sections. Lack of direct income distribution estimates based on microeconomic evidence prior to 1973 led me to resort to an indirect macroeconomic approach to appraising inequality (Sect. 5.2), and on the basis of the available information, to reconstruct the Gini coefficient and provide an aggregate picture of the evolution of inequality since the mid-nineteenth century (Sect. 5.3). Section 5.4 offers some explanatory hypotheses for inequality trends. And Sect. 5.5 attempts to calibrate the impact of growth and inequality on absolute poverty. The chapter closes with some hypotheses for further research.

Fig. 5.1

Real per capita net national disposable income, 1850–2020 (2010 Euro, natural logs) (Prados de la Escosura, 2017, updated)

The main findings are as follows. The evolution of income inequality resembles a wide inverted U with a peak in 1916, and when the Gini coefficient is plotted against real per capita income, a single Kuznets curve results. Economic rather than political forces appear to have driven long-run trends in Spanish income distribution. Stolper-Samuelson forces only partially explain inequality trends. World and civil wars affected inequality but lacked permanent effects, and progressive taxation had no impact until the 1980s. Economic growth, together with a decline in inequality during the interwar years and between the mid-1950s and the early 1970s, led to a long-run reduction in absolute poverty. The fall in inequality since the mid-1950s and the eradication of absolute poverty by the early 1970s represented major departures with respect to Latin America’s patterns and matched those followed by OECD countries.

The chapter’s results provide some hypotheses for further research. The Civil War (1936–1939) occurred after one and a half decades of declining inequality and an alleviation of poverty, offering an interesting paradox. There was an ‘overshooting’ of inequality, possibly as a consequence of the Civil War, during the early years of Franco’s dictatorship, in which an association between isolation, sluggish growth, and inequality resulted in high levels of absolute poverty. The late Francoist period appears as a benign phase of economic development in which growth and structural change contributed significantly to alleviating inequality and eradicating absolute poverty.

2 Assessing Long-Run Inequality

In the absence of direct estimates of income distribution (household budget surveys) prior to 1973, an alternative approach is needed.¹ 1999Historical evidence on income distribution in Spain in the ‘pre-statistical era’ (i.e. before 1973) is as unsatisfactory as is often the case for present day developing countries.² Any attempt to provide orders of magnitude for personal income distribution over such a long time span is perhaps too audacious, but could be justified in so far it provides hypotheses for future researchers to test.

The scattered and asymmetric time coverage (mostly post-1960) of conventional inequality datasets across countries has prompted attempts to construct alternative inequality measures on the basis of miscellaneous information (factor incomes, salary differences across professions, tax returns, etc.). My approach here is an eclectic one, in which choosing between wage and salary dispersion and property income’s share in total income is avoided, and all are used to depict trends in aggregate inequality.³ Thus, for example, the association between the functional and the personal distribution of income is explored.⁴

I will begin with the simplest case in which only two social groups, property owners (who do not receive returns for their labour) and workers (who do not own property) exist. In order to ascertain the evolution of income inequality, we need to know the gap between the average income of the two groups, as well as the dispersion of income within each group. Classical economists stressed the breach between average returns to proprietors and to workers. As David Ricardo (1817) asserted,

The produce of earth—all that is derived from its surface by the united application of labour, machinery, and capital, is divided among three classes of the community, namely, the proprietor of the land, the owner of the stock or capital necessary for its cultivation, and the labourers by whose industry it is cultivated. To determine the laws which regulate this distribution is the principal problem in Political Economy.

The classical economists’ focus on the functional distribution of income was grounded on the implicit assumption that, as the overwhelming majority of workers were unskilled (and lived near subsistence), the variance of labour incomes was very low. Later, as the economy developed and physical and human capital deepened, skilled workers increased their share within the labour force and, hence, the dispersion of labour returns rose (Kuznets, 1955). Thus, the implied conjecture is that, in early stages of development, income inequality is driven by the gap between average returns of proprietors and workers and only later, as economic progress takes place, is personal income distribution driven by dispersion of factor returns (labour, in particular). If confirmed, this interpretation would help to explain why societies are more sensitive to different types of inequality over time.

Thus, in order to ascertain long-run trends in personal income distribution, we need to assess both between- and within-group inequality. However, historians and social scientists often focus on only one of these at a time. Thus, while the Williamson index, the property (capital and land) share in national income, and, it could be added, the top income shares approach are examples of between-group inequality measures, the skill premium, skilled-unskilled wage gaps, and wage dispersion illustrate the emphasis on within-group inequality. Let us briefly examine some of these approaches for the Spanish case.

A major endeavour to derive yearly series of top income (and wealth) shares in national income for a growing sample of countries in the twentieth century was undertaken by Atkinson, Piketty, Saez and their associates on the basis of income tax statistics. This appealing approach, rooted in Kuznets (1953) classical work⁵ has, nonetheless, important shortcomings: only a very small fraction of the population was subjected to individual income taxation in many countries prior to the late-twentieth century, while fraud and tax evasion challenge the reliability of fiscal records as we move back in time or focus on countries with low quality-institutions. The historical case of Spain seems to fit this picture. High levels of fiscal evasion characterized the Spanish economy until the late twentieth century. Lack of political will to enforce taxation implied that no actual means (statistical records, bureaucracy) were available to fight evasion and fraud until the 1980s.⁶ In fact, income tax only became widespread from 1979 onwards, after a fiscal reform took place, and its share of total tax receipts went up from less than 2% over 1940–1978 to 30% in the early 1980s (Comín, 1996).⁷ Alvaredo and Saez (2009) applied this approach to Spain since the early 1930s. One of their main findings is that income concentration was much higher in the 1930s than at the end of the twentieth century. Their figures for the top 0.01% income share show a dramatic decline between 1935 and 1961,⁸ especially marked throughout the 1940s, and suggest stability between 1961 and 1981.⁹ Top income shares increased in the last two decades, as the joint outcome of top salary increases and capital gains.¹⁰

An alternative measure of inequality has been put forward by Jeffrey Williamson (1997), who proposed an ‘inequality index’ defined as the ratio between GDP per worker and the unskilled wage (y/wus), which has the advantage of being easily computable for most countries over long time spans.¹¹ The rationale for y/wus is that while the numerator reflects returns to all factors of production, the denominator only encapsulates returns to raw labour, so it compares the middle to the bottom of income distribution. Nonetheless, it is worth stressing that as societies develop and broad capital deepening takes place, the proportion of unskilled workers within the labour force dwindles. In this scenario, comparisons over time tend to be inconsistent and inequality measured by y/wus tends to be over exaggerated (upward biased). An alternative is to use the average returns to all labour (w), including both skilled and unskilled workers, as the denominator in the inequality index.¹² This alternative measure (y/w) is similar to the inverse of share of labour compensation in national income¹³ under the assumption that the return per head of self-employed workers matches the average compensation of employees in their corresponding industry.¹⁴ In other words, this approach identifies the functional with the personal distribution of income.

As returns to unskilled workers represented most of labour compensation in national income until the second half of the twentieth century, inequality indices computed with either unskilled (y/wus) or average wages (y/w) might be expected scarcely to differ up to the 1950s. Thereafter, as skilled labour increased its share in national income, large disparities between these alternative indices can be anticipated.¹⁵ The two short-cut measures are opposed in Fig. 5.2 and, as predicted, no major discrepancy between their trends is observed, except for the lower level of y/w in the nineteenth century, until 1970, when a widening gap between the two inequality indices steadily opened up and the Williamson index y/wus experienced a sustained and dramatic increase.¹⁶ Thus, as the unskilled labour share in the workforce declines, the significance of y/wus as a measure of inequality fades away.

Fig. 5.2

Alternative Williamson indices, 1850–2020 (1913=100). Source: See text

The y/w provides a measure of inequality only in so far as the dispersion within labour and property compensation does not change significantly.¹⁷ The assumption of stability in wage dispersion as the proportion of skilled workers within total employment increases is, however, entirely unrealistic (Kuznets, 1955). In fact, within-group inequality measures such as wage inequality or wage gaps have often been used as a short-cut for the evolution of personal income distribution.¹⁸ The bottom line of this assessment of alternative inequality measures is that no conclusion can be reached about trends in total inequality unless different components, namely, the gap between property and labour returns and the dispersion within both property and labour, are simultaneously considered. This this suggests the need for a historical reconstruction of total (between- and within-group) inequality.

3 A Reconstruction of Aggregate Inequality: The Gini

Income inequality over the long run can only be estimated on the basis of scattered and miscellaneous information. One possibility is to start with the breakdown of an inequality index and to build this by estimating each of its components and adding them up. Branko Milanovic (2005: 20–2) proposed a decomposition of the Gini coefficient as follows,

$$ Gini=\sum {G}_i{n}_i\;{\uppi}_i+\sum \left(\left({y}_p\hbox{--} {y}_w\right)/{y}_l\right){\uppi}_w\;{n}_p+L $$

(5.1)

Where the first part of the right hand term, ∑ G_i n_i π_i (Gini A, hereafter) is a weighted sum of within-group inequality, G being the Gini coefficient for each group (_i) and n_i and π_i the group’s shares in population and national income, respectively. In this case, I have only distinguished two groups, workers and proprietors.

The second element, ∑ ((y_p – y_w)/ y_w) π_w n_p (Gini B, henceforth), corresponds to between-group inequality. Groups are ranked according to their mean income, so property owners (y_p) appear above those getting labour returns (y_w) and their relative distance ((y_p – y_w)/ y_w) is weighted by the product of the labour returns’ share in national income (π_w) and the property owners’ share in population (n_p).¹⁹ Average incomes of proprietors and workers have been obtained as follows,

$$ {y}_p={\pi}_p\;\mathrm{GDP}/{n}_pN $$

(5.2)

$$ {y}_w={\pi}_w\;\mathrm{GDP}/{n}_wN $$

(5.3)

where N is total population.

Finally, L is the overlapping, or residual component, and it accounts for the fact that someone who is a property owner may still have a lower income than someone who is a worker and only gets labour returns.

How can the different components of the Gini, Gini A and Gini B, be estimated? Since GDP and population are available (Prados de la Escosura, 2017, updated), all we require is the Gini of earnings within each group, proprietors and workers, and the shares of labour (w) and property (p) in national income (π_w and π_p) and in population (n_w and n_p).

In the case of labour returns, inequality has been proxied by the dispersion of average annual nominal wage earnings across industries (1850–1900, 19 sectors; 1900–1954, 21; 1954–1985, 24; 1985–1995, 53; and 1995–2021, 63). Subsequently, the resulting inequality measures for each of these five periods have been spliced into a single one using their ratios in overlapping years. Thus,

$$ {{\mathrm{Gw}}_{\mathrm{i}}}^{\prime }=\left(\mathrm{G}{\mathrm{w}}_{\mathrm{o}}^{\prime }/{\mathrm{Gw}}_{\mathrm{o}}\right){\mathrm{Gw}}_{\mathrm{i}} $$

(5.4)

Where Gw_i’ represents the wage Gini series closer to the present (and with wider coverage of industries) and Gw_i, the more remote one (with narrower coverage), while Gw’_o/ Gw_o represents their ratio in the year they overlap (Fig. 5.3) (see Appendix, A.1 Sources).

Fig. 5.3

Spliced Wage Gini, 1850–2020. Source: See text

In the case of returns accruing to property, lack of direct evidence has forced me to assume that their dispersion was higher but evolved with that of wages. Property ownership of capital and land has been highly concentrated in Spain (Martin, 1990; Simpson and Carmona, 2020: 157) and the distribution of property has usually been considered to be more uneven than that of labour incomes (Pigou, 1920, cited in Dumke, 1988: 12). Since the highest wage inequality corresponds to 1850, I allocated an arbitrary value of 0.8, twice the peak for wage dispersion, to that year and moved it through time with the rate of variation of wage dispersion.²⁰

Comparing the wage dispersion with the top income shares in national income provides a crude test for my proposition, as the latter could be largely seen as a historical proxy for the concentration of proprietors’ earnings.²¹ It appears that, except for the early 1950s and from the late 1990s onwards, their tendencies are largely coincidental (Fig. 5.4).

Fig. 5.4

Wage Gini and top income share (0.01%), 1929–2015. Sources: Wage Gini, text; Top income share, Alvaredo and Saez (2009, updated)

The next step is to ascertain the shares in national income and in population of those who get returns exclusively from either labour or property. For the period 1850–1954, I obtained the amount of labour compensation by multiplying daily wage rates by the number of days worked in each industry, and adding them up. For the post-1954 period, modern national accounts distinguish two income components: compensation of employees (wages and salaries) and gross mixed income, which includes incomes accruing to proprietors and to the self-employed. Income components from different rounds of official national accounts were spliced through linear interpolation to obtain a consistent series for the entire period (see Prados de la Escosura, 2017: 173–174).

But what proportion of gross mixed income corresponds to returns to labour? Colin Clark (1957) and Simon Kuznets (1966) favoured the approach of attributing to entrepreneurs and self-employed workers an average labour income equal to the average employee compensation.²² I have, therefore, assumed an average return for non-wage labour identical to that of wage earners in each industry, and derived the income accruing to labour by dividing the amount of wages and salaries by the share of wage earners in the labour force. Then, the labour income share (π_w) was obtained by dividing total labour compensation by GDP at market prices (Chap. 4).²³

The two labour income share series (1850–1954 and 1954–2021) overlap in 1954 but their respective levels do not match. As compromise solution, I have distributed the gap between the two series in the overlapping year T (1954) at a constant rate over 1945–1954.

$$ {\uppi}_{wt}={\uppi^n}_{wt}\ast {\left[{\left({\uppi^n}_{wT}/{\uppi^o}_{wT}\right)}^{1/n}\right]}^t\kern2em \mathrm{for}\ 0\le t\le T $$

(5.5)

π_w being the linearly interpolated new series, πⁿ_w and π^o_w the values pertaining to the labour share according to the 1850–1954 and 1954–2021 series, respectively; t, the year considered; T, the overlapping year (1954) between the two series; and n, the number of years in between the initial (0) (1945), and the final (T) (1954), dates considered. Then, the property income share (π_p) was derived as a residual (π_p = 1 − π_w).²⁴

The breakdown of the population into the ‘equivalents’ of those whose income exclusively accrues from property and from labour, while avoiding any overlapping between these two groups, provides a further challenge and only a crude and arbitrary procedure has been possible in its estimation. As for the first 100 years considered, population censuses only provide figures of proprietors for 3 years, 1860, 1920, and 1950. I computed the share of proprietors in working age population (15–64) for these 3 years plus 1960 and linearly interpolated the resulting figures to derive a crude annual series. As regards the post-1954 era, I firstly computed the proportion of property income in gross mixed income and, then, applied this ratio to the share of non-occupied population in working age population in order to obtain a rough proportion of ‘equivalent’ property owners (that is, the share of economically active population whose income derives exclusively from property).²⁵ However, a possible objection to the estimate is that the average proprietor was probably richer than the average person earning non-wage income, so their actual number would be lower. Moreover, the estimate may include the self-employed and, hence, overstate the number of proprietors. In order to allow for this objection, I have assumed that the income of the average proprietor was twice that of the average self-employed person, and proportionally reduced the number of proprietors previously estimated. Interestingly, the share of proprietors in working age population (n_p) for the late 1950s obtained this way matches closely that derived through interpolation for the pre-1960 period. Then, the pre-1954 series were re-scaled with the average ratio between the two estimates for the overlapping years 1954–1960 (1.038). Lastly, I obtained the share of the ‘equivalent’ population whose returns derived exclusively from labour as a residual (n_w = 1−n_p) for the entire time span considered, 1850–2020.

As regards the overlapping L component, since it cannot be computed directly, an indirect procedure has been used. Household expenditure Gini on the basis of household surveys are available for 1973/1974, 1980/1981, and 1990/1991 (Goerlich and Mas, 2001) and from 1993 onwards (National Statistical Institute [INE]).²⁶ I have computed the annual ratio between the directly computed Gini and the ‘historical’ Gini estimate (that is, Gini A + Gini B) over 1973–2000. The average ratio can be employed to correct the ‘historical’ Gini over 1850–1972. For the missing years (1975–1979, 1982–1989, and 1994), the Gini was interpolated by projecting back and forth the closest available direct Gini with the “historical Gini” and, then, computing a variable weighted geometric average in which the closest benchmark receives a larger weight. The overlapping component L results from the difference between the aggregate Gini and the ‘historical’ Gini (Gini A + Gini B). It is worth noting, however, that resulting overlapping component L not only captures the fact that someone who is a property owner may still have a lower income than someone receiving labour returns, but also any measurement errors in the computation of Gini A and B.²⁷

Trends in aggregate inequality, measured by the Gini coefficient, are presented in Fig. 5.5. Needless to say, they merely represent a set of explicit hypotheses about income distribution in modern Spain. The evolution of inequality presents the shape of a wide inverted U between 1880 and 1976 with a peak in 1916.

Fig. 5.5

Income inequality in Spain, 1850–2020: Gini coefficient. Source: See text

Different long swings can be observed in the evolution of inequality. A long-term rise is noticeable during the early phase of globalization that peaked by World War I. The interwar period shows a marked reduction in inequality in two phases, up to 1923 and in the early 1930s, stabilised during the Civil War (1936–1939) and sharply reversed during the autarchy years, with peaks in 1944 and 1953. After a dramatic fall during the second half of the 1950s, inequality stabilised, before exhibiting a steady contraction in the early 1970s. From the mid-1970s to the present, the Gini has fluctuated within a narrow range (0.31–0.35), with peaks in 1997 and 2014.

If we now look at the composition of the Gini, two distinctive phases emerge (Fig. 5.6). From 1850 to 1950s, Gini B, i.e. between-group inequality, dominated personal income distribution. The reason is that, as unskilled labour represented the overwhelming majority of employment, the gap between property and labour returns drove aggregate inequality. Then, from the mid-1950s onwards, as the economy initiated a process of accelerated growth and structural change, skilled labour increased its share of employment and the dispersion of labour returns rose; thus, Gini A, or within-group inequality, became the main driver of personal income distribution.

Fig. 5.6

The Gini and its Gini A and Gini B components, 1850–2020. Source: See text

The fact that differences between returns to property and to labour dominated inequality trends during the first century of modern economic growth in Spain confirms that functional distribution of income is an appropriate proxy for personal income distribution in early stages of development.

Does the evolution of personal income distribution fit a Kuznets curve? In the historical literature, there have been challenges to this venerable hypothesis (Lindert, 2000; Rossi et al., 2001). The Kuznets hypothesis associates the evolution of inequality with economic growth (Kuznets, 1955). Thus, the relevant test is to compare levels of inequality and per capita income. In Fig. 5.7, the Gini Hodrick-Prescott trend is plotted against the natural logarithm of real GDP per head, and a single Kuznets curve emerges.²⁸

Fig. 5.7

The Kuznets curve (Kernel Fit, Epanechnikov, h=0.4042). Source: See text

4 Interpreting Inequality

How can these inequality trends be interpreted? Different hypotheses have been put forward in the literature. Alvaredo et al. (2013) have underlined external shocks and progressive income tax as major determinants of inequality trends. Specifically, World Wars and the Great Depression negatively affected the top incomes share in national income (in particular, capital income concentration) while progressive taxation did not allow its recovery. Significant changes, not always coincidental with those taking place in Western Europe, occurred in Spain during the period 1914–1950. Besides, the potential impact of progressive taxation was reduced by its delayed introduction in Spain (1979).

World War I represented a major shock for Spain: relative prices changed so dramatically that they may have affected income distribution (Prados de la Escosura, 2017; Rosés and Sánchez-Alonso, 2004). The increase in inequality observed in Spain during World War I has also been identified in other neutral countries (Denmark and the Netherlands) as profits rose due to increases in foreign demand and import substitution, while wages did not keep up with rising prices (Morrisson, 2000: 249). This evolution is at odds with that of belligerent countries during World War I. Moreover, the fall in income inequality resulting from ‘destruction, inflation, bankruptcies, and fiscal shocks for financing wars’ (Atkinson et al., 2011; Alvaredo et al., 2013) that occurred in France, Japan, or the U.S. is missing in Spain (a non-belligerent country during World War II), where the decline in inequality in the early 1930s was more than offset by the re-distribution of income towards property owners after the Civil War.

Alvaredo and Saez (2009) suggest a dramatic fall in top income shares inequality during the first two decades of Francoism. However, the behaviour of top income shares does not explain the evolution of the Gini in post-World War II Spain (Fig. 5.8). It could be argued that, in fact, the rise in total inequality was not determined by changes in the concentration of capital income—that would have fallen, according to the decline in top income shares (Alvaredo and Saez 2009)—, but by an increase in the share of property income within total income (Fig. 5.6). Thus, the distinction between Spain, where the Civil War had a divisive effect in the society, and most Western European countries, where World Wars tended to increase social cohesion, may be relevant to understand their post-war differences.

Fig. 5.8

The Gini and top income share (0.01%), 1850–2020. Sources: Gini, see text; Top income share (0.01%), Alvaredo and Saez (2009, updated)

How can we explain changes in the functional distribution of income? A clue is provided by Christian Morrisson’s (2000: 251) remark that the institutional design historically guaranteed rents to proprietors but not to unskilled workers. Tariff protectionism, for example, could be interpreted in this light and the Stolper-Samuelson model used to provide explicit hypotheses about inequality trends (Williamson, 2002). Does this model apply satisfactorily to the case of Spain?

The fall in inequality during phases of opening up to international competition (the late 1850s and early 1860s, the late 1880s) and the rise in inequality (from the late 1890s to the end of World War I) coinciding with a return to strict protectionism, could be predicted within a Stolper-Samuelson (1941) framework that posits that protectionism favours the scarce factors (land and capital, in this case) while it penalizes the abundant one (labour). In Spain, at the turn of the nineteenth century, redistribution towards the owners of scarce factors would have been reinforced by the fact that tariff protection did not drive out workers as in other protectionist European countries (i.e. Italy and Sweden). The depreciation of the peseta in the 1890s and early 1900s made the decision to migrate more difficult, as the cost of passage increased dramatically (Sánchez-Alonso, 2000, 2007). The Stolper-Samuelson model fails, however, to explain the rise in inequality between the mid-1860s and early 1880s.²⁹

The reduction in inequality during in both the early 1920s and 1930s, within a phase of globalization backlash, would not be consistent within a Stolper-Samuelson framework.³⁰ Other major forces conditioned the evolution of inequality. Accelerated growth and structural change all helped to reduce total inequality in the 1920s. Wage inequality rose with rural-urban migration and urbanization, given that urban wages were higher than rural wages, but the gap between returns to property and labour declined.³¹ Institutional reforms that included new social legislation, especially the reduction in the number of working hours per day, and the increasing voice of trade unions, contributed to a rise in wages relative to property incomes (Cabrera and del Rey, 2002; Comín, 2002).

The fall in inequality during the early 1930s, at the time of increasing restrictions to commodity and factor mobility, is, again, at odds with the Stolper-Samuelson view. Forces pushing for re-distribution were in place in Spain. On the whole, a reduction in the gap between returns to property and labour more than offset the rise in wage inequality (See the behaviour of Gini B and Gini A in Fig. 5.6). The Great Depression may have had a negative impact on top income shares by reducing property income concentration, as Piketty and Saez would expect.³² Wages (nominal and real) certainly rose in a context of increasing bargaining power of the trade unions and labour unrest. In the early 1930s, new legislation that tended to increase labour costs, threats to land ownership, and attempts at factory control by workers created insecurity among proprietors, leading to a severe investment collapse and polarization in Spanish society (Comín, 2002: 294–295, Cabrera and del Rey, 2002: 221–235; Simpson and Carmona, 2020: 201–204).³³

How can the evolution of inequality during the post-Civil War, autarchic years (1939–1953) be interpreted? After the inequality reduction resulting from the war itself and from the pro-labour policies of the II Republic, Franco’s victory reversed the inequality decline. Wage compression resulted from the re-ruralisation of Spanish economy (the share of agriculture increased in both output and employment) and the ban on trade unions. Simultaneously, there was a parallel decline in the 0.01% top income shares during the 1940s. Thus, while inequality was falling within both labour and capital returns, the gap between property and labour widened, leading to a rise in total inequality. The autarchy years provide, hence, a mirror image of the early 1930s. International isolation, resulting from autarchic policies, would intensify these trends, with inequality rising as scarce factors, land and capital, were favoured at the expense of the abundant and more evenly distributed factor, labour.

A dramatic decline in inequality occurred during the 1950s, that is, prior to the conventional phase of liberalization and opening up that followed the 1959 reforms (Chap. 8). A possible hypothesis is that this was triggered by economic agents’ increasing confidence in the viability of Franco’s dictatorship after the U.S.-Spanish cooperation agreements (Calvo-González, 2007) that led to imports of new vintage equipment and to an increase in the investment rate. Between 1953 and 1958, a spurt of economic growth led to improvements in living standards (private consumption grew parallel to per capita GDP), urbanization, and an increase in the labour share within national income (Prados de la Escosura, 2017). Furthermore, the populist policies of Franco’s Minister of Labour led to a substantial pay rise across the board in 1956 (Barciela, 2002).

It appears, then, that international economy forces were not alone in playing a role in reducing inequality during the second half of the twentieth century. Growth and structural change played a not inconsiderable role. The rise in savings, helped by the financial development that accompanied economic growth (Comín, 2007; Martín-Aceña and Pons, 2005), facilitated access to housing ownership which, in turn, helped reduce the concentration of property incomes. The diffusion of education (Núñez, 2005) certainly played a role in the decline of inequality by reducing the concentration of human capital. Furthermore, the decrease in regional disparities, conditioned by technological catch-up, the generalization of basic education, and the spatial redistribution of employment (de la Fuente, 2002; Martínez-Galarraga et al., 2015; Díez-Minguela et al., 2018), must also have impacted income distribution.

The coincidence between the social policies of the late Francoism and the cautious opening up of the economy could perhaps be interpreted in terms of an association between exposure to international trade and the weight of the government sector (Rodrik, 1997). Even though the modern welfare state was not fully introduced in Spain until the transition to democracy, social expenditures had already increased during late Francoism and must have had an effect on reducing inequality.³⁴ The share of social spending in GDP went up from 3.9% in 1958 to 12.1% in 1974, representing limited catching up with Western Europe’s share (Espuelas, 2012: 214).³⁵

Increasing political participation after democracy was reinstated in 1977 led to a progressive fiscal reform and to substantial increases in public expenditure on social transfers (unemployment, pensions, education, and health) that had a substantial redistributive impact, as observed when inequality before and after taxes and social transfers are compared (Fig. 5.9). However, the Gini of disposable income has remained trendless, fluctuating within a 0.31–0.35 Gini range since 1973. It clearly emerges that progressive redistribution accounts for the stability of disposable income distribution, while the market or pre-fisc Gini (that is, prior to taxes and transfers) has increased to levels comparable to the 1916 peak (or, by the same token, to present day Brazilian levels). However, the stability of the post-fisc Gini poses the question of why the inequality of disposable income has not fallen since the instauration of democracy in Spain (Torregrosa-Hetland, 2015).

Fig. 5.9

Inequality before and after taxes and transfers, 1970–2019: Post-fisc Gini and progressive redistribution. Sources: 1970–1990, Torregrosa Hetland (2015); 2007–2019, OECD

How does the case of Spain compare to other historical experiences? Estimates for aggregate income inequality over the long run are only available for a few OECD countries. Denmark, Norway, Italy, and the U.K. have Gini estimates dating back to the late nineteenth century, as do Japan and the U.S. Some crude historical estimates of inequality for Latin America are also available (Prados de la Escosura, 2007). However, comparability problems between Gini estimates constructed using different kind of data have led to a focus on trends rather than on levels (Gottschalk and Smeeding, 2000: 285). Hence, the historical evidence on Gini estimates I am presenting for a handful of countries should be taken with a grain of salt. Figure 5.10 indicates that Spain matched the behaviour of OECD countries except for the autarchic period that followed the Civil War.³⁶ Interestingly, the comparison with Italy in the twentieth century depicts the latter as a case of more benign development. The contrast with the case of Latin America is illuminating (Prados de la Escosura, 2007). Contrary to the usual assumption of high and enduring inequality in Latin America since colonial times, an upward trend until the 1960s brought inequality to the high plateau, where it stabilized for the rest of the twentieth century. Spain and Latin America followed similar patterns until the mid-1950s, when Spain shifted away to converge towards OECD inequality levels.

Fig. 5.10

Inequality in OECD countries, 1870–2019. Source: See text, fn. 37

5 Trends in Absolute Poverty

How do trends in inequality and economic growth impinge on poverty reduction over the last century and a half? In this section, I will calibrate trends in absolute poverty from which hypotheses for further research could be derived.

I will focus on the absolute growth of the incomes of the poor (Ravallion and Chen, 2003) rather than on whether these experienced a relatively disproportionate growth (Kakwani and Pernia, 2000); therefore, the evolution of absolute poverty will be defined with reference to a fixed international poverty line.

If a fixed poverty line (PL) is defined at $2.10 (1990 purchasing-power adjusted international dollars) per person and day,³⁷ it was not until 1880 that Spanish average incomes (real net national disposable income per capita) doubled the poverty line and until 1930 that the latter was trebled. If we bear in mind the results from empirical research in developing countries (for example, Bourguignon, 2002; Klasen, 2004; López, 2004; Ravallion, 1997, 2004) such a low level of development probably hampered the impact of growth on poverty reduction (Deininger and Squire, 1998). In the ongoing debate on pro-poor growth, few views are shared. One of them is that the higher the initial level of inequality, the lower the reduction in poverty for a given rate of growth in GDP per head. Thus, poverty reduction would depend on the initial level of average income and its subsequent growth, on the initial income distribution and its evolution over time, and on how sensitive poverty is to growth and inequality changes (Bourguignon, 2002; Ravaillon, 2004; López and Servén, 2006).

How much impact would average incomes growth and distribution changes have had, then, on absolute poverty in the case of Spain? During the nineteenth century and up to World War I, low per capita income and increasing inequality may have drastically reduced the impact of economic growth on poverty. High initial inequality would have also mitigated the effect on poverty of the acceleration in economic activity during the 1920s, as would have been the case during the 1953–1958 growth recovery. Moreover, faltering growth in the early 1930s would have weakened the effect of falling inequality on poverty reduction. The unprecedented growth of the 1959–1974 years suggests, however, that as the low initial income constraint was removed, the impact on poverty intensified.

Can these hypotheses be put to the test? Alas, no microeconomic data are available on Spain’s household expenditures to compute poverty levels and trends before the late twentieth century. In these circumstances, François Bourguignon and Morrisson’s (2002) assumption that income distribution remained unaltered in Spain from the early nineteenth to the mid-twentieth century has much in its favour. In such a case, it would suffice to know the growth rate of GDP per head to assess the evolution of absolute poverty over time. In fact, some researchers suggest that a large proportion of long-run changes in poverty are accounted for by the growth in averages incomes (Kraay, 2006), and, consequently, emphasize the protection of property rights, stable macroeconomic policies, and openness to international trade as simultaneous means to achieve growth and suppress absolute poverty (Klasen, 2004). Assuming a one-for-one reduction in poverty with per capita GDP growth seems to be, however, a gross misrepresentation,³⁸ and I have therefore preferred to rely on the available macroeconomic evidence on growth and changes in income distribution to propose conjectures about historical trends in absolute poverty.

I have calibrated the impact of growth and inequality changes on absolute poverty (those living below G-K 1990 $2.10) for the case of Spain on the basis of López and Servén (2006), who, drawing on a large micro database for a wide sample of developing and developed countries over the last four decades and using a parametric approach, found that the observed distribution of income is consistent with the hypothesis of log-normality. Under log-normality, the contribution of growth and inequality changes to poverty reduction only depends on the poverty line/average incomes ratio, and on a measure of inequality (the Gini coefficient). The poverty headcount, P_o, that is, the share of population below the poverty line, is derived as,

$$ {\mathrm{P}}_{\mathrm{o}}=\Phi \left(\log \left(z/\upnu \right)/\upsigma +\upsigma /2\right), $$

(5.6)

$$ \mathrm{where}\ \upsigma =\surd 2\;{\Phi}^{-1}\left(\left(1+G\right)/2\right) $$

(5.7)

in which Φ, is a cumulative normal distribution; ν, the average per capita income; z, the poverty line; σ, the standard deviation of the distribution; and G, the Gini coefficient. Thus, all I need to calibrate the poverty headcount is the poverty line/average income ratio and the Gini coefficient.

A long-run decline in absolute poverty is observed in Fig. 5.11 (continuous line). Poverty reduction occurred, nonetheless, at differing speeds over time—a result that supports the view that the impact of growth on poverty is weakened in the presence of rising inequality and low initial levels of development—, while once the initial income constraint is released, its effect heightens. A major contraction took place between 1850 and 1880, which reverted its trend and peaked at the beginning of the twentieth century. Growth underlies the fall in absolute poverty over the third quarter of the nineteenth century, as inequality did not change substantially. Sluggish growth and rising inequality explain the increase in absolute poverty from the 1880s to the end of World War I. The sharp decline in absolute poverty during the interwar years (1919–1935) was the combined outcome of a sustained fall in inequality in the early 1920s and 1930s and the fast growth of the 1920s. This constitutes a counterintuitive result, as an association between staggering inequality and extreme poverty and the break-up of the Civil War has often been hinted at, though never proved, in the historical literature (cf. Pérez Ledesma, 1990, and Payne, 1993). During the early years of Francoism (1939–1953), rising inequality and poor economic performance brought the share of those below the poverty line back to late 1920 levels. Conversely, the late phase of Franco’s dictatorship appears as an epoch of falling inequality and increasing per capita income, factors that jointly eradicated absolute poverty by the early 1970s.

Fig. 5.11

Calibrated and counterfactual* poverty headcount (Poverty Line 1990 G-K $2.10 day/person). *Assuming a fixed 1913 Gini. Source: See text and Table 5.2

A glance at Fig. 5.5 might suggest, however, that given the similar level of inequality in the mid-nineteenth and in the late twentieth century, growth alone would explain the eradication of absolute poverty. Was this the case? I have carried out a counterfactual exercise in which I computed the poverty headcount under the assumption that inequality remained unchanged at a high level (that of 1913) over time. The results for the counterfactual and the calibrated poverty headcounts are shown in Fig. 5.11 (dotted line). It turns out that although economic growth was the main force behind the long-run fall in absolute poverty, during episodes of intense poverty decline, a significant contribution came from the rapid decline in inequality (such as the late 1920s-early 1930s, and from the late 1950s to the early 1970s).

The case of Spain presents interesting similarities to and differences from Latin America. Spain shadowed the evolution of Latin American poverty until the 1950s, when inequality levels in Spain departed from those prevailing in Latin America and initiated a fast convergence towards OECD patterns.³⁹ Thus, the growth of per capita income had a higher payoff in terms of absolute poverty suppression in Spain than in Latin America, where the poverty headcount still remained at the end of the twentieth century.⁴⁰

6 Conclusions

In Spain, inequality rose during the late nineteenth century and up to World War I, reversed during the interwar years, witnessed an upsurge in the post-Civil War autarchy, and fell between the mid-1950s and the early 1970s, stabilising thereafter. During the first 100 years considered, the gap between property and labour returns drove aggregate inequality. Then, from the mid-1950s onwards, as growth and structural change accelerated, skilled labour increased its share of employment and the dispersion of labour returns became the main determinant of personal income distribution.

The contrast between Spain and Latin America offers a parallel long-run evolution up to the mid-twentieth century, when Spain deviated to converge towards OECD levels. World and Civil Wars affected inequality—although they did not have permanent effects—, and progressive taxation only had an impact as of 1980.

In modern Spain, no trade-off between inequality and growth is observed. In its most dynamic phases, inequality declined—the 1920s, the Golden Age (1954–1973)—but also increased (1850–1883), while in years of sluggish performance, inequality deepened (1880s-1920, the post-Civil War autarchy) though it shrank too (during the II Republic, 1931–1936). Furthermore, economic growth and declining inequality had dramatically different outcomes during the world crisis of the 1930s and 1970s, with political and social strife leading to civil war in the former period, and democratic stability and social consensus in the latter.

Absolute poverty experienced a long run decline. Growth prevailed over falling inequality as the main cause of poverty reduction, but a more egalitarian income distribution played a significant part in crucial phases of absolute poverty decline. The contrast with Latin America reveals that thanks to a lower degree of initial inequality, Spanish economic growth in the third quarter of the twentieth century had a much larger payoff in terms of absolute poverty alleviation.

From this preliminary assessment of modern Spain’s experience, some hypotheses about the connections between growth, inequality, and social conflict emerge. Attempts to introduce institutional and social reforms during the II Republic (1931–1936) were accompanied by increasing social turmoil and political unrest that led to General Franco’s uprising and to the Civil War (1936–1939). Were there economic reasons for this conflict? Was there a war of attrition on income and wealth distribution at the roots of the Spanish Civil War (Boix 2004)? The fact that it broke out after one and a half decades of inequality decline and poverty alleviation demands new explanatory hypotheses. Unfulfilled hopes of sharing share increases in wealth on the part of those at the bottom of the distribution may have contributed to the social unrest that preceded the Civil War. Furthermore, the shrinking gap between returns to property and to labour in a context of social unrest, including threats to property, during the early 1930s, provides a potential explanation for the support lent by a not inconsiderable sector of Spanish society to the military coup d’état that triggered the Civil War.

The outcome of the Civil War, Franco’s long-lasting dictatorship (1939–1975), encompassed two distinctive phases: autarchy and sluggish growth, in the former; in the latter, cautious liberalization and fast economic progress. My estimates suggest that a dramatic increase in inequality, possibly a consequence of the Civil War, together with sluggish growth, resulted in striking poverty, with one out of four Spaniards below the poverty line by the early 1950s. A benevolent picture emerges, however, from the mid-1950s onwards, since, as income distribution became more egalitarian and growth accelerated, absolute poverty was practically suppressed by the early 1970s. Did the successful transition to democracy in the last quarter of the twentieth century have its roots in the late Francoism?

Appendix

1.1 A.1 Sources

Nominal GDP derives from Prados de la Escosura (2017, updated). Data available at https://frdelpino.es/investigacion/en/category/01_social-sciences/01_spanish-economy/02_historical-perspective-spanish-economy/

Nominal unskilled wage correspond to that for agriculture over 1850–1954. From 1954 onwards, the weighted average unskilled wage rate per hour (weights are the number of hours worked in each branch of economic activity).

Nominal average wage is the nominal weighted average wage rate per hour (weights are the number of hours worked in each of branch of economic activity).

Employment

Only the latest series of national accounts (CNE15, CNE10) provide full-time equivalent (FTE) workers and its distribution by industry from 1995 to 2020. However, the 1995-based quarterly national accounts (CNTR95) present data on FTE workers for 1980–1995. I have, then, spliced the two sets of FTE workers through linear interpolation to get consistent estimates over 1980–2020.⁴¹

For the pre-1980 years, García Perea and Gómez (1994) provide estimates of employment back to 1964 that can be pushed further back to 1954 with the rate of variation of employment provided in earlier national accounts (CNE64) (Instituto de Estudios Fiscales, 1969: 33–34). I have assumed that the number of FTE workers evolved alongside employment and projected its 1980 level backwards to 1954 with the employment rate of variation in order to derive FTE employment series for the period 1954–2020.

Linking the post-1954 series with the historical evidence back to 1850 represents a challenge. On the basis of population censuses I constructed yearly employment estimates for 1850–1954 for the four main sectors: agriculture, forestry, and fishing; industry, mining, and utilities; construction; and services. Major shortcomings appear in Spanish census data: working population is only available at benchmark years and refers to the economically active population [EAN, thereafter], with no regard of involuntary unemployment.⁴² Moreover, censuses tend to only record one activity per person, that which individuals consider being their principal activity (farmer usually). However, in a developing society the division of labour is low and a single person might undertake various work tasks over the course of a year.⁴³ Henceforth, activities corresponding to the industrial and, particularly, service sectors end up being underestimated in population censuses.⁴⁴ In addition, figures for female EAN in agriculture seem to be inconsistent over time.⁴⁵ Therefore, I have been forced to make some choices. For example, in order to derive consistent figures over time for EAN in agriculture, I excluded the census figures for female population, while assumed that female labour represented a stable proportion of male labour force in agriculture and, hence, increased the number of days assigned to each male worker (see below).⁴⁶ Moreover, as the share of EAN in agriculture is suspiciously stable over 1797–1910, in spite of industrialization and urbanization, I corrected it by assuming that the agricultural share of EAN moved along, and could not exceed, the proportion of rural population (living in towns with less than 5000 inhabitants) in total population.⁴⁷ Thus, I adjusted downwards the percentage of EAN employed in agriculture between 1887 and 1920 by redistributing ‘excess’ agricultural workers proportionally between industry, construction, and services.⁴⁸ The next step was to obtain yearly EAN figures through log-linear interpolation of benchmark observations. Since the resulting estimates do not capture yearly fluctuations in economically active population, a partial solution has been, firstly, to compute EAN share in working age population (WAN) and WAN share in total population (N), being WAN and N computed through linear interpolation (i) between population censuses.⁴⁹ Then, these ratios have been multiplied by the yearly population estimates (N) (Prados de la Escosura, 2017) to derive annual figures of economically active population (EAP). Thus,

$$ \mathrm{EAP}=\left({\mathrm{EAP}}^{\mathrm{i}}/{\mathrm{WAN}}^{\mathrm{i}}\right)\left({\mathrm{WAN}}^{\mathrm{i}}/{\mathrm{N}}^{\mathrm{i}}\right)\mathrm{N} $$

(5.8)

Later, in order to adjust for differences in labour intensity across main economic sectors and obtain a crude measure of full-time equivalent worker by industry, the data on EAP was converted into days worked per year. I assumed that each full-time worker was employed 270 days per annum in industry, construction, and services. Such figure results from deducting Sundays and religious holidays plus an allowance for illness. This assumption is in line with contemporary testimonies and supported by the available evidence.⁵⁰ In agriculture, however, contemporary and historians’ estimates point to a lower figure for the working days per occupied, as full employment among peasants only occurred during the summer and, consequently, workers were idle for up to 4 months every year. It can be assumed that the working load per year for the average male worker in agriculture would range, at most, between 210 and 240 days.⁵¹ However, in order to make for the exclusion of female employment in agriculture (due to the absence of consistent data), I increased the number of days assigned to male workers employed in agriculture to match the figure used for the rest of economic sectors (270).⁵²

Lastly, full-time equivalent employment figures by economic sector for 1850–1953 were derived by assuming that their yearly changes mirrored those in economically active population and, thus, FTE employment estimates for 1954 were backwards projected with those for economically active population (EAN). It is worth noting that, in 1954, the ratio between FTE employment and EAN for each economic sector is 1.003 (agriculture), 0.872 (industry), 1.095 (construction), and 1.069 (services), and 1.000, for the aggregate. The implication, in the case of agriculture, is that, the upper bound figure for male employment (resulting from an attempt to make for missing female labour figures) matches that of full-time equivalent total employment (including female work).

Wages:

The quality and availability of wage data necessary to construct these estimates vary enormously through time. Different periods can be distinguished:

1850–1908: Agricultural wages come from Bringas (2000). Wages in construction and services from Reher and Ballesteros (1993) were re-scaled to the national levels provided by Rosés and Sánchez-Alonso (2004). Wages for mining are from Chastagneret (2000) and Escudero (1998). Levels of manufacturing wages in all industry and services sectors at different dates (1850, 1880, 1905) were obtained, respectively, from Cerdá (1867), the U.S. Department of Labor (1900), and Anuario Estadístico de Barcelona (1905). Benchmark wage levels were interpolated with Fisher indices constructed with yearly data from Camps (1995), Llonch (2004), and Soler Becerro (1997) for consumer industries, and Escudero (1998) and Pérez Castroviejo (1992), for the rest.

1908–1920: the detailed wage enquires conducted by the Instituto de Reformas Sociales (various issues) with information by gender on minimum, maximum and average wages for 20 branches of industry (kindly supplied by Javier Silvestre) were used. Wages in agriculture and services were taken from Bringas (2000) and Reher and Ballesteros (1993), respectively.

1920–1954: Wage levels from a detailed survey for 1914, 1920, 1925 and 1930 (Ministerio de Trabajo, 1931) were interpolated with wage variation rates provided in Anuario Estadístico de España [AEE] (various issues) (only 9 occupations up to 1925, 15 thereafter) to derive nominal wage series, classified by industry, for the period 1920–1936. For 1936–1954, wage levels for 1930 and 1955 were projected back and forth with wages’ rates of variation taken from data in AEE and Vilar (2004) and the resulting series were combined as Fisher index to obtain yearly wage levels.

1954–1985: Unit labour costs by sectors of economic activity from Fundación BBV (1999).

1985–1995: Unit labour costs for 53 Industries come from National Accounts (CNE86).

1995–2021: Unit labour costs for 63 Industries come from National Accounts (CNE2010, CNE2015).

See Tables 5.1 and 5.2.

Table 5.1 Gini and its components, 1850–2020

Table 5.2 Poverty headcount, 1850–1975: calibrated and counterfactual

Fig. 5.12

Unadjusted Gini with fixed and variable proprietors income inequality

See Fig. 5.12.