Van Zanden's surprise

Van Zanden certainly does not denounce Adam Smith. His primary yardstick for judging institutions is their efficiency in reducing transaction costs and enhancing market exchange (p.294). Which market economist could cavil at that? He finds the appropriate institutions appearing, or re-appearing, in the High Middle Ages (ca. 1000–1250) in Western Europe. They include the rule of law, the notion of citizenship, and the corporate body (such as the guild). The first two are clearly ‘inclusive’ institutions, in Acemoglu and Robinson's sense, and the third could be. But he gives similar importance to a much less obvious facilitator of market exchange: the European Marriage Pattern, which appeared in the Low Countries and England in the 14th century and in time – very slowly – spread as far as the eastern frontiers of medieval Catholicism. This we may call van Zanden's surprise: a profoundly inclusive institution quite unknown to the other authors, which raised both the status and the economic contribution of the lowest of the low. We shall examine the European Marriage Pattern in the next section.

The medieval period

North et al. (Reference North, Wallis and Weingast2009) give the most detailed account of how institutions change between their four stages of social organisation, of which the first three are types of ‘natural state’. ‘A natural state manages the problem of violence by forming a dominant coalition that limits access to valuable resources – land, labour and capital – or access to and control of valuable activities – such as trade, worship and education – to elite groups’. (p.30). In ascending order, there are ‘fragile’, ‘basic’ and ‘mature’ natural states. They insist that ‘no teleology pushes states through the progression from fragile to basic to mature natural states’ (p.73) – or further.

Indeed; but can we not get some clues as to what may drive states in one direction or another? ‘In mature natural states, credible institutions evolve that provide organisations a measure of rule of law. . . .sustaining [this] for elite organisations appears to be incredibly complicated to pull off and is the beginning of the doorstep conditions [which make ascent to the summit stage, ‘open access orders’, possible].’ (p.73) Agreed (by Acemoglu and Robinson, van Zanden and the present writer). Rule of law is a key expression of impersonality. ‘The doorstep conditions represent institutional and organizational support for increased impersonal exchange. . . .’:

What then may favour the increasing ascendancy of the rule of law? Hicks in Theory of Economic History (1969) has an answer – the growth of trade and with it of more-or-less mercantile city-states, which the Greek poleis like Athens (at least) famously were – even if still dominated by a landowning aristocracy. Rulers above the law are the bane of merchants. Van Zanden does not appear to know Hicks, but agrees: he gives a multitude of examples from the High Middle Ages – the city-states of Northern Italy, like Venice and Florence; the Hanseatic League of maritime city states in the North; and of course his own Low Countries. The city-states innovate in the rule of law, some quite inclusively, even democratically, as in 6th century BC Athens, and van Zanden's succession of medieval insurgencies from Italy and the Low Countries. They prosper in consequence. But what mercantile city states never learn is how to expand their form of organisation across wider and wider territories; and so ‘the Greeks become subjects. . .of Macedonian monarchs; the Italian republics fall prey to France and Spain’ (Hicks, Reference Hicks1969 p.61). Then, in the happier cases, the ‘military monarchs’ who take over, ‘may wish to retain some part of the mercantile institutions’ (p.61).

We must focus on the mercantile innovators of medieval Europe and the institutional changes in which they led and the monarchs partially followed. They had two striking advantages. First, they had the remembered, or rediscovered, legal legacy of Rome – that one city-state (scarcely mercantile) which had succeeded in greatly expanding its territory, before falling under monarchical rule – of its own Caesars. And they had the Church. Acemoglu and Robinson virtually ignore the Church at any period. North et al. (Reference North, Wallis and Weingast2009) look at the medieval Catholic church rather closely, between pp.62–69. They say nothing about its doctrines, but a great deal about its power structures and their interaction with those of the secular nobility. They are as usual well-informed and accurate – and yet mystifying in what they do not say about the relationship of church and state. They start with the unarguable: ‘Under the original institutions by which the Catholic church became the state church of the Roman Empire, there was no confusion. . .the emperor was the source of law and above the law’. (pp.63–4).

Indeed: and this is the situation in almost every natural state in recorded history: religion is completely subordinate to or integrated in the secular power. It remained the situation in the Byzantine Empire, the continuation of the Roman Empire in the East (van Zanden, Reference van Zanden2009 p.66). But North et al.'s account makes clear (as does van Zanden's chapter 2), that it was absolutely not the situation in medieval Catholic Europe – less and less, after the ‘papal revolution’ from 1070–1122 which increased the cohesion of the Church. Canon law – church law – was independent of any secular power, and woe betide the monarch who forgot that – as the story of Henry II of England and his chancellor-turned-Archbishop, Thomas Becket, reminds us (Poole, Reference Poole1951). And yet they miss the point of this by suggesting that this extraordinary situation was within the normal range of variation of natural states: ‘The close interrelationship between the church and the state in medieval Europe from Charlemagne through the 16th century illustrates the futility of drawing hard and fast lines around governments in natural states and calling them the state. Power is dispersed in natural states’ (pp.72–3).

Dispersed? Of course, a natural state cannot be (or remain) a complete autocracy: there must be some kind of sharing of power with members of the elite or elites below the ruler. In a large natural state, there must be a large measure of spatial dispersion – decentralisation to provincial governors or local nobility or some such. But in no other case has there been ‘dispersion’ remotely like that of medieval Catholic Europe – with the primary, secular power territorially fragmented, and curbed and challenged by a multinational organisation operating over a million square kilometres; that organisation holding a virtual monopoly of education and thus having everything to gain from an ever-expanding corpus of written laws. Van Zanden understands the dynamic significance of this unique situation: ‘From 1072 to 1150 a systematic set of laws was created, regulating not only the governance of the Church, but also large parts of social and economic (and political) life. In response to these claims by the Church for a decisive influence in secular affairs, the emerging states of Western Europe began to create their own legal systems’. (pp.46–7). The result was to elevate the status of law: ‘Be you never so mighty, the law is above you’.Footnote ⁴

The medieval Church appears also to have played a key role – the key role – in the genesis and spread of van Zanden's surprise, the European Marriage Pattern. It is a cliché of demographers that birth rates fall as economic development proceeds – the ‘demographic transition’ – and that this is accompanied by a rise in the age of first marriage. We learn however that a sort of demographic transition with far-reaching consequences was under way in England and the Netherlands by 1400. The European Marriage Pattern, which first appeared in these two countries, before spreading ultimately to the rest of then-Catholic Europe, had three key, related features:

1. (1) The freedom of both sexes to make their own choice of partner: both sexes – a girl was free to make her own choice of husband.

2. (2) On marriage, the new couple set up their own household, separate from both sets of parents.

3. (3) In order properly to launch this establishment, they would leave home, usually in their early teens, and work for some other household, accumulating savings and competence.Footnote ⁵ The time this took, meant delay in marriage.

It should be noted that all this applied first and most completely to the poor.

Why then? Why there? The trigger for the appearance of the European Marriage Pattern may have been the transformation of the labour market in the Black Death years in the mid-14th century: pay rates rose very sharply, particularly for unskilled workers, and prospective employers were disinclined to worry about the sex or origins of their workers. Yet the Black Death affected almost the whole of Eurasia. The medieval Catholic church had come out strongly in favour of the rights of the young couple in the 12th century: but this applied to all its territories. Van Zanden finds certain variations in tradition on the rights of women within medieval Catholic Europe – with rather better treatment in the north. But the great innovation came in just two northern countries. The Netherlands, as van Zanden points out in the quotation above, had already moved out of feudalism – and that may have weakened patriarchy; it certainly made sure that there was a labour market on which the female poor could take advantage of scarcity. That leaves England.

We may suppose that there was something about English law or culture which made it less resistant to the new ‘girl-friendly’ teaching of the Church. Indeed in England it was not entirely new. Leyser (Reference Leyser1995) finds Theodore, the Archbishop of Canterbury, insisting on the absolute freedom of daughters to reject suitors chosen for them in the early 8th century. Theodore was Greek, not English; but in England he seems to have been pushing at an open door. In Anglo-Saxon England, women could hold land in their own right. That, they lost with the Norman Conquest of 1066 (Leyser). They could also reign in their own right: Aethelfled was sovereign Queen of Mercia around 900 – and a very effective ruler. One might have thought that here too, French practice would prevail after the Conquest: and in France never has a woman reigned in her own right. But already in the early 12th century Henry 1^st – whose wife was descended from Anglo-Saxon kings – was succeeded, albeit briefly, by his daughter Matilda. In the 16th century, of course, as in every subsequent century, women reigned in England in their own right.

Whatever the causes of the unusual nature of marriage in the ‘North Sea region’, the consequences, as van Zanden shows, were profound and long-lasting – pithily expressed by the title of his chapter 4 (written with Tine de Moor): ‘Girlpower’. Young men and young women invested in their own human capital and indeed their own financial capital, from the earnings of their employment. As real wages fell back, with the recovery of the population after the plague, the response was to postpone marriage further, in order nonetheless to accumulate enough, while searching for the right position, or farm tenancy. Those who never reached their goal might never marry – and the proportion never married rose to exceptional levels. This slowed down population growth, particularly in England, where average female age at first marriage rose to 26 by 1600 (Laurence, Reference Laurence1994), and thus helped to hold up wages.

An exceptionally large proportion of the labour force worked for wages – of the order of 50% in Holland and eastern England, in the early 16th century. (As van Zanden points out, this contrasts with a mere 1–2% in Ming China at the same period: an economy of similar level of development at that time, but much poorer performance subsequently). Not only was there a large and fluid labour market, but there was an excellent supply of skills, reflected by an unusually low skill premium (since the poor were both investing in their skills and restraining their fertility) and very high rates of literacy, particularly in Holland – with an exceptionally small difference between the sexes.

Financial markets were also very well developed, with a remarkably low rate of interest in both countries, the lowest anywhere in pre-modern times – falling to 5% in Holland in the mid-16th century. This must owe something, probably much, to the European Marriage Pattern. Young people were saving, and needed to invest. And the loosening of the bonds of the family meant that parents could count less than before, and elsewhere, on the care of their children in old age: so the investing continued, or resumed, after marriage.

The 16th century onwards: progress and regress

With the assistance of the Church, by the early 16th century Western Europe had moved a long way towards mature natural states, in the terms of North et al. (Reference North, Wallis and Weingast2009) – and none further than the Low Countries, which for some time had been under the rather light hand of the Dukes of Burgundy. But it is their backward English neighbours which Acemoglu and Robinson choose for close examination. We have already noted their preamble on the decline of the feudal order in England; but they make little of the supposed advantage this had given:

The significance of 1588 as a starting date is that in that year a Spanish fleet (the ‘Armada’) came close to landing a strong army in England which might well have overthrown Elizabeth and changed the course of economic and political development in England. However: ‘Against all odds, the English destroyed much of the fleet of their more powerful opponents. . . . Without this unlikely victory for the English, the events that would create the transformative critical juncture and spawn the distinctively pluralistic political institutions of post-1688 England would never have got moving’. (p.111). This is one of what Diamond (Reference Diamond2012) in his otherwise sympathetic review calls Acemoglu and Robinson's ‘Just So stories’Footnote ⁶ – brief explanations which read like blatant oversimplifications giving credit often to chance factors. Nonetheless, their story about the consequences of 1588 is highly plausible – except that it ignores the Dutch, who by then had made far more progress towards pluralistic political institutions than the English. Their history offers an even better story from fourteen years earlier – the Relief of Leiden:

1574 was a year of crisis. By dynastic inheritance the progressive Low Countries had fallen under the rule of Philip of Spain in 1555. He cared little for their traditional liberties and much about the Protestantism spreading across them, which he had called in the Inquisition and the Jesuits to suppress. In defence of religious tolerance and of their liberties in general, the Low Countries had risen in revolt under the reluctant leadership of William of Orange, the greatest noble in the land and the erstwhile favourite of Philip's father. Now the revolt looked like failing: the South had fallen, and the Spanish army was taking town after town in the rebel heartland, Holland. In the spring of 1574 the Spaniards laid siege to Leiden.

Leiden must not fall, and William and his officers were agreed, there was only one chance to save it: open the dykes – destroy the crops and homes of much of Holland, but make a water way through to Leiden. But he had no legal right to give the order: it must be voted by the Estates of Holland. The revolt was in defence of law: he would not break the law to save it. On July 30th, after seven weeks of argument, the Estates finally agreed. The dykes were opened. But to get the water high enough to float the relievers’ barges through, and flood the Spanish camp, the right wind was needed: from the sea. On the first of October, with Leiden within days of starved surrender, a steady north-west wind finally coincided with a high tide. Two days later, a Sunday, the barges reached Leiden's gates, and the joybells rang across Holland for Leiden's relief (Wedgwood, Reference Wedgwood1944).

The revolt of the United Provinces of the Netherlands was never so close to extinction again (with some help from England and its queen), and they were soon leading the way towards modern political and economic institutions. How critical a juncture the siege of Leiden was, we can better understand from van Zanden, who shows the particularity of the ‘North Sea region’ (mostly, England and the Netherlands) in the two, crucial, centuries from 1500–1700. During this period, most of the great medieval innovators – the German and Italian cities – fell victim, Hicks-style, to the armies of their feudal inferiors. The Reformation and consequent Counter-Reformation turned what remained of the Catholic church into a largely reactionary institution, politically; and even the Protestant churches were regressive in some respects, particularly in Germany, where they were most dependent on princely power. The Protestant kings and princes mostly followed the lead of their Catholic rivals in the move towards absolutism during the 17th century; England and the Netherlands stood out against it.Footnote ⁷

Strikingly, the same tendency, with almost the same exceptions, prevailed at the bottom of the social pyramid. As princes, so fathers: ‘Protestants as well as Catholics strengthened clerical, paternal and state control of marriage. . . .[only] England remained outside these reforms. . . Montesquieu [18th century] mentioned that English daughters frequently married. . . .without consulting their parents, because they were allowed to do so by law. . .’ (van Zanden, Reference van Zanden2009: 137). So Archbishop Theodore's rule, improved, held for English girls, after a thousand years: and it held up women's age at first marriage, which held down their fertility. This no doubt contributed to the rise of unskilled workers’ wages in London to the highest level in Europe, overtaking Amsterdam, in the early 18th century (van Zanden, Reference van Zanden2009: 97).

The economic consequences seem clear. Van Zanden's well-accepted findings on Europe's GDP per head from 1500–1800 are striking: outside Britain and the Netherlands, it stagnated, at best – falling in Italy. The movement of the European average for real wages was worse: in 1800 it was only 60% of its level in 1500, which itself was well below its post-Plague peak (p.305).Footnote ⁸ Yet it was the exceptionally high level of British wages for unskilled labour which played a key role in the Industrial Revolution by making it profitable to substitute capital for labour (Allen, Reference Allen2011).

In sum: this was a continent which had lost its medieval dynamism, and become more unequal and less ‘inclusive’, except for the Dutch and the English: had they succumbed to the power of Spain and the Counter-Reformation, what then? That wind from the sea may have saved a lot more than Leiden.

Prime movers

From whom have we learnt most about the prime movers of institutional change? From van Zanden, because he looks when and where the institutions of modern Europe largely originated – in the mercantile city-states and surrounding rural areas of northern Italy and the Rhine valley after 1000 AD; and at the right social levels. ‘At four levels – the micro level of the household, the meso level of communes, guilds and universities, the macro level of the state, and the meta level of the law – relatively democratic institutions originating in the High Middle Ages were at least partially the result of bottom-up processes of institution building’. But how did those medieval cities get the opportunity for their extraordinary creativity? From the meta-institution par excellence – the medieval Catholic Church. In section 3, we looked at its contributions to the micro level (the European Marriage Pattern) and the meta level of the law. The Church also made an indispensable contribution to the keeping of order in what would otherwise have been a hopelessly-anarchic political patchwork of 200 states (van Zanden's chapter 2). Further, ‘a high level of human capital formation and a strong civil society are required for revolutions to be successful against an established regime and to develop more democratic institutions’. (p.300). For that, first and foremost literacy was needed. And which institution can be thanked for the fact that ‘from 900–1300 a society emerged which believed in the power of the written word, and preferred the written text as evidence in court. . .’? (p.90).

The medieval Catholic church collapsed during the 16th century – partly because its followers insisted on reading its own written words in their own language. Its replacement by Catholic and Protestant churches largely under secular control, removed the key dynamic meta-institution and made further progress dangerously dependent on the fate of two weak states. This allows Acemoglu and Robinson, shaky as their history isFootnote ¹¹ , to make an important contribution with their ‘critical junctures’ – whether we emphasise the Armada or Leiden.

Interfering effectively

How can policy-makers in rich ‘inclusive’ societies help bring about progressive institutional change in poor ‘extractive’ ones? At least some of them are condemned to try – now that the expansion of the European Union has taken in a number of countries with extractive institutions, and other such countries are in acute crisis at the EU's gatesFootnote ¹² .

For Acemoglu and Robinson, as for the other authors, sustained economic growth requires inclusive institutions – today as much as ever. They use the prevalence of extractive institutions to explain the poor performance of many countries, and to express pessimism about the prospects of others, such as China. But in doing so they beg almost as many questions as they answer, because they do not make it entirely clear what institutions they are referring to, and they (clearly) exclude the least visible but potentially powerful institutions which we may call culture. They touch on but never really follow through the implications of the ‘resource curse’ best known from the case of oil: where wealth arises from natural resources, and particularly where it is concentrated to start with, it is easily exploited by extractive institutions and helps to perpetuate them. They might have pointed out that extractive institutions in Russia, with its enormous resource wealth/curse of oil, gas, and other minerals, are inherently more secure (other things being equal) than in mainland China, where income is mostly generated by manufacturing and services which are not concentrated and which depend critically on the initiative and enterprise of those in charge of and involved in producing them. Just such economic conditions underlay the successful transition to inclusive institutions of South Korea and Taiwan.Footnote ¹³

Acemoglu and Robinson are critical of most actions by ‘Western’ countries to export their inclusive institutions to less-developed (thus ‘extractive’) countries, whether by force (e.g. Iraq, Afghanistan) or through pressure (e.g. conditional aid), essentially on the grounds that extractive institutions will sustain or recreate themselves. North et al. (Reference North, Wallis and Weingast2009) are critical of such interventions too, and given their more careful specification of what institutions are, their analysis of what goes wrong is more revealing:

This has no chance of success (North et al., Reference North, Wallis and Weingast2009 contend) unless the society is close to being an open access order already – that is, it meets the three ‘doorstep conditions’, which few less-developed countries do. It may well make matters worse by unleashing disorder. This applies to the ‘transplanting’ of political as well as economic institutions – elections being a case in point. ‘For democracy to work, elections must be embedded in an institutional and competitive environment that allows political competition to convey information to and constrain politicians. Elections in natural states typically do not provide these functions or do so incompletely’. (p.267). What they might have recalled here is that their study of transitions to open access orders in Britain and the United States shows a rather gradual process of widening of access to power. This included the franchise: as representative government developed in these countries, so the right to vote, initially restricted, was steadily extended. Yet nobody presses the limited franchise on developing countries now.Footnote ¹⁴

Nor does anybody press protection on developing countries: on the contrary, tariffs and above all controls over inward investment by multinational firms are strongly discouraged. As Chang points out in Kicking Away The Ladder (2002), this flies in the face of the history of (above all) the United States and Japan, as well as Britain. Protection has political as well as economic implications – what better support for the development of inclusive political institutions can there be than the building up of competitive domestic businesses? The riposte may come, protection may build up uncompetitive domestic businesses dependent on political favour, as did Import-Substituting Industrialisation in Latin America (Allen, Reference Allen2011). But it did not do so in Japan, because as Allen points out, Japan modified advanced Western technology to suit its own factor endowment. Other countries could be helped to do likewise.

Concluding appreciation

In conclusion: Violence and Social Orders is worth reading on the grounds that it even half justifies its sub-title: A conceptual framework for interpreting recorded human history. North et al. (Reference North, Wallis and Weingast2009) have done a careful exercise in what can be called comparative statics: they do little to indicate what drove the progressive changes that they delineate, and almost nothing to explain the different evolution of other countries. But the understanding that they offer of the interrelations of institutions at a point in time and space would be (as indicated above) vital in ‘interfering effectively’. Why Nations Fail, by contrast, is exhilarating in its range. Acemoglu and Robinson are careless in their use of concepts, above all in what they mean by institutions. Their grasp of history is uncertain, and of politics, a little crude: thus their Just So stories. Nonetheless, here are influential mainstream economists who have abandoned the assumption that institutions should be presumed to be functional, and who now make clear that economies can benefit greatly from activist states and (in a range of circumstances) lose much by trade. And they do dynamics, often persuasively: a major achievement.

The most remarkable contribution is van Zanden’s. Unlike Acemoglu and Robinson, his arguments are always soundly based historically and politically. Unlike North et al. (Reference North, Wallis and Weingast2009), they are dynamic. He explains the extraordinary institutional fertility of the High Middle Ages in Western Europe, and shows where it led. Few readers, medievalists included, will get through van Zanden and de Moor's chapter 4, Girlpower, without astonishment. It would already seem, from the literature on economic development over the last century, that one of the best leading indicators of successful development is the status of young women (WBGDG, 2003; World Bank, 2001); now we see that holds too for the dawn of the modern age. In that case, it is difficult to exclude culture and religion entirely from the ranks of likely determinants of institutional change.