The Gentry Strikes Back

The Dissolution of the Monasteries and British Economic Growth

TL;DR: I review “The Long-Run Impact of the Dissolution of the English Monasteries,” by Leander Heldring, James Robinson, and Sebastian Vollmer. I laud the paper's data collection, research question, and design and definitely buy the main results on social change and commercialization. But I'm skeptical about the paper's pitch as an explanation of the Industrial Revolution.

Main points to be addressed:

1. The Dissolution of the monasteries and resultant land sales allowed a new class of middling landowners, the gentry, to accumulate holdings.

2. Hereditary copyhold was flexible enough to allow productivity growth, but did retard commercialization and agricultural productivity on the margin by inhibiting the efficient allocation of holdings.

3. However, the commercialization that followed the Dissolution (small, diffuse effects) was of a different nature than that which characterized British industrialization (enormous but regionally concentrated boom).

4. Monastic lands were mostly in the “agrarian periphery” of the south and east; few of the northwest's high-skill, low-wage textile regions were monastic.

5. Thus there probably is little connection between the mechanization of the cotton sector and the Dissolution VIA the specified channels of A) commercialization B) Catholicism (except on a national level) and C) gentry entrepreneurship.

6. Since industrialization was heavily regional, studying the IR in Britain at the scale of the parish might be looking at the wrong aggregation level.

7. It's possible that early land reform did differentiate Britain from France and Germany, but the effects by the mid-nineteenth century are pretty small.

8. Two reasons for (7): first, this is basically a persistence paper, so effects tend to attenuate over time; and second, commercialization itself helped to erode feudal tenures.

---

There's another exciting new paper out on British development: "The Long-Run Impact of the Dissolution of the English Monasteries," by Leander Heldring, James Robinson, and Sebastian Vollmer and published in the QJE. The authors use a 1535 survey of ecclesiastical properties to show that the expropriation and sale of clerical land by Henry VIII wiped out the inefficient feudal land tenure institutions that had prevailed across much of England after the Black Death, thereby increasing innovation, raising crop yields, and promoting structural change and industrialization.

They link this shift to the Industrial Revolution and Great Divergence through the so-called "commercialization hypothesis"—that Britain's precocity was due to the pervasiveness of its markets in land and labor. Markets could only form, however, once property relations had been clarified, and this happened (thanks to the Dissolution) 250 years earlier in Britain. Indeed:

The lagged abolition of feudal land tenure in France and Germany may be behind why England pulled ahead on the world stage in the eighteenth century. Continental Europe only transformed after their political revolutions in the nineteenth century finally did away with servile labor and customary land tenure relationships.

It's clearly a great paper, using a ton of novel data at a high level of spatial disaggregation—thousands of parishes!—and a plausibly important (if not exogenous) shock to answer a big historical question. But is it the right answer, and if so, how much does it explain?

"Mostly" and "a little," I'd say. But first, some background is in order.

In 1300, nearly one-half of the English population were serfs—unfree peasants who were bonded to a lord or a specific piece of land. This status was hereditary, not contractual, and was subject to the arbitrary authority of lord-controlled manorial courts. Serfs, also known as "villeins," also had their freedom of movement restricted and were forced to pay fines and labor dues to their masters. But the Black Death, which arrived in England in June 1348, brought serfdom to an end. The decline in population—and resulting labor scarcity—gave villeins increased bargaining power, which they used to renegotiate their tenurial status.

One of the best deals a villein could strike was a copyhold, so called because a copy of the contract was kept at the local manor court. There were two kinds: of inheritance, which was perpetual, and for lives, which generally lasted for three generations; both allowed the ex-villein to pay a low fixed nominal rent, useful in inflationary times. Since copyhold rents were lower during the Tudor period than for leaseholders, freeholders, and tenants-at-will, villeins had the incentive to take the arrangement when they had the power.

Heldring et al. argue that the Church was able to resist peasant bargaining power better than the lay lords—though, oddly enough, they attribute this to "more aggressive" behavior and being "less willing to give way." Why weren't the lords? I would have suggested that the Church as an organization was stronger than most individual lords. Consequently, copyhold was less frequent on monastic lands. They confirm this by looking at 2,136 manorial contracts owned by monasteries, finding that 13% were copyholds of inheritance, versus an estimated 33% for England at large. Elsewhere, copyhold predominated—2/3 of land in 1688—and endured until the nineteenth century.

What's so bad about copyhold? Using a formal model, Heldring et al. predict that copyhold of inheritance would lead to lower investment, reduced labor mobility, and inefficient matching between farmers and land. The reductionist summary: tenants pay a fixed nominal rent and thus get all of the net cash flows from investments. But the investment is specific: if they leave, they forfeit it, because they can't sell. With rising mobility and increasingly attractive outside options, a tenant invests less than if the landlord A) farmed the land or B) let it at market rent. Mobility is reduced because tenants want to keep their investments. And there will be little incentive for the most productive landlords to buy the best lands because the productivity gains would all go to the tenant. The problem was less bad with copyhold for lives, whch could be renegotiated upon expiry (100 years), and freehold.

In The Great Divergence, Kenneth Pomeranz cites the early collapse of hereditary land tenure regimes in northwest Europe as a key factor in that region's rise. "Much of western Europe’s farmland was far harder to buy or sell than that of China," he writes, and even in 19th-century England, "50 percent of all land in England was covered by family settlements, which made it all but impossible to sell." Hereditary tenure made it expensive, if not impossible, to consolidate and enclose land, which in turn prevented investment and the planting of fodder crops for livestock feed. The Anglo-Dutch adoption of the "new husbandry" began the Agrarian Little Divergence within Europe, helping to promote urbanization and raise incomes.

In the 1530s, there were 825 English monasteries owning 1/3 of the land in England and Wales. These holdings were spread out across the country, but as can be seen on the map, were concentrated in the South and East. In 1534, Henry VIII declared himself head of the Church of England. His goal, as a centralizing monarch, was to appropriate the tax revenues that the churches and monasteries had previously remitted to the pope, and so he conducted a survey of all ecclesiastical properties—the Valor Ecclesiasticus—in 1535 to figure out how much he could get, and where he might get it. Then between 1536 and 1540, the Dissolution got underway. Parliament passed a series of acts transferring the ownership of all English monasteries to the Crown, sometimes including the monastic buildings themselves. Monks and nuns, in this case, got state pensions and were told to pack their bags. When the expropriation was less peaceful, the Dissolution ended in the destruction of hated Catholic relics.

Henry's original plan was to manage his new holdings as a long-term tax base. Land revenue was an important way that late medieval and early modern kinds, who were expected to "live off their own", could end-run the power of their elites over the purse strings. But with war with France and the Holy Roman Empire looming at the end of the 1530s, Henry decided to dump the properties for immediate cash. He sold the buildings to his friends and commissioners of the Court of Augmentations, appointed to oversee the sales, generally offloaded individual manors at the fixed price of 20 years income, many of which went to the former managers—bailiffs and stewards.

So you can basically think of this as a great big land reform. Factors of production—land and people—get reallocated via the market for the first time. Holdings get owners with the commercial incentive to cultivate.

Why does this matter? There are basically two stages—the agrarian and the social. I'll take these in turn.

Land Reform

We've already talked about the modeled deleterious effects of copyhold of inheritance—it reduced investment, mobility, and productive land allocation. This, if correct, would have kept yields low and inhibited structural change. Does this play out in practice? Heldring et al. pitch their paper as a test of the "commercialization hypothesis"—that "the ability of factors of production to be allocated commercially through the market, rather than via feudal regulation or custom" explains Britain's precocious growth. They attribute this to Henri Pirenne, Karl Polanyi, and John Hicks (poor Robert Lopez! Paul Sweezy! though it should be noted that the Pirenne thesis concerned medieval Europe and the decline of feudalism!).

But I think their model has a tiny bit more Brenner (sans class struggle) than Pirenne. The old commercialization thesis is about market commerce eroding the inefficient feudal system; here, however, the authors turn the causality around. You need to get a fluid market for land going so that landlords can consolidate holdings and lease them to "capitalist tenant farmers" who could invest in improvements, exploit scale economies, and hire wage labor. Even though Brenner emphasizes the autonomous role of pre-existing class formations, he does argue that the sixteenth century was the critical stage of English development, when the landlords forestalled the move toward freehold tenure, suppressed peasant revolts, and began "to engross, consolidate and enclose." In France, meanwhile, the state—which sought to derive revenues from a peasant base—protected hereditary tenure and fixed fines, forcing landlords to buy up many small holdings to put together a contiguous unit.

This is an intuitive theory. But it goes against some relatively recent British historical literature. Bob Allen, for example, has argued that the copyholder was effectively the owner of his plot and thus could sell, sublet, or innovate as he pleased. French and Hoyle (2007, p. 11), whom the authors cite, state that “it is not clear why the survival of copyhold should have inhibited capitalist development, because copyholds could be bought, sold and let just like any form of freehold property.” And Allen's fabulous book Enclosure and the Yeoman (1992) strove to demonstrate that open field farmers—yeomen—on copyhold land were equal to, if not leading, the agrarian capitalists in advancing productivity during the Agricultural Revolution.

The first-order results look mostly good. The authors don't have data on the thickness of land markets. But they do show that parishes with monastic property were 9 percentage points more likely to have a market survive from 1516 to 1600 (vs a sample mean of 33%) and had fewer copyholds in 1842. For the second result, the OLS coefficient across all parishes is essentially zero, meaning that non-monastic and monastic parishes had the same number of copyholds. Using a long-difference strategy on a much-reduced sample (310 vs 2394 obs) consisting only of parishes with known copyholds pre-Dissolution, they do get a significant negative effect. So places that had copyholds lost them faster if they were monastic.

The Rise of the Gentry

Instead of citing Brenner and looking for agrarian capitalists, however, Heldring et al. situate their analysis in the context of the longest-running debate about the effects of the Dissolution: the "storm over the gentry," as it was humorously termed by the historian J. H. Hexter. The furor was kicked off by R. H. Tawney—he of the EHS's Tawney Lecture—back in 1941. In a famous (to my mind) paper, "The Rise of the Gentry," Tawney argued that the century-and-a-half leading up to the Glorious Revolution saw the emergence of a new class of middling landowners-cum-commercial-farmers. The "gentry" were

the landed proprietors, above the yeomanry, and below the peerage, together with a growing body of well-to-do farmers, sometimes tenants of their relatives, who had succeeded the humble peasants of the past as lessees of demesne farms; professional men, also rapidly increasing in number, such as the more eminent lawyers, divines, and an occasional medical practitioner; and the wealthier merchants, who, if not, as many were, themselves sons of landed families, had received a similar education, moved in the same circles, and in England, unlike France, were commonly recognised to be socially indistinguishable from them.

Mocked as crude and gauche, this "upper layer of commoners" absorbed an increasing area of land from the yeomanry and aristocracy. Heldring et al. contend that the proportion of land owned by the gentry increased from 25% in 1436 to 45-50% in 1688 while the shares of the yeomanry and great landowners remained stable. The Church and Crown share, however, crumbled from 25-35% to just 5-10% over this period—indicating that the soon-to-be-gentry picked up and profited from post-Dissolution sales. There were 200 knightly and 4,000-5,000 esquire/gentlemen families in 1524, according to the lay subsidy of that year; 500 and 16,000 respectively in 1600; and 620 knights, 3,000-3,500 esquires, and 12,000-20,000 gentlemen (per Gregory King) in 1688.

For Brenner, the gentry were, like the "great landlords," powerful enough to undermine peasant property rights and promote enclosure and consolidation. But Heldring et al. are interested in a slightly different angle: as proto-industrialists. One weakness here is that the gentry aren't rigorously defined—not in Tawney's work, and even less so by Heldring et al. We are simply to take them as being "more entrepreneurial" than either the major landlords, the Church, or the yeomen. That said, while Eric Jones (2021) has questioned whether the gentry were wholly economically motivated, he confuses absolute levels for relative status. The gentry only have to be marginally better than other proprietors at land management and business practice to have some effect.

The authors muster some anecdotal evidence that this is true, though I'd have to say it's a little disappoiniting. They cite John U. Nef, the great English economic historian, to the effect that the gentry of Lancashire and the West Riding mined and rented out the coal pits on their lands; and if Langton (1979) is to be believed, nearly 50% of the region's collieries were gentry-owned. Some members of the gentry also formed partnerships with entrepreneurs, including Thomas Bentley's financing of Josiah Wedgewood. But one can equally well find examples, as Jones does, of gentry frivolously dissipating their wealth, and of peers exploiting the coal mines on their property. Some better research on the social origins of the gentry and a more concrete definition of that class would make the "improving gentry" thesis a little more plausible.

So the "rise of the gentry" channel works as follows: Dissolution => liquid land market => gentry land purchases => resource exploitation + industrial financing => industrialization.

Religion and Reformation

The second channel the authors explore is the Reformation, which, in England, involved the systematic ideological and economic persecution of Catholics. By the 1559 Act of Uniformity, all English men and women were forced either to attend Protestant churches on Sunday or pay a monthly fine of 12 shillings, raised to 20 pounds in 1581. So-called "recusants" were also subject to imprisonment and the forfeiture of land and goods. Between 1600 and 1642, 102 recusant Yorkshire families had their estates seized, and religious fines could erode as much as 20% of a landowner's income.

It got worse, too. James I tacked on a ban of Catholics from the professions and public offices, supplemented by an oath of allegiance whose abnegation could result in life imprisonment and the confiscation of all property. William and Mary forced Catholics to pay double the rate of land taxes, forbid them to buy land, and allowed Protestant next-of-kin to claim their inheritances. Unsurprisingly, English people preferred Protestantism to persecution; the number of Catholics declined from, well, the whole population in 1530 to just 40,000-60,000 in 1600.

The authors argue that conversion depended on the opportunity cost. Since the Dissolution increased the value of land in monastic regions, the incentive to convert and not get expropriated must have been larger. So altogether:

In sum, we hypothesize that the Dissolution’s immediate effect was on markets and the allocation of factors of production. Following Tawney, we hypothesize that there was an intermediate impact of the Dissolution on social change. Finally, we hypothesize that ultimately there was a reduced-form effect of the Dissolution on industrialization, in line with the commercialization explanation of the English Industrial Revolution.

The intermediate results for gentry and Catholic presence are very strong. Controlling for parish area (though not population), they find that the number of gentry was higher by .23, relative to a mean of .67, in monastic parishes; this is basically confirmed in the long-diff specification. They also see a one-percent decline in the Catholic share against a mean of three percent. These are highly significant effects (1 percent), but at first glance... they're not very big? Unfortunately, they are not able to discern how much parish land the gentry owned—so maybe an additional gentleman would have bought up the entire parish, or maybe just a little fraction.

Commercial Capitalism?

The authors test a battery of outcomes for the commercialization hypothesis, including shares in agriculture and trade/handicrafts and the presence/number of textile mills (in 1838). The agricultural share fell by three percentage points (against a mean of 62 percent) in monastic parishes, while the trade/handicraft share rose by just two pp (vs a larger mean of 18 percent). Heldring et al. argue that the 3 pp fall in the ag share all went into industry, which thus increased by 11 percent... but that's not really even a large fraction of an already-small fraction. It's hard to see that making a critical difference. They also can't distinguish between the addition and the movement of gentry to monastic parishes. By their own data, the average parish had fewer gentry in 1700 than pre-Dissolution, and while the monastic mean rose from .77 to .87, the non-monastic mean fell by an even greater proportion, from .73 to .58. How great was the Dissolution really if it just polarized parishes?

The authors also find that monastic parishes were more likely to house (.01) and had more (.11) textile mills in 1838. Which again, isn't a lot (especially for 1838, when industrialization is well underway), but since it's a solid percentage of the sample mean (.04 and .16) the authors interpret this as a substantial effect.

However, this presumes that the spatial distribution of economic activity was relatively even—in other words, that monastic parishes had a few more textile mills than non-monastic parishes. But 15,588 of the 16,290 parishes in the sample had no mills at all and 453 more had just one, meaning that fewer than three hundred parishes (1.5%) had more than one mill. Textile production wasn't evenly distributed throughout the country, but concentrated in a couple of highly localized regions like Lancashire and Yorkshire. The first 42 parishes ranked by number of mills were in these two counties alone! And the biggest, Whalley, was way out in front of the others with 141 mills. The next largest was Ashton Town with 86. Only the top 7 had more than 50 mills. This polarization calls into question the use of the probability of having a mill as the main measure—having one mill does not tell you about industrialization at all

And as you can see on the map, monastic properties (marked in blue) were mostly concentrated in the south and east, NOT in the industrial northwest. Lancashire ranks 34th out of 42 counties in the fraction of parishes monastic. #1 on that list, Bedfordshire, actually has no mills at all. None. Indeed, a simple regression of the fraction of monastic parishes on the mean number of mills at the county levels is negative and statistically significant. It's definitely driven by outliers! Drop Lancashire and there's no correlation. But that's my point.

Also, the Kelly et al. mechanics paper shows quite compellingly that textile employment was concentrated in high-skill areas of the northwest—but not because they were great farming regions. Rather, it's because their poor soils and agricultural productivity led, with declining internal transport costs, to regional specialization in "industry" according to comparative advantage. The rich agricultural south, by contrast, became Britain's domestic agricultural periphery. If you take a look at the map in the top right pane, you can see that the real agricultural wage maps onto the Dissolution zone almost perfectly. And the high-skill/textile region fits into the no-monastery area almost as well.

The regression specification used in the paper has county fixed effects, which means that parishes are compared only to parishes within the same county. So the differences that you're seeing are real—a monastic parish probably was more likely to have a textile mill than a non-monastic mill in 1838. But I'd argue that this misses conceptually the defining characteristic of the IR, which was extreme regional concentration.

So we have a conceptual problem here. We're looking at the wrong kind of differences here, and possibly at the wrong level of aggregation. Controlling for counties actually negates the variation (region/county-level) which is actually how industrialization operated. In a way, it's very similar to an issue that we talked about in the context of Kelly et al.'s "Mechanics" paper. They can explain variation in economic development in Britain, but not Britain's advantages over Europe and Asia, which limits the project as an explanation of industrialization and the Great Divergence. Here, Heldring et al. can explain economic development within a parish, but the same doesn't hold true at the regional level—the scale that is most relevant for describing Britain's economic geography.

So insofar as this paper purports to be an explanation of the Industrial Revolution, I don't think it quite succeeds. But that's not all that it has to offer. Fundamentally, as I said before, this is a paper about how land reform leads to agricultural dynamism and commercialization. And I think it generally does do that. The authors show, for example, that of 234 agricultural patentees with occupation data, the single largest class (52 individuals) was the gentry, dwarfing engineers (16), though artisans defined more broadly together (as expected) made a larger contribution. Monastic parishes did have significantly more patents, as well as more threshing machines. My caveat (again) is that the small number of positive outcomes (234 and 409 respectively) relative to the sample size (16,000+) makes this hard to interpret, but the results are really precisely measured!

Heldring and co. also suggest that the gentry may have successfully pushed for enclosure, a la Brenner, a process that had to be initiated by the landlord. True to form, enclosure was 8% more likely in monastic parishes (sample mean 37%). Given their other paper showing that the parliamentary enclosure increased wheat yields by 44%, they suggest that this may have been a major channel for raising productivity. Wheat yields were also slightly higher (.24 bushels per acre against a mean of 21.71) in a reduced sample of 4,000 parishes, albeit at a lower significance level.

It's also worth revisiting the original Tawney thesis here. Tawney focused on the political economy consequences of the "rise of the gentry," especially the creation of a faction in Parliament with vested interests in commercial development.

With the growth of speculative dealings in land, the depreciation of the capital value of certain categories of real property by the antiquated form of land-taxation known as the feudal incidents became doubly intolerable. The more intimately an industry-agriculture or any other-depends on the market, the more closely is it affected by the policy of Governments, and the more determined do those engaged in it become to control policy. The fact that entrepreneur predominated over rentier interests in the House of Commons, was, therefore, a point of some importance. The revolt against the regulation by authority of the internal trade in agricultural produce, like the demand for the prohibition of Irish cattle imports and a stiffer tariff on grain, was natural when farming was so thoroughly commercialised that it could be said that the fall in wool prices alone in the depression of 1621 had reduced rents by over L800,000 a year (Tawney 1941).

It's curious that they focused on the gentry-industrialist connection, because to my mind the original Tawney formulation is more interesting. How you get elites on the side of (or at least indifferent to) your development project is super important, and a lot of pre-industrial countries just couldn't do it. On the continent, we know, it took the French Revolution to wipe out feudalism and its attendant aristocracy. But what if Britain basically sold tickets to a new elite? That would've diluted the power of the old, consumption-based aristocracy and made compromises between manor, merchant, and manufacturer more likely (a la Mokyr-Nye).

Speaking of the French Revolution, the authors suggest that the Dissolution helps to explain why Britain set out early on the special path to capitalism. In one of the many papers that we all wish we'd written, Acemoglu et al. (2011) argue that feudal land tenure (as well as guilds and mobility restrictions) lapsed in Europe only after Napoleon's conquests. Using the French invasion of Germany as a natural experiment, they show that places invaded by the French experienced more rapid urbanization and structural transformation during the nineteenth century (with the caveat that much of it came after 1850).

Another paper, Finley et al. (2021), basically replicates the Dissolution analysis for France, studying the effects of the confiscation and sale of Church properties during the Revolution. The area sold comprised 6.5% of France's territory and had been plagued by a maze of restrictions that inhibited investment and proper land allocation, including uncertain common rights and feudal exactions due to the Church. Even the great spurt of enclosures that took place after 1750 (belatedly) was rendered nugatory by hereditary tenures, as in pre-Dissolution Britain. The situation was probably worse in France thanks to the monarch’s support for peasant rights as a bulwark against the nobility.

They find that more confiscated land led to higher wheat yields and larger farms, which they attribute to reduced transaction costs in the land market. So it's plausible that the greater persistence of hereditary tenure in the Continental agrarian system might have delayed Franco-German development for 250 years.

What do I take away from this paper? I think we learn the following:

1. Market-based land reform can lead to major social change by allowing new, commercially-minded classes to collect and manage property.

2. Changing the property structure of landholding has a positive impact on agricultural innovation, organization, and productivity, with some consequent effects on structural change.

3. The shift toward pro-landlord agrarian property relations came later than Brenner thought and relied not just on pre-existing, immutable class formations but also on historical policy shocks, institutions, and cultural change.

4. Single shifts, even large ones, produce effects that pale in comparison to the dynamic changes that swept the early modern British economy; much more of the wheat yield improvement, for example, still appears to have come from manuring/rotations/legumes/seed varieties that evolved slowly over time.

NB: It’s important to remember that this is an economics paper published in an economics journal. Proving that the mechanism outlined in the model could have some effect probably outweighs the historical argument either way. If you have an important economics-related contribution, you can—for better or worse—be flexible with the history.