The rise and fall of paper money in Yuan China, 1260–1368

Update (Ocotber 2023): This paper is now forthcoming at the Economic History Review

Hanhui Guan (Peking University), Meng Wu (Manchester) and I have a new working paper with the above title (available in open access here). We have been working on this for some time, and we are happy that we finally have the first version of our paper ready. In my view, the case of historical China is not only interesting in its own right but provides a helpful mirror to understanding later Western European monetary and financial history better, too.

The abstract for our new paper is here:

Following the Mongol invasion of China, the Yuan (1260–1368) dynasty was the first political regime in history able to deploy paper money as the sole legal tender. Drawing on a new dataset on money issues, imperial grants, and prices, we show that a silver standard initially consolidated the Chinese currency market. However, persistent fiscal pressures eventually compelled rulers to ease the monetary standard, and a fiat standard was adopted, leading to high inflation levels. We show that military pressure generated fiscal demands which led to over-issuance, and we reject the role of excessive imperial grants in triggering the over-issue of money.

Annual nominal money issues, 1260–1355

Kublai Khan, founder of the Yuan dynasty

1 guan note Zhongtong Yuanbao Jiaochao (中统交钞) issued by the Yuan

Interview to Atlantico

Here’s the transcript, in English, of a recent interview I gave to the French media Atlantico, concerning this paper (joint work with Jaime Reis and Lisbeth Rodrigues).

The interview concerns the causes of the Little and Great Divergences, with particular attention to the “Girl Power” hypothesis. (A recent VoxEU column is also available in this link.)

My French is not great so I responded in English. I copy below the responses that I sent them.

In your study “Historical gender discrimination does not explain comparative Western European development”, you mention the fact that the slow economic growth of south-western Europe since the Middle Ages is often attributed to the lesser influence of women compared to northern countries such as England or the Netherlands. Where do we find such statements? Why is this explanation so common?

Around 1900, Global Inequality between countries was at an all-time high. Most of Europe and its offshoots (such and the USA and Australia) had been growing systematically for a long time, while much of the rest of the world had not, including the largest regions by population like China and India.

In the prior centuries, from the late Middle Ages until the early Twentieth Century, the world had witnessed two important economic divergences. One was the Great Divergence: how the richer parts of Europe (e.g. England) became much richer than the richer parts of the rest of the world (e.g. the Yangtze Delta). And even the poorer parts of Europe became considerably richer than the poorer parts of the rest of the world.

The other was the Little Divergence: how the richer parts of Europe (Northwestern Europe) became much richer than the poorest parts of Europe (Southern and Eastern Europe).

There are different explanations in the literature for why these divergences happened. One explanation is cultural, and states that there is a particular “European” way of thinking and acting, and one manifestation of this was comparatively high female agency and an original European Marriage Pattern (EMP), which can be characterized by later female first marriage ages than other parts of the world, a relatively high celibacy rate, and marriages that were monogamous, exogamous, based on consensus and neo-locality. Regardless of the ultimate cause of these cultural behaviors — I personally believe that political institutions are jointly determined with and can shape culture over time and not just the opposite – it does seem undeniable that there was particular, individualist, comparatively liberal way of thinking and acting in Europe, coming from the Middle Ages already – an European culture, if you will.

This European culture was reflected in comparatively high female agency, by contrast with most other parts of the world. It was, in particular, Western European: the first to propose the EMP was a Hungarian scholar, John Hajnal, who in 1965 proposed what is now known as the Hajnal line: an imaginary line running between Saint Petersburg and Trieste, to the West of which the EMP operated. A good book to understand the deep origins of a Western European mentality and political culture is in my view Siedentop’s Inventing the individual: The Origins of Western Liberalism.

Concerning the Little Divergence, one well-known and increasingly popular explanation for it, associated in particular with Jan Luiten van Zanden and his multiple co-authors, is essentially that Southwestern/Mediterranean Europe was not really part of Western Europe in this EMP sense. According to this explanation, the lower female agency which characterized regions such as Spain, Portugal, or even south France and south Italy explain their comparative underdevelopment. Women, this literature claims, married early, had high fertility (with consequences for low human capital accumulation, of themselves and their children), did not participate in the labor market as much as elsewhere, and faced comparatively high gender wage gaps, which were supposedly determined by social norms, not market forces. A related literature also exists which splits Europe in religious terms, for instance Joseph Henrich argues that the EMP was stronger in Protestant regions of Europe.

What we’ve done in our present paper was show this was not, empirically, the case. Culturally, something we can call “Western Europe” did exist, with variations within, but these differences could not be first-order to explain the Little Divergence. Dennison and Ogilvie had already shown in fact that the parts of Europe where the EMP were historically stronger were not the most successful in a development sense.

But let me be clear. The evidence does not support the view that the Southern or Catholic part of Western Europe discriminated against women more, but it does largely support Hajnal’s original point that there is such a thing as a “Western European” culture reflected in high female agency and the EMP, among other aspects. In the long run, this may have been a key advantage of Western Europe relative to other parts of the world. In this sense my disagreement with Jan Luiten van Zanden – a scholar for whom I have the utmost respect – is only partial.

One of your conclusions is that economic development fosters the improvement of women rights and not the over way round. Is that the explanation for usual misreading of the situation?

The usual explanation – most associated with Jan Luiten van Zanden and his co-authors –  is that the high female agency was a cause of development: the regions of Europe that had it grew more. We are pointing out that, within Western Europe, there were not important historical differences in female agency. Consequently the differences in rights that emerged – mostly only visible only by the early Twentieth century and for the most part gone by the late 1970s – were more in fact a consequence of differential development.

You’ve compared the discrimination against woman in several countries, including, Portugal, England, Netherlands, etc. You’ve found no significant differences in the discrimination towards women. How do you measure that?

We rely on a new dataset of thousands of observations from archival sources covering six centuries, and we complement it with a qualitative discussion of comparative social norms. Compared with Northwestern Europe, women in Portugal faced similar gender wage gaps, married at similar ages, and did not face more restrictions to labor market participation.

What are the influences of the European Marriage Pattern (EMP) according to your findings? 

Please see above

You write that “an explanation for the growing income inequality between European countries during the early modern period, especially from the mid-seventeenth century onward – the ‘Little Divergence’ – must be found elsewhere” than gender discrimination. Do you have leads on what might be the explicative factor?

As I mentioned, by the late Middle Ages Western Europe was characterized by an individualist culture that was jointly determined with representative and even proto-democratic political institutions. Parliaments, the judiciary, and the independent power of the Church led to checks and balances to executive power in a way that was particularly Western European and absent from all other parts of the world. Scholars such as Acemoglu and Robinson believe that the Crowns of Spain and Portugal were absolutist already around 1500, but the evidence does not in fact support this claim. Instead, the evidence suggests that Iberia [later] suffered from a resource curse. Note that the second country to have an Industrial Revolution was [Catholic] Belgium, and France followed shortly afterwards. At the same time, Italy and Germany were able to industrialize quickly once they unified politically, in the second half of the nineteenth century.

Historical gender discrimination does not explain comparative Western European development

Historical gender discrimination does not explain comparative Western European development: This is what we argue in a new paper (joint work with Jaime Reis and Lisbeth Rodrigues). Also available as a CEPR discussion paper.

Here’s the abstract:

Gender discrimination has been pointed out as a determining factor behind the long-run divergence in incomes of Southern vis-à-vis Northwestern Europe. In this paper, we show that there is no evidence that women in Portugal were historically more discriminated against than those of other parts of Western Europe, including England and the Netherlands. We rely on a new dataset of thousands of observations from archival sources which cover six centuries, and we complement it with a qualitative discussion of comparative social norms. Compared with Northwestern Europe, women in Portugal faced similar gender wage gaps, married at similar ages, and did not face more restrictions to labor market participation. Consequently, other factors must be responsible for the Little Divergence of Western European incomes.

Comparative gender wage gap (f/m) for unskilled casual workers

Monetary Capacity: some background

New voxeu column and accompanying CEPR discussion paper, covering my work in co-authorship with Roberto Bonfatti, Adam Brzezinski, and K. Kivanç Karaman.

The paper touches a number of themes across economic history, historical political economy, and macro/monetary economics; but here’s a 1-sentence summary: we argue that monetary and fiscal capacity and, by extension, markets and states have a symbiotic relationship. And we provide causal evidence, too. If this peaked your curiosity, please have a look at the column and the paper itself.

In this short post, I’ll give some background about this work which makes more sense to be in a blog post than elsewhere.

We have been working in this paper for a long time – more than three years – and it feels great that we finally have a working paper. It’s worth pointing out our new paper is related to prior work by Kivanç Karaman and co-authors, as well as my own such as this piece, briefly summarized here.

For a long time, I’ve had the greatest admiration for Kivanç Karaman and the exciting work he has done, some of which with the equally impressive Sevket Pamuk, including the work they did in starting off a literature (which has since grown greatly) on empirical measures of historical state capacity. If you asked me who’s my favorite young economist in the world – and if I really can only pick one – then I’d have to answer it’s Kivanç. Whatever you do, do yourself a favor and read his work (you can thank me later!).

Our current work is in some sense the intersection of his research program with mine, and it has been wonderful working with him as well as my long-time friends and collaborators Adam Brzezinski and Roberto Bonfatti, who both also provided critical input and without whom the current paper would also not have been possible.

Ridolfi on premodern France (Hightlight I)

As announced in the previous post, there will be from now on once in a while posts written by guest scholars, both junior a senior. This post has been written by Leonardo Ridolfi of the IMT School for Advanced Studies, Lucca. You can find Leonardo’s most recent working paper here.

The French economy in the longue durée. A study on real wages, working days and economic performance from Louis IX to the Revolution (1250-1789)

This work addresses a gap in the literature concerning living standards in pre-industrial France.

While traditionally research had an eminently localized character, focusing on the experience of specific regions or what might be called “local economics,” still to date, there is no consolidated understanding of the long-term development of wages and prices from a broader national perspective.

Building and improving upon the precious contributions offered by the many compilers of wage and price data in France, this study is an attempt to provide a solid empirical characterization of the principal macro-economic aggregates of pre-industrial France and trace the main contours of economic growth in the country from the phase of early state formation to the Revolution.

Delving into the vast set of secondary and printed primary sources, the first section presents new series of real wages for male agricultural and construction workers in France from 1250 to 1789 (now updated to 1860) following Allen (2001)’s barebones basket methodology.

The analysis highlighted three main issues.

First, our series offer little support to the argument that there were appreciable long run improvements in living standards for French wage earners before the Industrial Revolution. Indeed, real wages displayed no substantial trend improvement between the thirteenth and the mid-nineteenth century.

Second, the estimates reveal that the period 1350-1550 saw the rise and consolidation of a real wage gap between France and England as well as other leading European cities. Still in the decade prior to the Black Death the real wage differential between French and English workers of the construction sector was remarkably low. A century later, in the 1450s, French building labourers had between about 25 and 40 percent less of the income of their European counterparts.

Comparing real wages of French farmers to those of their English counterparts I found a similar pattern and few traces of a French “golden age” of labour. Indeed, after a first phase of rapid expansion following the Black Death, by the1370s real wages grew less and for a shorter period than elsewhere in Europe where the welfare gains consolidated almost until up the 1450s. At a more disaggregated level, similar trends are discernible by comparing Paris to London.

As a first step, I decomposed the proximate causes of this gap between prices and wages. I found that France and England witnessed similar deflationary trends between the 1370s and the 1450s. Yet, it was the decline of French silver wages (apparently driven by falling production and reduced labour demand especially during the worst phases of the Hundred Years War) and the contemporaneous increase of English salaries, that explain the “dampened” Malthusian cycle of real wages in France as opposed to the “full” Malthusian cycle experienced by England and Central-Northern Italy.

Figure 1: Real wages

Notes and Sources:  French labourers: this study (updated version of the thesis). England: Clark (2005).

Finally, even if demographic data before the 1550s are fragmentary, it is possible to argue, consistently with the Malthusian interpretation, that the dynamics between real wages and population was characterized by a long-lasting inverse relationship. Nevertheless, while this mechanism appears to hold in general, at least by the mid-seventeenth century one can detect a weakening of the inverse relationship. Indeed, the long phase of demographic expansion that brought population almost to triple between the 1600s and the mid-nineteenth century, was paralleled by a mild decrease or a substantial stagnation of real wages.

The second section provides a broad characterization of working time in pre-industrial Europe concentrating on three dimensions of time: the calendar working year corresponding to the calendar year net of general holidays and religious festivities; the actual working year and the implied working year defined as the annual number of days of work required by a male breadwinner to provide for a notional family of five components (Allen and Weisdorf 2011).

Due to the dearth of compelling evidence on work intensity for workers employed in agriculture, I looked at the experience of construction workers on site providing new estimates of trends in calendar, actual and implied working year in France and England from the fourteenth to the eighteenth century.

By analyzing the joint evolution of these three dimensions of time and comparing the patterns of change of time-use, and their response to variations in the institutional and market conditions, I identified two distinct regimes of industriousness featuring France and England in the pre-industrial era.

In France, the annual number of days required by a male breadwinner to provide for his family (the implied working year) was greater than the actual number of days worked per year, meaning that women and children’s labour force participation as well as the presence of additional sources of non-labor income were necessary to assure the basic levels of consumption. This implies that expansions in the offer of labour were primarily driven by raising inflation and economic hardship (Figure 2).

Figure 2: The French case

Sources: Calendar, actual and implied working year: this study.

Notes: Surplus (deficit) labour input: The positive (negative) difference between actual and implied working year (shaded area).

By contrast, I found evidence of the existence of two phases where English regular construction workers supplied more days of work to the market than required by basic household subsistence (Figure 3).

The first episode occurred between 1400 and 1500, while the second corresponds to the industrious revolution originally described by De Vries (2008).

Several hypotheses are discussed to shed light on the origin of these phases of surplus labour input and their implications on the structure of consumption and production. These episodes differed in two fundamental ways.

First, they originated from different dynamics.

Indeed, the episode of surplus labour input located by De Vries in the seventeenth century England and the Low Countries, derived from an upsurge in actual workloads and a contemporary drop of work requirements necessary for family subsistence in a context of progressive expansion of the frontier of working possibilities.

On the contrary, the episode of surplus labour input detected in the post-plague period was characterized by the contemporary reduction of actual, calendar and implied working year.

Received wisdom would suggest that workers should have totally (or in large part) compensated the post-plague increases in real wage rates by reducing labour supply of approximately the same amount consuming a considerable proportion of their augmented purchasing power in the form of leisure (Blanchard 1994). However, actual workloads decreased much less than implied by the contemporary increase in real wage rates. This incomplete adjustment, that reflected a rather inelastic labour supply of construction workers, could depend on two main factors.

First, the existence of technical requirements and institutional settings, including the rhythm of the construction process, the rests dictated by calendar working year as well as the recruiting schemes of contractors and the organizational forms of entrepreneurs, limited voluntary reductions of actual workloads.

Second, the incomplete response of actual workloads could reflect the rise of a new attitude toward higher quality consumption from an increasing share of workers (seemingly skilled and urban) that was “aping the lesser gentry” (Dyer 1988).

In this respect, these episodes had different implications for the relationship between labour offer, consumption and production.

Indeed, the phase of surplus labour input in the seventeenth century England was seemingly related to a consumer revolution (Allen and Weisdorf 2011) and could be thought of as a transition from traditional consumption cluster to a broader and more modern one that included colonial products and luxuries (De Vries 2008).

The episode of surplus labour input in the late medieval England was not marked by more and new items entering the basket but seemingly ran in parallel with a relocation of consumption choices within the horizon of traditional consumption that reflected structural changes in the economy after the Black Death and the aspiration of a growing share of population for higher alimentary standards less dependent upon cereal-based and lower quality foodstuff (Dyer 1988).

From the production side, while the seventeenth century phase of surplus labour input saw the rise and consolidation of new sectors outside agriculture, the first episode (seemingly did not cause but) coincided in time with a shift of agriculture from arable to pasture. This process is consistent with a large body of empirical evidence documenting changes in alimentary regimes during the fourteenth and fifteenth centuries.

Figure 3: The English case

Sources: Calendar year: this study. Implied working year: Allen and Weisdorf (2011). Actual working year: Period 1300-1559: this study. Between 1560 and 1732, Clark and Van DerWerf (1998) and by 1750 Voth (2001) as reported in Table 2 of Allen and Weisdorf (2011).

Notes: Surplus (deficit) labour input: The positive (negative) difference between actual and implied working year (shaded area).

Finally, in the last section I present new estimates of agricultural and total output per capita in France between 1280 and 1789 using the demand side approach. The study suggests that GDP per capita displayed no substantial trend improvement over this period. At the death of King Philip the Fair in 1314, France was a leading economy in Europe and output per capita averaged 900 dollars per year. Almost five centuries later, at the beginning of the 18th century, this threshold was largely unchanged and GDP per capita was slightly above 1000 dollars, about half of the level registered in England and the Low Countries (Figure 4).

These estimates document quantitatively and in the aggregate what was previously known only qualitatively or for some regions by the classic works of the French historiography (Goubert 1960; Le Roy Ladurie 1966) thus offering support to Le Roy Ladurie (1977)’s characterization of the pre-industrial French economy as a stagnating, growthless system.

Nevertheless, GDP per capita was highly volatile and experienced multiple peaks and troughs alternating phases of economic crisis to periods of economic expansion. These include the “efflorescence” of economic growth that took place between the 1280s and the 1370s and the growth trend since the mid-16^th century that ran in parallel with the consolidation of the French state and the opening of new trade routes from Europe to Asia and the Americas.

Overall, our estimates suggest that the evolution of GDP per capita in France can be suitably interpreted as an intermediate case between the successful example of England and the Low Countries and the declining patterns of Central-Northern Italy and Spain. Being neither a southern country nor a northern one, the growth experience of France seems to reflect this geographic heterogeneity.

Figure 4: GDP per capita in Europe

Sources: England: Broadberry et al. (2011); France: this study; Holland: van Zanden and van Leeuwen (2012); Italy: Malanima (2011); Portugal: Palma and Reis (2016); Spain: Álvarez-Nogal and Prados de la Escosura (2013); Sweden: Schön and Krantz (2012).

References

Allen, Robert C. “The great divergence in European wages and prices from the Middle Ages to the First World War.” Explorations in economic history 38, no. 4 (2001): 411-447.

Allen, Robert C., and Jacob Louis Weisdorf. “Was there an ‘industrious revolution’ before the industrial revolution? An empirical exercise for England, c. 1300-1830.” The Economic History Review 64, no. 3 (2011): 715-729.

Álvarez‐Nogal, Carlos, and Leandro Prados de la Escosura. “The rise and fall of Spain (1270-1850).” The Economic History Review 66, no. 1 (2013): 1-37.

Blanchard, Ian. Labour and Leisure in Historical Perspective, Thirteenth to Twentieth Centuries: Papers Presented at Session B-3a of the Eleventh International Economic History Congress, Milan, 12th-17th September, 1994. No. 116. F. Steiner, 1994.

Broadberry, Stephen et al. “British Economic Growth, 1270-1870: An Output-Based Approach”, London School of Economics, 2011. http://www2.lse.ac.uk/economicHistory/whosWho/profiles/sbroadberry.aspx.

Clark, Gregory, and Ysbrand Van Der Werf. “Work in progress? The industrious revolution.” The Journal of Economic History 58, no. 3 (1998): 830-843.

Clark, Gregory. “The condition of the working class in England, 1209–2004.” Journal of Political Economy 113, no. 6 (2005): 1307-1340.

De Vries, Jan. The industrious revolution: consumer behavior and the household economy, 1650 to the present. Cambridge: Cambridge University Press, 2008.

Dyer, Christopher. “Changes in diet in the late middle ages: the case of harvest workers.” The Agricultural History Review (1988): 21-37.

Goubert, Pierre. Beauvais et le Beauvaisis de 1600 à 1730: contribution à l’histoire sociale de la France du XVIIe siècle: atlas (cartes et graphiques). Paris: SEVPEN, 1960.

Le Roy Ladurie, Emmanuel. Les paysans de Languedoc. 2 vols. Paris: SEVPEN, 1966.

Le Roy Ladurie, Emmanuel. “Motionless history.” Social Science History 1, no. 2 (1977): 115-136.

Malanima, Paolo. “The long decline of a leading economy: GDP in central and northern Italy, 1300-1913.” European Review of Economic History 15, no. 2 (2011): 169-219.

Palma, Nuno and Reis, Jaime. “From Convergence to Divergence: Portuguese Demography and Economic Growth, 1500-1850” (September 13, 2016). Available at SSRN: https://ssrn.com/abstract=2839971 or http://dx.doi.org/10.2139/ssrn.2839971

Schön, Lennart, and Olle Krantz. “The Swedish economy in the early modern period: constructing historical national accounts.” European Review of Economic History 16, no. 4 (2012): 529-549.

Van Zanden, Jan Luiten, and Bas Van Leeuwen. “Persistent but not consistent: The growth of national income in Holland 1347-1807.” Explorations in economic history 49, no. 2 (2012): 119-130.

Voth, Hans-Joachim. “The longest years: new estimates of labor input in England, 1760-1830.” The Journal of Economic History 61, no. 4 (2001): 1065-1082.

 

 

Introducing “highlights”: Ridolfi on premodern France and Jongman on the Roman empire

From now on there will be once in a while posts written by others in this blog. These will be written both by young scholars and by more senior, established scholars.

The idea is that these scholars will write short essays about the main conclusions (and possibly policy implications) from their overall work. Scholars will write about their work overall: the forest, not the trees. Speculation about future work and general considerations about the state of the field are welcome. Hence the logic is different (and a complement, not a substitute) from that of blogs such as EHES’s Positive Check or EHS’s The Long Run, where people write about one specific paper at the time.

Consistent with this policy, the two inaugural posts will be written by:

• Leonardo Ridolfi, who recently wrote an excellent PhD thesis which builds estimates for French real wages and GDP over the long run. This is an exciting development: those of you who have read this blog’s very first post will know that we previously didn’t have estimates for premodern France. This absence has been largely a sad heritage of the French academia’s aversion to quantitative history since the third generation Annales School. It is significant that it was an Italian, working at an Italian institution, that has finally done this. And Ridolfi has done an excellent job, going back to 1250.
• Willem M. Jongman, a well-known scholar of the economic history of antiquity. I first realized how outstanding his work was when I first heard him speak at the LSE, and read this chapter in the Cambridge History of Capitalism, Re-constructing the Roman economy. He will write about what have we learned in the last 15 years, and future projects.