The Law of Primogeniture and the Transition from

Landed Aristocracy to Industrial Democracy∗

Graziella Bertocchi

Universita’ di Modena e Reggio Emilia, CEPR and IZA

January 2006

Forthcoming in the Journal of Economic Growth

∗I would like to thank Matthias Doepke, Oded Galor, Thorvaldur Gylfason, Andy Mountford, Mike

Spagat, Yong Wang, Fabrizio Zilibotti, and workshop and seminar participants at ESSIM, SED, SET,

NBER Summer Institute, the Workshop on Intergenerational risk sharing and contracts, the CEPR Con-

ference on Dynamic Aspects of Policy Reforms, ASSET, the Winter and the Summer Meetings of the

Econometric Society, the Workshop on Economic Growth over the Very Long Run, and the University of

Modena, for helpful comments and suggestions. All errors remain of course mine. Financial support from

the Italian University Ministry and the European Commission is gratefully acknowledged. Author’s ad-

dress: Dipartimento di Economia Politica, Universita’ di Modena e Reggio Emilia, Viale Berengario, 51,

I-41100 Modena, Italy, phone 39 059 2056856, fax 39 059 2056947, e-mail bertocchi.graziella@unimore.it,

web http://www.economia.unimore.it/Bertocchi Graziella/.

0

ABSTRACT

We study the connection between inheritance systems and the historical evolution of the

relationship between a society’s economic structure and its political system, with a focus on

Europe from feudal times. The model predicts that, in an early agrarian phase, aristocratic

political systems prevail, while democracies tend to emerge with industrialization. At the

same time, as indivisible landed estates are replaced by capital as the primary source of

wealth, the inheritance system evolves endogenously from primogeniture to partition. The

dynamics of output, distribution, class structure and political participation are in turn rein-

forced by the system of intergenerational wealth transmission, with primogeniture tending

to concentration and partition to equalization.

JEL Classification Numbers: O40, P10, J12 , N10 , K19, Q15.

Key Words: Primogeniture, land, aristocracy, industrialization, democracy.

1

“But the law of inheritance was the last step to equality.”

Alexis de Tocqueville, Democracy in America (1835).

1 Introduction

Back in the Middle Age, Europe was a predominantly agrarian economy ruled by a feudal

aristocracy which based its power on the land. Several centuries later, Europe’s societies

are now democracies which have completed, and passed, their industrialization process. Few

attempts have been made to explain the long-term determinants of wealth distribution,

class structure and political participation and their connection with the process of structural

reallocation from agriculture to manufacturing. This paper fills this gap by focusing on the

role of the system of intergenerational transmission of wealth as a crucial link between a

society’s economic structure and its political system.

We start from the observation that, at least in the European experience during the past

ten centuries, agrarian societies tend to be inherently aristocratic, while industrialized soci-

eties, as capital replaces land, tend to move into the direction of democratization. At the

same time, we observe an evolution in the legal system regulating the intergenerational trans-

fers of property rights, with primogeniture prevailing in agrarian aristocracies and partible

inheritance gaining ground as industrialization and democratization advance. Our conjec-

ture is that the source of this association is to be found in the fact that feudal estates were

subject to a politico-economic indivisibility constraint while capital, being reproducible and

divisible, could be easily partitioned from the beginning of the accumulation process. With

wealth at the roots of political power, when there is a change in the primary source of wealth

political power will shift as well, and the effect is amplified if land and capital are subject

to different rules of intergenerational transmission, with primogeniture reinforcing wealth

inequality in aristocracies and partition promoting equalization in democracies.

Our main focus is on the period running from 1300 to 1800, which witnessed the transition

from feudalism to early forms of capitalism, starting from the decline of the primarily agrarian

and aristocratic society of the 14th century and ending with the emergence of manufacturing

2

and the establishment of bourgeois political power. Nevertheless, our investigation contains

implications also for subsequent economic and political developments.

The agrarian basis of Europe’s political order dates back to the introduction of feudalism

at the turn of the first millennium. Feudalism consisted of the following set of arrangements.

First of all, feudal lords were granted by kings the income from great estates, in return for

military service and for the local administration of justice. Indivisibility of the estates was

crucial to generate sufficient income and was enforced through a system of legal provisions.

The nature of maintenance costs evolved over the centuries, but they always remained very

large. Initially, military costs prevailed, while the costs of supporting a court and sustaining

high living standard for reasons of prestige became more substantial later on.

Another crucial characteristic of feudalism was the imposition of feudal rights on peasants,

in the form of a variety of services and dues. Despite the fact that agrarian relationships were

often conflictual, the feudal organization also accomplished the goal of providing peasants a

safe and stable environment. The organizational form of the landlords-peasants relationship

evolved over time The organizational form of this relationship has changed over time, as first

documented by Smith ([1776] 1937), who described how serfdom was initially replaced by

sharecropping contracts and then by leases involving a rent plus other serviced performed

by small farmers for the benefit of the landlord. Cross-country comparisons indicate that

modernization occurred first in England, while Continental Europe only joined in after the

French revolution. Generally speaking, as manufacturing and urbanization developed, feudal

lords started to lose control on workers’ mobility.

Finally, class structure in the feudal society was based on the distinction between lords

and peasants, with the nobility formally involved in the exercise of exclusive political power.

The first challenge against aristocratic power was represented by enriched merchants de-

manding freedom from feudal control on society. The bourgeois revolutions in England and

France established the role of a newly-formed socioeconomic group opening the way for the

supremacy of democratic capitalism over agrarian elitism.

In this framework, primogeniture emerged around the 13th century as a reaction to in-

tensified demographic pressure and kept spreading across Europe up to the 17th century.

3

Despite marked regional differentiations, for centuries it remained the rule among the nobil-

ity, and whenever land constituted a dominant portion of wealth, because of a need to keep

large estates together. The French revolution abolished entails, but it is only with the 19th

century that most European countries finally outlawed them, despite all the controversies

and the fact that the demographic, social and economic reasons for them to emerge were

already changing much earlier.

Eminent economists and political and social scientists contributed to the debate on pri-

mogeniture, which spread from Europe to colonial America. The relevance of the inheritance

law is witnessed for instance by Smith ([1776] 1937), who writes: “When land was considered

as the means, not of subsistence merely, but of power and protection, it was thought better

that it should descend undivided to one (...). The security of a landed estate (...) depended

upon its greatness. To divide it was to ruin it, and to expose every part of it to be oppressed

and swallowed up by the incursions of its neighbors (...).” While from de Tocqueville ([1835]

1956) we read: “But the law of inheritance was the last step to equality (...). When framed

in a particular manner, this law unites, draws together, and vests property and power in

a few hands; it causes an aristocracy, so to speak, to spring out of the ground. If formed

on opposite principles, its action is still more rapid; it divides, distributes, and disperses

both property and power.” Marx and Engels ([1844-45], 1975) also remarked how one of the

victories of the bourgeoisie was the one of the “division of land over primogeniture”.

The goal of this paper is to establish how the evolution of the law of inheritance is

connected with the growth path of the economy, the sectoral composition of aggregate output,

class structure, wealth inequality, and political participation. To achieve this goal, the

paper embeds the above-mentioned features - indivisibility of landed estate, feudal rights

on agricultural production, wealth as the basis of political power, primogeniture over the

land - within an otherwise standard specific-factor model with an agricultural sector and

a manufacturing sector. There are two groups in the population, landlords and landless

workers. Workers can migrate from agriculture to manufacturing. The analysis is framed

within an overlapping-generations model with an altruistic bequest motive. The model’s

innovations are founded on the following three assumptions, which are further justified in

4

Section 6 on the basis of the available historical evidence.

First, there exists a politico-economic constraint on the minimum size of landed estates,

dictated by the need to raise sufficient income to ensure their protection and maintenance.

As an implication, land cannot be alienated and primogeniture emerges as the efficient

way to regulate the intergenerational transmission of property rights on land. Since no

indivisibility constraint is imposed on capital, equal partition among all children instead

regulates transmission of this component of society’s wealth.

Second, political power is determined by wealth. Land ownership immediately entitles

the aristocracy to political power, while there exists a minimum capital requirement for

active participation of landless workers in the political process. We assume majority voting

with restricted suffrage as a simple device to capture the idea that the aristocracy begins

with all the power but gradually loses it as the economy develops. Specifically, individuals

vote on the removal of feudal privileges, with the median voter deciding the outcome.

Third, feudal privileges take the form of the appropriation of a portion of the agriculture

product by the landlords. This organizational form implies non-competitive agricultural

wages, while in the manufacturing sector the labor market is competitive.

Without these three assumptions the model would generate a transition from agriculture

to manufacturing driven purely by capital accumulation, as in standard dual growth models.

Under our assumptions, however, the model generates much richer dynamics, with structural

transformation interacting with the political balance of power.

In the initial, primarily agrarian phase, land is the main form of wealth, so that land-

lords hold all political power and can impose feudal rights on peasants. As capital accu-

mulates, workers start migrating towards the manufacturing sector, while the agricultural

sector shrinks. As the workers’ capital bequests increase with accumulation, the process of

democratization begins by allowing active political participation. Once a majority formed

by the richest workers reaches the minimum wealth requirement, the distortive feudal rights

are abolished. As a result of this political choice, capital accumulation and income growth

accelerate, so that wealth equalization and democratization further progress.

Throughout the process just described, primogeniture over the land emerges as the ra-

5

tional choice of altruistic parents subject to a politico-economic indivisibility constraint,

while capital can be partitioned freely among all children. However, since the importance of

land as a portion of total wealth declines with capital accumulation, we observe over time a

reduction of the incidence of primogeniture. Thus, the mechanism that generates the tran-

sition towards an industrial democracy is reinforced, since primogeniture preserves wealth

concentration while partition promotes its dispersion.

To sum up, the model establishes a connection between the evolution of society from

an aristocratic political system into a democracy, the process of capital accumulation and

sectoral reallocation from agriculture into manufacturing, and the economic impact of pri-

mogeniture, thus generating a theory of endogenous determination of the inheritance system

based on the evolution of wealth composition.

The rest of the paper is organized as follows. Section 2 explains how the paper relates to

the literature. Section 3 describes the basic model. Section 4 analyzes the model and derives

the main results. Section 5 considers a number of extensions of the basic model. Section 6

looks at the historical evidence and provides for it an interpretation in light of our results.

Section 7 draws some conclusions and indicates lines for future research.

2 Related literature

This work is primarily connected with the research program which has focussed on the

connection between growth, development and institutions in a long-term perspective. Early

examples from this literature are North (1981), Engerman and Sokoloff (1997), Acemoglu

and Robinson (2002), Bourguignon and Verdier (2000), and Bertocchi and Spagat (2004),

who study the behavior of elites in the face of economic and institutional developments. The

joint evolution of the economic and political system is modeled, among others, by Justman

and Gradstein (1999) and Cervellati et al. (2005), while a specific stream of this literature has

focused on franchise extension (see Acemoglu and Robinson (2000), Bertocchi and Spagat

(2001), Lizzeri and Persico (2004), and Llavador and Oxoby (2005)). Closely connected,

another research body has built a unified theory that captures the historical evolution of

population, technology and output from Malthusian stagnation to Solowian growth (see

6

Galor and Weil (2000) and Hansen and Prescott (2002)). Galor and Moav (2004, 2006)

and Doepke and Zilibotti (2005) are also closely related for their analysis of the long-term

evolution of class structure. We contribute to this literature by showing the connection

between politico-economic evolution and intergenerational wealth transmission.

There is also a literature on inheritance rules and the family, with some contributions

attempting at finding an endogenous explanation of primogeniture and family structure, for

which Becker (1981) provides a general introduction. Chu (1991) and Bergstrom (1994)

develop a lineage survival approach, where primogeniture can emerge as an optimal dynastic

policy to minimize the probability of extinction, while Faith and Tollison (2001) motivate

partition as a device to minimize rent seeking among children. Ekelund et al. (2002), within

their analysis of the Protestant Reformation, treat primogeniture as a proxy for stability of

wealth distribution, which was higher in countries with strong feudal elites, lower economic

growth and less-developed entrepreneurial activity.1 Nobody however has yet looked at the

primary source of wealth - land vs. capital - as an explanation of the prevailing law of

inheritance.

On the role of land in dynamic models, we were inspired by the dual growth tradition

initiated by Lewis (1954), and in particular by the work of Eaton (1987) and Drazen and

Eckstein (1988), from whom we take the basic structure of the model although they assume

constant population. Laitner (2000), in a similar setup, allows for population growth but

not for capital accumulation. Sectoral shifts of the labor force from agriculture to man-

ufacturing are at the basis of the Kuznets (1955) inverted-U relationship between income

inequality and income. Recent work on the role of agriculture in development includes

Caselli and Coleman (2001), who focus on regional growth patterns in the U.S. during the

past century, and Gollin et al. (2002), who stress the role of agricultural productivity in

explaining cross-country income disparities. Different patterns of agrarian production orga-

nization are studied by Conning (2002) and Mookerjee and Ray (2002), while the adverse

impact of landowners’ interests for education reform is explored by Galor et al. (2002) and

Falkinger and Grossmann (2004). Boix and Garicano (2001) stress the relevance of country-

1Botticini and Siow (2003) study the optimal joint determination of bequests for sons and dowries for

daughters, while Bernheim and Severinov (2003) compare differences in division rules for gifts and bequests.

7

specific wealth, of which land is the most typical and extreme example, in delaying the

democratization process.2 None of these papers however explicitly analyzes the role of the

intergenerational transmission of property rights on land within a neoclassical growth model

with capital, population growth and technological progress.

3 The model

At some point in the development process, a combination of events initiated the transition

from a static agrarian economy to a dynamic two-sector economy with manufacturing. The

new political and economic order established through feudalism had permitted the expansion

and the improvement of cultivated land and therefore the emergence of technical progress in

agriculture. The resulting wage and consumption dynamics had pushed towards a permanent

increase in the rate of population growth and the accumulation of an agricultural surplus

sufficient to start the process of capital accumulation.

We start modeling our economy once the above process has been completed. The newly-

determined economic conditions, characterized by technical progress in agriculture, popu-

lation growth, and capital accumulation, are fully embedded in our basic framework. The

model is also designed to capture the gradual decline of feudalism and the associated loss of

control on workers’ mobility under the competition of the growing modern sector.

3.1 Production

The economy displays two specific-factor technologies producing an agricultural good and a

manufacturing good using three factors: land, capital, and labor. Land is in fixed supply,

and equal to L. At time t, production of the agricultural good, Y At , uses labor NAt and land

L, according to

Y At = (GtL)

αNAt1−α (3.1)

2On the related issue of land reform, see Horowitz (1993), Grossman (1994) and Conning and Robinson

(2005).

8

where 0 < α < 1, Gt is the level of technological progress in the agricultural sector and

Gt+1 = (1+ g)Gt. The quantity of effective land is therefore given by Lt = GtL. Production

of the manufacturing good, Y Mt , uses labor NMt and capital Kt according to

Y Mt = Kα

t NMt1−α (3.2)

The agricultural good is not storable and can be used only for consumption, while the

manufacturing good can either be used for consumption or bequeathed to children, who will

in turn employ it as capital in the manufacturing sector. Labor is perfectly mobile between

sectors at no cost, consistently with our goal to describe the declining phase of strict feudal

control.3

3.2 Population

We consider an overlapping-generations model with bequests where individuals live for two

periods. Total population grows at a rate n and consists of two groups: landlords Lt and

landless workers Wt.4 At time 0, the size of the landlords’ group, L0, is small if compared

to the size of the workers’ group. When our story starts each landlord owns an equal share

e of the total amount of land L, where e represents the minimal size of an estate. The size

of the estate cannot fall below e because the costs of protection and administration, to be

covered out of the agricultural rents, are assumed to be infinite if its size is smaller than e.

For simplicity we can set costs to 0 for any estate equal or larger than e.5 Since eL0 = L,

setting e = 1 we obtain L0 = L. Land is passed on from parents to children. Since land is

not allowed to be put for sale, there exists no market for land. Given our assumption on

minimal estate size, land is passed on to one child only (unigeniture). This implies that the

size of the landlords’ group cannot increase over time. Therefore, Lt = L, for any t. It is

3Frictions determined by the landlords’ ability to retain a partial degree of control could be easily added

to the analysis, without modifying the main results.4While population growth is assumed to be exogenous, it could be endogenized as a positive function of

income, so that landlords would have higher fertility than workers. This would reinforce the process we are

about to describe.

5More realistic assumptions about estate costs would complicate but not affect the results.

9

natural to assume that the heir of the estate is always the first child (primogeniture).6 We

assume that all individuals, including landlords, receive a labor endowment. The landlords’

younger, landless children - the cadets - become part of the workers’ group.

Each individual also receives a capital bequest. As standard, being a reproducible, divis-

ible resource, capital can be divided up among all children in equal parts without restrictions

(partition) even among the landlords’ class.7 Without loss of generality, we assume that all

individuals start with an identical level of initial capital bequest, b0, and that the relative

size of e is larger than b0, so that at the start capital represents a negligible portion of the

landlords’ wealth. Moreover, the aggregate size of the initial capital bequests is also small,

so that aggregate initial wealth in the economy is composed mostly of land.

While from the above analysis we have obtained a constant size of the landlords’ group,

the workers’ group grows over time at a rate nt ≡ n Nt

Nt−L > n because of the inflows of cadets.

The workers’ group is internally stratified since it includes, beside the descendants of the

original workers, first-generation workers, i.e., cadets just coming from the aristocracy, as

well as previous-generation workers, i.e., children of cadets who came from the aristocracy at

earlier times. Figure 1 maps the pattern of population growth for each group, for t = 0, 1, 2,

and for the simple case of n = 1 and only one individual per group at t = 0, a landlord l0

and a worker w0.. The subscript indicates the period in which each individual is born, while

the superscript is introduced only for individuals during whose family history a change of

group has occurred.8

6While the model does not explicitly predict that primogeniture would prevail over unigeniture, traditional

arguments - such as the advantages of the first born in providing for the old and/or taking care of the land

- can be invoked to explain the prevalence of the former in actual practice.7Our results would hold even under the assumption of unequal division for capital as well, as long as

the degree of indivisibility for land remains higher than for capital. In an extension in Subsection 5.1 we

introduce the possibility of compensatory capital bequests for cadets.8The difference between the subscript and the superscript indicates how many generations have past since

the change occurred so, for example, w01 denotes a worker born at time 1 whose father was a landlord (i.e.,

w01 is the the second born of the landlord), while w02 denotes a worker born at time 2 whose grand-father

was a landlord.

10

Figure 1: The inheritance tree

3.3 Political system

In this model, political power is acquired through wealth. Landed property is therefore a

source of political power, and in fact the only one in the early phase of the development

process. In this early phase, the landlords therefore constitute an aristocracy i.e., literally,

a political system that exclusively assigns power to a few. Accordingly, as we shall see, they

are in a position to impose feudal rights on the peasants, i.e., the workers in the agricultural

sector. Initially, capital plays no role in the political equilibrium of the model since we

assume there is a minimum wealth requirement, eb > b0, for participation of workers in the

political process. The wealth threshold is not to be interpreted as a rule specified by the

voting mechanism and thus chosen by the elite in power, but as a parameter meant to capture

the broad fact that a minimum, exogenously-determined socioeconomic status in necessary

to permit political activism.9

It is only when all workers reach a level of wealth eb that the political system turns into

a full democracy, i.e., literally, a political system that assigns power to all the people, on

a one-person-one-vote basis. However, for workers to play a role in the political process, a

9Our wealth threshold therefore differs from the formal wealth, income, and education requirements which

were later introduced in the 19th century legislation concerning suffrage, as the expression of a choice by the

elite.

11

partial degree of democratization will suffice. And, since we assume that decisions are made

through majority voting, workers will be in a position to determine the political outcome

as soon as enfranchised workers outnumber the members of the aristocracy. As we shall

see, enfranchisement in this model will be gradual and will progress at an endogenously-

determined speed, even if all workers start with an identical initial capital bequest.

The analysis will prove that a landlords’ and a workers’ majority express different policies.

In particular, under aristocracy, landlords’ privileges consist of the imposition of feudal rights

on agriculture, which are abolished once the median voter is a worker. We assume that, once

feudal rights are abolished, they can never be reinstated. In the following discussion, even

though in this historical phase the structure of the political system is rudimentary, and our

tax simply proxies for more complex feudal relations, we shall refer to “voting” and “tax

policies” to indicate the means and the outcomes of the political process.

3.4 Preferences

In the first period of life, individuals are born with one unit of labor that they supply

inelastically, and they receive their land and capital bequests, which are employed in the

production processes together with their labor. Only the first born of the aristocracy receive

a land inheritance. Production takes one period to be completed, so during their second

period individuals receive their wage, interest, and rental income. Rental income only accrues

to the landlords. In the second period individuals optimally choose their level of consumption

and a capital bequest for their children. The landlords also bequeath their estate. Second-

period consumption of an individual born at t is given by ct+1 = cAt+1 + cMt+1, where cAt+1

and cMt+1 are the individual consumption levels of the agricultural and of the manufacturing

goods, respectively. The agricultural and the manufacturing goods are assumed to be perfect

substitutes in consumption, i.e., their relative prices are fixed and set equal to 1. Preferences

are given by

uit = δ log cit+1 + (1− δ) log bit+1 + et+1 (3.3)

12

for i = L,W , where δ is a preference parameter, with 0 < δ < 1, cit+1 is consumption, and

bit+1 and et+1 are the capital and the land bequests to the children, respectively.10 To be no-

ticed is that we have assumed an identical utility function for landlords and workers, so that

primogeniture is not to be attributable to different preferences. We have also assumed iden-

tical labor and capital endowments for all individuals, so the only exogenous heterogeneity

in the economic fundamentals of the model has to do with land endowments.

3.5 Factor prices

We shall now focus on the implication of feudalism for the landlords-peasants relationship.

To capture its non-competitive nature, we assume that peasants do not receive their marginal

product, but an average of what is left after landlords appropriate a fraction λ of the total

agricultural product, with 0 < λ < 1. We assume that subsistence is always guaranteed,

and that λ is exogenously determined as the result of a landlords-peasants conflict so that

its value increases in the relative strength of the landlords.11 In this setup each landlord

receives from his estate a rent which is given by

ρt = λGαt (NAt

L)1−α (3.4)

and is increasing in the level of employment in agriculture and in the level of technological

progress.12 Note that, were land partitioned among all the landlords’ children, the appropri-

ate denominator Lt would also grow over time eroding this component of landlords’ income,

but this is prevented by primogeniture. The individual wage is

ωAt = (1− λ) (GtL

NAt)α (3.5)

Even though feudal rights represent, by their nature, the appropriation of a fraction of

total output, it is instructive to derive their impact on individual wages by introducing a

parameter τ such that λ = [α+ τ(1− α)]. It follows that factor prices can be written as

10A non-linear functional form for the land bequest component et+1 would not affect any result.11In a bargaining framework, λ would be equivalent to the exponent of the landlord’s payoff in the expres-

sion for the Nash product. Mookerjee and Ray (2002) fully develop a bargaining framework to determining

landowners’ taxes on workers’ wealth accumulation.

12In terms of land units, the rent received by each landlord is instead given by λ(NAt

GtL)1−α.

13

ρt = [1 +τ(1− α)

α]αGα

t (NAt

L)1−α (3.6)

ωAt = (1− τ) (1− α) (GtL

NAt)α (3.7)

where λ > α for τ > 0. Therefore, for τ = 0 factor prices coincide with their respective

competitive levels and the feudal “tax” τ can be interpreted as a measure of the degree to

which agrarian relations depart from competition.13

Maximization of (3.6) subject to the choice of τ reveals that the relationship between rents

and feudal rights is humped-shaped, which implies the existence of an optimal level of the

tax from the landlords’ vantage point. However, to keep the following analysis manageable,

we assume that the level of the tax rate τ (or, equivalently, λ) is an exogenous parameter of

the model, set lower or equal than the maximum rate at which the rent with tax is equal to

the rent without tax. For future reference, we highlight this condition as follows.

Assumption 1. At any t the individual rent in the presence of feudal rights is higher than

the individual rent in their absence.

Under the above setting, tax policy will simply be given by the choice over two possible

tax levels, either a fixed τ > 0, or 0.

In the manufacturing sector both the capital and the labor market are competitive.

Indeed even in its early phase the manufacturing sector has been characterized historically

by a totally different structure of property rights and organizational forms, if compared

to agriculture. This justifies a different modeling strategy, in keeps with the development

literature previously cited. Assuming perfect competition in the manufacturing labor market

is an extreme assumption. However, our results are robust to the introduction of a tax

imposed by capitalists on manufacturing wages as long as this tax is lower than the one

on agricultural wages, a reasonable assumption to make since the level of the tax depends

on the relative strength of the groups involved, and manufacturing workers were certainly

stronger and better organized than peasants.

13The competitive rent for each unit of effective land Lt is given by α(NAt

GtL)1−α.

14

Profit maximization yields the following wage and interest rate:

ωMt = (1− α) (Kt

NMt)α (3.8)

rt = α(NMt

Kt)1−α (3.9)

The capital stock is subject to full depreciation. Migration ensures the equality of the

wage between the two sectors, i.e., ωAt = ωMt ≡ ωt. Equating (3.7) and (3.8), we obtain the

following expression for NAt :

NAt = GtL(1− τ)1αNMt

Kt(3.10)

which shows that, as expected, the number of workers in agriculture is a decreasing function

of the level of the capital stock.14 More interestingly, (3.10) also shows that employment in

agriculture decreases in the intensity of feudal rights. This is consistent with the fact that,

once the development process has opened up new opportunities and control over mobility

has been relaxed, whenever landlords-peasants relations are less favorable towards peasants

the rural system suffers more keenly the competition of the growing modern sector, which is

characterized by very different organizational forms.

Figure 2 illustrates wage and employment determination in each sector at any t.When the

tax is in place, the demand for labor in agriculture shifts to the left, implying a wage ωT lower

than the fully-competitive level ωC , lower agricultural employment and higher manufacturing

employment. As we shall see, over time the demand for labor in manufacturing will shift

up and to the left, pushing up the wage level and absorbing more workers.

We use the notation yit, i = L,W, to denote individual i0s total income inclusive of capital

inheritance but exclusive of land inheritance at time t. Therefore we have yLt = bLtrt+ ρt+ωt

and yWt = bWtrt + ωt.

14Were the two technologies characterized by a lower labor share in agriculture, as widely accepted, equa-

tion (3.10) would predict a smaller size of agricultural employment and output, relative to manufacturing.

15

Figure 2: Wage and employment determination in the two sectors

16

4 Analysis of the model

4.1 Utility maximization

Each individual maximizes utility (3.3) subject to the constraints

(i) cit+1 + bit+1 ≤ yit

(ii) et+1 ≥ 1(iii)

Pj b

jit+1 = (1 + n) bjit+1

for i = L,W, where bjit+1 is the individual capital bequest received by child j. The first

constraint simply states that the individual’s consumption and capital bequest have to be

less or equal to total income inclusive of the capital bequest received by the individual. Note

that, independently on which sector individuals receive their income from, they can consume

any good since they can trade intragenerationally at the unit price. The second constraint

captures the indivisibility of estates due to our assumptions about maintenance costs, where

1 represents the normalized minimum estate size. The third constraint imposes identical

capital bequests to all children, i.e., rules out compensatory capital bequests.

The solution is cit+1 = δyit, bit+1 = (1 − δ)yit, bjit+1 =

bit+11+n

, for i = L,W, and et+1 = 1.

The last expression implies that bequeathing land to a single child is not an assumption or

the expression of parents’ preferences, but is the rational response of altruistic parents to

the indivisibility of landed estates.

In their first period, those individuals whose capital holdings are higher than eb vote onthe tax policy which will affect their remunerations at t+ 1.

4.2 Market clearing

Given the above results, the market-clearing condition for the capital market is

Kt+1 = (1− δ)yLtL+ (1− δ)yWtWt (4.1)

Aggregating over individual incomes, we obtain

Kt+1 = (1− δ)(Y At + Y M

t ) (4.2)

17

where employment in each sector is given by (3.10) and by the following condition for market

clearing in the labor market:

NAt +NMt = Nt (4.3)

Note that at the aggregate level the unit price assumption, together with our preferences,

implies that some manufacturing goods must be consumed, otherwise the price of the man-

ufacturing good would have to be higher than the price of the agricultural good. In other

words, Kt+1 < Y Mt and the saving rate, given by (1− δ) = Kt+1

Yt, is bounded above by

YMt

Yt,

where Yt = Y At + Y M

t .

The agricultural and the manufacturing goods markets must also be cleared at each t,

implying

Y At = cALtL+ cAWtWt (4.4)

Y Mt = cMLtL+ cMWtWt +Kt+1 (4.5)

where cAit and cMit , for i = L,W , are the individual consumption levels of the agricultural and

of the manufacturing goods for the currently old.

As previously explained, no market for land exists since land is never put for sale. Another

feature of the model is that, while the saving rate is constant over time, the portion of

the manufacturing output which is invested at each t must vary, being given by Kt+1

YMt

=

(1 − δ)(1 + Y At

YMt) which is larger than (1 − δ) and, as we shall show, decreases over time as

Y At

YMtdeclines.

4.3 Political-economic equilibrium

Definition 1 A perfect-foresight political-economic equilibrium is a sequence

{cALt, cAWt, cMLt, c

MWt,Kt, Y

At , Y M

t , Gt, NAt, NMt}∞t=0 such that at each t utility and profits are

maximized, all markets clear, and tax policy is set by the median voter, starting from given

initial values of K0, N0, and G0, and given the set of parameters n, g, α, δ, L, τ and eb.18

The equilibrium path solves simultaneously equations (3.6)-(3.9) and (4.2)-(4.5), plus the

wage-equalization condition (3.10).

4.4 The output-maximizing tax rate

From (3.10) and (4.3) we can derive the following expression for total output at each t:

Yt = {[ψ(τ,Kt)]1−αKα

t + [1− ψ(τ,Kt)]1−α(GtL)

α}N1−αt (4.6)

where ψ(τ,Kt) ≡ Kt

Kt+GtL(1−τ) 1α. Output maximization by choice of the sectoral allocation

of labor yields a tax rate equal to 0, with ψ(0, Kt) ≡ Kt

Kt+GtL. It follows that the level of

employment in agriculture increases with tax abolition. Therefore, a change in tax policy

positively affects the process of capital accumulation and income growth by removing the

distortion on sectoral labor allocation. This is summarized by the following proposition.

Proposition 1 (i) At each t, the level of output for τ = 0 is higher than the level of output

under a positive tax; (ii) at each t, employment in agriculture (manufacturing) for τ = 0 is

higher (lower) than employment in agriculture (manufacturing) under a positive tax.

For τ = 0, the agricultural sector is characterized by a new, competitive organizational

form, so that the process of structural adjustment is temporarily slowed down with a reallo-

cation of workers towards agriculture. However, as we shall clarify below, this return to the

land is only temporary and does not prevent the industrialization process.

4.5 The balanced-growth path

The dynamic evolution of the system can be analyzed using the capital market clearing

condition (4.2), starting from an initial condition K0, where K0 is assumed to be small.

Along the equilibrium path, workers migrate to the manufacturing sector as the capital

stock grows, while the agricultural sector shrinks. Therefore, the agriculture share of out-

put varies inversely with the level of development. In the long run, the dynamical system

evolves towards a unique and stable balanced-growth path, where all the relevant variables,

19

and in particular Kt, Lt and Yt, must grow at a rate n.15 The rate of growth of technological

progress in agriculture, g, has to be equal to n in order to sustain such a path with an

appropriate expansion of the quantity of effective land since, with land as a fixed factor,

population growth has to be balanced by an identical growth of effective land. In other

words, despite the presence of technological progress, in this model a balanced-growth path

involves constant land-labor and capital-labor ratios. Moreover, as implied by Proposition 1,

the higher the intensity of feudal power, the lower capital, output and agricultural employ-

ment, the higher manufacturing employment. After the abolition of the tax, the process of

structural adjustment progresses at a higher speed leading to higher balanced-growth levels

of Kt and Yt. We can therefore state the following.

Proposition 2 (i) In the long run the economy reaches a balanced-growth path if and only

if g = n; (ii) in balanced growth, the capital-labor ratio and the output-labor ratio for τ = 0

are higher than their respective levels under a positive tax.

Figure 3 illustrates the evolution of the capital-labor ratio kt ≡ Kt

Nt, given by kt+1 =

(1 − δ)yt, where yt ≡ YtNt. The highest trajectory describes the evolution of kt for τ = 0,

i.e., under competitive factor prices in agriculture. The corresponding balanced-growth

level of the capital-labor ratio is kC . The lowest trajectory is drawn under the hypothesis

of ψ(τ,Kt) = 1, i.e., as if the entire labor force were employed in manufacturing so that

yt = yMt , with yMt ≡ YMt

Nt. The intermediate trajectory finally describes the evolution of

kt when the tax is in place and the capital-labor ratio converges to kT .16 The effect of the

tax is to shift the competitive trajectory down by compressing employment in agriculture.

The higher the tax, the stronger the effect and thus the discrepancy between kC and kT .

Starting from an initial situation in which the tax is in place, the economy moves along the

intermediate trajectory until the time at which the tax is abolished, and then jumps on the

15See Eaton (1987) for a formal proof.16Since all sources of income accrue to individuals in the same period of life, the impact of feudal rights

on accumulation depends on sectoral reallocation and not on intergenerational redistribution as in Drazen

and Eckstein (1988).

20

Figure 3: The evolution of the capital-labor ratio

21

highest.17

4.6 Evolution of income and wealth distribution

We shall keep track of the evolution of factor prices, income and wealth accumulation as

time goes and sectoral reallocation progresses, taking into account the impact of a change

in tax policy. From (3.7) and (3.8) the evolution of wages can be expressed as

ωt+1ωt

= (1 + n

1 + nAt)α = (

Kt+1/Kt

1 + nMt)α (4.7)

where nAt and nMt are the rates of growth of employment in agriculture and manufacturing,

respectively. Before the system converges to the balanced-growth path, nAt < n since nAt < 0,

and capital grows more than employment in manufacturing, so wages increase over time. In

balanced growth they will be constant, since employment in the two sectors grow at the

same rate as population and capital. Examine now the evolution of the interest rate, which

is given by

rt+1rt

= (1 + nMnKt+1/Kt

)1−α (4.8)

which implies declining rates until they settle at a constant balanced-growth level. Rents

per unit of effective land show a similar pattern, given by(1+nAt

1+n)1−α. However, what matters

for each individual landlord is the evolution of actual rent per estate, i.e., their individual

rent, which is given by

ρt+1ρt

= (1 + n)α¡1 + nAt

¢1−α(4.9)

with individual rents declining less than effective rents - or possibly growing depending on

the relative size of n and nAt - before balanced growth is reached, and growing at a rate

17Balanced growth can be reached before or after the tax is abolished. In the former case, if the economy

has already settled at kT , tax abolition constitutes a permanent shock that causes the system to move to

the new, upper path leading to a new balanced-growth equilibrium, kC .

22

n = g in balanced growth. Therefore, as the size of the landlords’ group is kept constant by

primogeniture, the actual rental income of each landlord grows.18

Let us turn now to the between-classes differential evolution of income and wealth. Out-

side balanced growth, income grows for each individual - albeit at different rates. In balanced

growth, the wage and interest rates are constant, while the rent component of landlords’ in-

come keeps growing indefinitely following agricultural productivity. The difference between

the level of income of a landlord and a worker is given by yLt − yWt = (bLt − bWt)rt+ ρt.

Clearly the gap is always positive even if we start with the same initial capital. Once balanced

growth is reached, the rent component keeps increasing while everything else is constant,

implying an increasing differential.19 The capital bequest of the landlords will always be

higher than the capital bequest of the workers. Thus the aristocracy will retain its economic

supremacy (even though, as we shall see, not its political supremacy) not only because of its

additional source of income through land ownership, but also because of its diversification

into capital and manufacturing. In fact, since land as a source of income becomes negli-

gible if compared to capital, the second factor plays a decisive role in this process. When

considering the distribution of total wealth, the weight of land will also be declining, since

before balanced growth is reached capital is growing more than effective land. This implies

a reduced impact of primogeniture on the intergenerational transmission of the economy’s

wealth, that further dilutes wealth concentration. So the economic importance of primogeni-

ture declines with capital accumulation and sectoral reallocation towards manufacturing. In

the long run, partition regulates intergenerational transfers for most of existing wealth and

for most families.

We shall now examine the impact of the feudal rights. By substituting (4.3) and (3.10)

into the expressions for the factor prices (3.6)-(3.9), we obtain their equilibrium values as

follows:

18Were primogeniture replaced by partition, the rate of growth of rent per landlord would be lower, and

given by a factor of(1+n)α(1+nAt )

1−α

(1+n)1−α , which reduces to (1 + n)αin balanced growth.

19In balanced growth the aggregate rent of the landlords’ group grows at n = g like the other aggregates.

23

ωt = (1− α) (Kt +GtL(1− τ)

1α

Nt)α (4.10)

rt = α(Nt

Kt +GtL(1− τ)1α

)1−α (4.11)

ρt = [1 +τ(1− α)

α](1− τ)

1−αα αGt(

Nt

Kt +GtL(1− τ)1α

)1−α (4.12)

While the wage increases with tax abolition, the interest rate decreases, and so does the

rent under Assumption 1. We collect the above results in the following proposition.

Proposition 3 (i) The landlords’ individual income and capital bequest are always larger

than the workers’ individual income and capital bequest; (ii) over time the proportion of land

out of total wealth declines, and so does the incidence of primogeniture; (iii) the abolition

of the tax has a positive effect on the wage rate and a negative effect on the interest and

individual rental rates.

4.7 Evolution of class structure

A crucial implication of primogeniture is the fact that the model endogenously creates mul-

tiple income and wealth classes, as new cadet lines are created at each t. By cadet lines,

we mean generations of younger children of the aristocracy that at each t are excluded from

land ownership and join the workers’ class, starting with a capital bequest which differs from

that of the original worker lines, as well as from that of the previously-created cadet lines.

In other words even assuming, as we did, identical initial capital bequests, at each time t we

observe t+ 2 types of individuals defined by the different level of the bequest they receive.

The first type, with the highest bequest, are the first born of the aristocracy who also receive

a land bequest (at each t, we have a line of landlords, i.e., of first born at each previous t of

the original landlords, for each original landlord). A second type is represented by landlords’

second born, who receive a capital bequest equal to that of the first born, but no land. A

third type, with the lowest bequest, characterizes individuals who come from the original

workers (again, we have a line of “pure” workers, expanding at the rate n, for each original

24

worker). Individuals of this third type are the most numerous. The remaining, intermediate

t−1 types come from subsequent generations of cadets who are expelled from the landlords’class at each t, with larger bequests for the more recently created cadet lines. For example,

as in Figure 1 assume that n = 1 and that at t = 0 we have only one landlord and one

worker, with bL0 > bW0. At t = 1, there will be one first born of the landlord, receiving a

certain capital bequest plus land, a landless second born of the landlord, receiving an iden-

tical capital bequest but moving into the workers’ group, and two children of the original

worker, receiving an identical bequest. So indeed at t = 1 there are three types of bequests

being received. We can therefore state the following proposition.

Proposition 4 At any t there exist t + 2 types of individuals characterized by the bequests

they receive, t+ 1 of which belonging to the workers’ group.

Given the stratification of bequests within the workers’ group, we shall indicate as the

“middle class” those workers who have received a sufficiently large bequest to achieve political

participation. By excluding younger children from landed property, primogeniture plays a

crucial role in affecting the composition of the emerging middle class and its chances of

upward social and political mobility. Moreover, capital accumulation and thus the growing

incidence of partition will intensify the pressure towards wealth redistribution.

4.8 Evolution of political participation

We shall refer to the extension of political participation as “democratization”. From the

above discussion, it is clear that most recent generations of cadets will always be at the

top of the wealth distribution within the workers’ group, so that they will be the first to

cross the wealth threshold eb. When the process of democratization begins, the number ofthe new voters is simply nL, so for n < 1 the newly enfranchised will be outnumbered

by the landlords and will not be able to affect tax policy. Period after period, however,

additional relatively wealthy cadet lines join the workers’ group and add to the middle class,

and at some point political participation will also be obtained by the rest of the workers,

with previous-generations cadets still coming first. Gradually, all individuals will be able

25

to participate into the political process, so full democratization eventually prevails for a

sufficiently low eb.However, for tax policy to be affected, it is sufficient that a majority of the enfranchised

is constituted by workers, so that a worker is the median voter, provided that indeed workers

prefer tax abolition and landlords prefer its retention. This is ensured under the conditions

spelled by the following lemma.20

Lemma 1. Under Assumption 1, for any bLt and bWt such that bLt > bWt, workers prefer

tax abolition and landlords tax retention if and only if the following conditions hold at any

t :

(i) bWt < βWt, and

(ii) bLt > βLt,

where βWt ≡ (1−α)(Σαt −Γtα)αNt(Γ

α−1t −Σα−1t )

, βLt ≡ {Gt[1+ τ(1−α)

α](1−τ) 1−αα Γt1−α−Σt1−α

(ΓtΣt)1−α(Γtα−1−Σtα−1) − (1−α)(Γtα−Σtα)αNt(Γtα−1−Σtα−1)}, Γt ≡

Kt +GtL and Σt ≡ Kt +GtL(1− τ)1α .

Proof: See the Appendix.

Lemma 1 establishes conditions under which indeed the abolition of the feudal rights

benefits the workers and hurts the landlords, so that it will be implemented only when the

median voter is a worker. Conditions (i) and (ii) have an intuitive interpretation. Notice

that, while the wage component of income increases with tax abolition, the interest rate

decreases, and so does the rent under Assumption 1. For conflicting preferences over tax

policy to arise, we need that, when the tax is abolished, landlords’ income goes down while

workers’ income goes up. The issue is not trivial because both workers and landlords own

capital and supply labor.21 So for workers the loss from capital holdings must be smaller than

the gain from labor remuneration, while for landlords the rental loss must not be compensated

by an increase in the labor-plus-interest income. Condition (i) ensures that workers’ capital

20Despite the heterogeneity of income and bequest levels among workers, with a slight abuse of notation

we shall still denote them by yWt and bWt without introducing any additional indexing.21Early two-class models in the tradition of Kaldor (1955) avoided this issue by assuming that workers

only receive labor income and do not save, while the elite only receives property income and invest it. This

simplification implies a set of highly ad hoc assumptions regarding preferences and distribution.

26

holdings are indeed small enough not to compensate with an interest-income loss the wage

gain. Condition (ii) in turn ensures that landlords’ capital holdings are large enough that

their labor-plus-interest income does not grow more than the rental loss or, equivalently,

that the combined decline of rent and interest income is larger than the wage gain. Note

also that Assumption 1 per se would not ensure that (ii) holds true, while it implies that

the capital threshold for landlords is smaller than for workers, i.e., βLt < βWt. To sum up,

Lemma 1 requires that workers must hold a relatively small amount of capital, while the

opposite must be true for landlords, in order to obtain conflicting preferences over tax policy

with workers voting for abolition. Given that accumulation is faster for landlords than for

workers, the conditions imposed in Lemma 1 are mild ones.

A corollary of Lemma 1 implies that each generation of cadets joining the working group

must hold a bequest in (βLt, βWt). The bequest must be below βWt for these individuals,

once turned workers, to prefer abolition, but since the bequest is identical to that of the first

born, it also must be above βLt for their landowning siblings to prefer retention.22

In the following analysis we assume that the conditions established for Lemma 1 are

satisfied for landlords and for all workers. The following lemma establishes a condition for

workers to translate their preferences into actual political power.

Lemma 2. The feudal tax is abolished when a number of workers larger than L holds a

bequest in [eb, βWt).

Proof: See the Appendix.

Lemma 2 restricts the size of the workers’ bequest to an interval over which they prefer

abolition and at the same time they can express their preference through voting. A number

of workers larger than the number of landlords must hold a bequest in this interval for

workers to set tax policy. A corollary of Lemma 2 is that eb must be smaller than βWt for any

worker to be able to vote. Under the assumptions spelled by Lemma 1 and 2 we can finally

derive the following proposition.

22A cadet richer than βWt would instead ally with the aristocracy in retaining the tax. An extension

explores this possibility in more detail.

27

Proposition 5 (i) The process of democratization is a gradual one; (ii) the tax is abolished

once the median voter is a worker; (iii) tax abolition can occur before full democratization

obtains.

The process of enfranchisement will therefore begin gradually, with cadets voting first,

and will be followed initially by the abolition of feudal rights and then by full democratiza-

tion, i.e., by full participation of the entire workers’ mass to the political process, with the

timing depending on n, L, eb and the technology parameters. For some parameter constella-tions, abolition can actually occur only when the entire mass of original workers’ children is

enfranchised. If instead it occurs earlier, democratization will be accelerated once the mid-

dle class has conquered political power and feudal rights in agriculture have been abolished.

From the previous analysis, we know that the abolition of feudal rights by a workers’ major-

ity reduces the income gap between landlords and workers and also affects the subsequent

evolution of bequests and wealth distribution.

4.9 Summary

To sum up, the model evolves from an initial aristocratic stage, where landlords hold all

political power and exploit their feudal privileges. As capital accumulates, workers start

migrating towards the manufacturing sector, while the agricultural sector shrinks. The tax

leads workers to leave agriculture but also depresses the economy-wide wage rate, putting

the economy on a lower balanced-growth path and thereby delaying the establishment of a

democratic majority to abolish feudal taxation. As the workers’ capital bequests increase,

the richest workers will reach eb, marking the emergence of a bourgeoisie, and the process ofdemocratization begins. Once the tax is abolished sectoral reallocation will proceed, with

an acceleration of capital accumulation and income growth because of the removal of the

distortion, so that wealth equalization and democratization further progress. Throughout the

process, the importance of land as a portion of total wealth declines implying a reduction

of the incidence of primogeniture and further equalizing pressure. Therefore, we observe

an association among the evolution of society from an aristocratic political system into a

democracy, sectoral reallocation from agriculture into manufacturing and, as land is replaced

28

by capital in the composition of society’s wealth, a shift from primogeniture to partition. It

should be stresses that, while primogeniture over the land follows directly from our estate

indivisibility assumption, the evolution of the principal mode of transmission of the whole

of society’s wealth is an endogenous outcome of the model. We can therefore conclude with

the following main proposition.

Proposition 6 There exists an association among sectoral allocation in production, type of

political system, and system of inheritance. This association is founded on the nature of

the primary source of wealth, with agrarian economies being organized as aristocracies where

primogeniture prevails, and industrializing economies evolving towards democratization and

equal partition.

5 Extensions

5.1 Compensatory bequests

The historical literature documents the widespread use of non-land, at least partially com-

pensatory bequests for cadets, as well as dowries for daughters. In the present model, these

provisions would amount to a distribution of the capital bequests among children in an

unequal way, with the first born getting a smaller share, in order to compensate his primo-

geniture on the land.23 Compensatory bequests reduce wealth inequality between landlords

and cadets and increase it between cadets and workers. Under the maintained conditions

for Proposition 1, this implies a process of democratization which starts relatively soon even

though it will proceeds relatively slowly. However, if cadets turn out to be sufficiently rich

(i.e., condition (i) of Lemma 1 is not satisfied) it will be in their interest to side with the

aristocracy. It is well known that many of the cadets were actually entering the clergy or

the military, and therefore constituted a well-defined, separate social group whose interests

23A rigorous comparison between land and capital would require to consider the price of the two assets,

given by the present discounted value of the income accruing to each asset, i.e., effective rents and interests,

respectively. Since they both decline through time, we can abstract from their explicit consideration without

affecting the qualitative results.

29

were often closer to the aristocracy’s own. In this case, the workers’ struggle against the

aristocracy and its allies will inevitably be a longer one.24

5.2 An agrarian aristocracy

To capture a widely-shared, but only partially-verified perception of the aristocracy as a

group that did not contribute to industrialization, and remained attached to the landed

estates as its only source of wealth, consider the following variant of the basic model, which

can capture the late latifundia economies of Southern Europe. Landlords have no initial

capital b0, only workers do. Moreover, to reflect restrictions that were enforced in some

historical circumstances, the aristocracy is prevented from entering activities other than

agriculture by purchasing capital at later dates, so that it can simply consume its entire

rental and labor income. This implies that cadets are left with their own labor to supply,

and no capital, so they will always enter the workers’ class from the bottom. Therefore, the

children of the original workers will be the first to reach political voice and will immediately

outnumber the members of the aristocracy. The predictions of this variant of the model are

an early conspicuous consumption for the aristocracy, its relative impoverishment due to

lack of diversification, and the emergence of a middle class consisting of individuals of low

social extraction who will be able to reverse tax policy as soon as democratization begins.25

6 Historical evidence

The main purpose of this section is to justify on the basis of the available historical evidence

the main assumptions on which our model rests. In particular, we provide support for

the assumption that aristocratic power was based on the land, and that primogeniture was

common practice among the aristocracy because of a need to keep large estates together. We

also provide information on the role of cadets, the nature of feudal rights and the evolution

24Bertocchi and Spagat (2001) analyze the case of a middle class that is co-opted by the elite against the

lower classes within a divide-and-rule strategy.25Doepke and Zilibotti (2005) offer an alternative theory of the demise of the aristocracy, according to

which the rural aristocracy proved too impatient to invest in the new industrial technologies.

30

of class structure and political power within the relevant time framework. We indicate

how geographical differentiation and time variations may have influenced different areas of

Europe. The evidence we bring also confirms the main results of the model concerning the

process of growth and sectoral reallocation, the evolution of feudal rents, the dynamics of

class structure and the diminishing relevance of primogeniture.

6.1 The time framework

Up to the year 1000, after the fall of the Roman Empire and continuing through the phase

of the barbaric invasions, Europe went through a deep economic and political crisis. Income

and population stagnated. Things turned for the better in the next millennium, in part

thanks to the introduction of feudalism in the 9th century. Maddison (2001) provides long-

term data on population and GDP growth for Western Europe. Population stagnated until

the year 1000, and started growing in the 1000-1500 period (at an average annual rate of

0.16%, despite the occurrence of natural catastrophes such as Black Death). Population

growth intensified in 1500-1820 (+0.26%) and took off after 1820. Per capita GDP growth

in Western Europe is estimated to grow at an average annual rate of 0.13% in the 1000-

1500 period, 0.14% in 1500-1600 and 0.15% in 1600-1820, to jump to 1.51% in 1820-1998.26

So the period we are interested in witnessed a slow but steady process of development,

which accelerated only towards the end. Together with the sizeable increase in population,

the 1000-1500 period came with an expansion of the area of rural settlement and with

the gradual incorporation of technological changes that raised land productivity. The 12th

century is commonly viewed as the golden age of the feudal economic and political system.

Afterwards, agriculture started its transformation, trade began its development, urban

growth slowly accelerated. The impact of these phenomena implied a deterioration of the

feudal organization of society, which reached a deep crisis in the 14th century, preluding to

the transition towards mercantile capitalism. In extreme synthesis, in turn during the 17th

26For Eastern Europe, the pattern is quite similar, with average annual population rates of 0.15%, 0.31%

and 0.72% in 1000-1500, 1500-1820 and 1820-1913, respectively, and per capita GDP growth at 0.04% in

1000-1500, 0.10% in 1500-1820, and 1.06% in 1820-1998.

31

century mercantile capitalism entered a crisis leading to the rise of industrial capitalism.

In fact, the most important difference between the preindustrial and the modern industrial

societies rested on the greatly diminished role of agriculture in the latter, following the

progressive rise of manufacturing which started, at least in Britain, at the beginning of the

18th century. Accordingly, urbanization rates in Western Europe are estimated by Maddison

(2001) at 6.1 in 1500, 7.8 in 1600, 9.9 in 1700, 10.6 in 1800 and 31.3 in 1890. Nevertheless,

the influence of feudalism on the economy and society did not fade until the 19th century

(Pirenne (1925, 1963), Goldstone (1991)).

Overall, our focus is on the period running from 1300 to 1800, which witnessed the

transition from feudalism to capitalism, i.e., started from the decline of the primarily agrarian

and aristocratic society of the 14th century and ended with the emergence of industrial

capitalism and the establishment of bourgeois political power. Of course the presence of

cyclical patterns and marked geographic differentiation makes it difficult to trace a simple

historical trend and will force us to unavoidable generalizations.27

6.2 The aristocracy and the land

In the critical scenario of Middle-Age Western Europe, feudalism was introduced by the

Frankish kings in the 9th century as a system of military and political relationships aimed

at maintaining external security and domestic order. In order to guarantee a stable flow of

income to support military expenses and an organized bureaucracy, the feudal lords were

granted the income from great estates in return for military service and for the local admin-

istration of justice (Pirenne (1963)). By bringing political and economic stability, feudalism

contributed to the solution of Europe’s crisis. At the same time, it established the agrarian

basis of Europe’s political order and the roots of aristocratic power. Despite regional dif-

ferentiations but with substantial homogeneity within Europe, at least until the end of the

18th century the nobility remained a ruling social elite whose economic power was based on

27For example, the demise of the feudal organization occurred as early as in the 12th century where trade

grew faster, i.e., in Lombardy, Tuscany, Northern France, and Flanders, while it manifested itself only in

the 13th century in Germany and England. On the other hand, the subsequent phase of technological

improvement was faster in Northern France and Southern Britain, slower in the Mediterranean area.

32

the land and that exercised exclusive military and political power (Labatut (1978)). In some

countries, for example Russia and Poland, literally all the land was in the hands of the no-

bility, while this was not true for example in Sweden or France. In England the nobility held

a portion of the land much higher than in the rest of Europe (Dewald (1996)). Nevertheless,

the control of the land by the aristocracy was preponderant overall.

In the first few centuries of the modern age the European nobility faced a series of

transformations linked to the advent of capitalism and its impact on the agricultural sector.

While in the beginning land was the basis of the nobility’s wealth, diversification occurred

even at a relatively early stage. In 1500 land was still the only form of investment, and

in fact it was forbidden for a nobleman to become engaged in commerce or industry, but

progressively this rule was relaxed. For example, France in 1669 allowed the nobility to be

involved with maritime trade. Manufacturing and banking were also allowed by 1767. In

England the nobility was deeply involved in mining and trade (Labatut (1978)). Therefore,

despite the incontrovertible decline of the landed basis of aristocratic power, at the same time

the nobility was able to adapt to the changing economic conditions and, with the gradual

loss of importance of agriculture, we observe a progressive diversification of its wealth, with

involvement in commercial and industrial activities (Dewald (1996)). This process intensified

after the 17th century and by mid-18th century many members of the nobility invested in

commercial and financial activities (Lindert (1986)).

To conclude, the evidence confirms the role of land as the basis of aristocratic power

and, on the other hand, the ability of the aristocracy to diversify its wealth into trade and

manufacturing as land loses its central role.

6.3 The land and primogeniture

During the 13th century, primogeniture emerged as a reaction to the intensified demographic

pressure and the consequent inefficient partition of land. It kept spreading across Europe

up to the 16th and the 17th century, although marked differentiations in the local laws

and in the actual customs were present. Generally speaking, however, for many centuries

it remained most common among the nobility and whenever land constituted a dominant

33

portion of wealth. In particular, primogeniture was linked to feudal tenure, where it was

favoured for military reasons and generalized control and it was also supported by Church

and State (Goody et al. (1976)).

One should be aware that the regulation of inheritance may be based on the laws, the

customs or the actual practices, which need not correspond. Moreover, customary systems

of law continued to change even after the customs were codified. Nevertheless, and despite

the presence of variations within Europe, the broad similarities within which these variations

occurred allow to offer a unified picture of these developments. In the period between the

15th and the 18th centuries, the descent of land among great landowners was indeed regulated

by male primogeniture, settled by a system of entails which forbid alienation of the family

land (Goody et al. (1976)). Entails were introduced to preserve the property and name of

great houses who were founded on landed wealth (Werner (1998)).

Primogeniture is a distinctive feature of Western European society, practically unknown

to antiquity, which also emerged in other societies, such as Japan, where wealth distribution

was largely determined by land concentration. Conversely, where most wealth belonged to

traders and nomads and was therefore easily moved and partitioned, as in ancient Western

Arabia, egalitarian principles prevailed. Kuran (2003) reports how the Qur’an prohibits ex-

clusionary practices for Islam. Similarly, Platteau and Baland (2001) document how partible

inheritance always prevailed in pre-agricultural SubSaharan Africa.

If we focus on Europe in 1500-1700, a map of the customs of inheritance shows primo-

geniture predominating in England, Scandinavia and parts of France and Eastern Europe.28

Castilla also adopted it later through the so-called mayorazgo. Partition was prevalent in

trade-oriented 14th century’s Italy, where fideicommissary entails subsequently spread due

to Spanish domination, and gained ground as the country turned into a primarily agrarian

28Ekelund et al. (2002) acknowledge the existence of a controversy among historians regarding the preva-

lence of primogeniture in England. They argue that after the Normans introduced it, it was systematically

avoided through a number of devices, until the Statute of Wills of 1540 de facto abolished it (see Spring

(1993)), in accordance with the needs of a dynamic economy subject to market forces. However, if this

is generally true for aggregate wealth, landed estates were and still often are subject to primogeniture in

England.

34

economy.29 A differentiation in the customs even existed within two areas of 16th-century

France, with primogeniture (preciput) prevailing in the South and in the Walloon and parti-

tion in the West.30 In Germany the tradition in favor of partitioning estates had been strong

in the Middle Ages, but primogeniture gained ground in the 16th and early 17th centuries.

It was, however, never adopted in Russia, where the nobility was never in a position of power

in the face of the Tzars’ absolutism (Goody et al. (1976)), so that partition was employed

as an effective policy to keep it in a weak position. The consequences of primogeniture on

younger children were often harsh. However, it should be noticed that cadets were always

taken care of somehow (Labatut (1978)). The exclusion of younger siblings was rarely if

ever total, and meant in most cases only exclusion from the land, with provisions in money

(alimenta) as a partial compensation.31

The predominance of primogeniture throughout the centuries occurred despite all the

controversies that it provoked, and the fact that the demographic, social and economic

reasons for it to emerge were already changing by 1800. As de Tocqueville ([1835] 1956) well

understood, it is the law of inheritance that produces aristocracies, since the inheritance laws

based on primogeniture were able to protect the cohesion of the family wealth on which the

political stability of monarchies was founded, while partible inheritance would undo them and

promote social mobility. With the Enlightenment the maintenance of a wider distribution of

property started to be viewed instead as the foundation of political stability and the French

29Among the Venetian nobility fideicommissa became general after 1550. They were not accompanied by

primogeniture and simply prohibited alienation but, since usually only one brother married, the result was

the same as under primogeniture. As a consequence the Venetian nobility was on the verge of extinction by

the 17th century. Habakkuk (1955) discusses the link between law of inheritance and population growth.30However even if the law did not allow entails, still de facto entails were used to set up rules of succession

for great families. In addition, again in the actual practice only one son used to marry.31A distinction between laws of inheritance can also be observed for social groups below the nobility, with

primogeniture always being associated with land holdings. In early 16th-century England primogeniture was

noticeably extending among the gentry and by the 19th century it was adopted even by the middle classes.

German peasants in regions characterized by large land holdings were also practicing primogeniture, while

for centuries in the British Midlands there existed a clear relationship between type of inheritance and nature

of wealth, with only reproducible resources such as fishing and forestry being subject to partition (Goody et

al. (1976)).

35

revolution eventually abolished entails. In the middle of the 18th century, fideicommissary

entails were outlawed in Piedmont and Tuscany and soon afterwords England witnessed

attacks on strict settlements and the campaigns for free trade in land. But it is only with the

19th century that most European countries outlawed primogeniture, with some exceptions

like Germany and with parents’ discretion still surviving in certain cases as Britain.32

Overall, the historical evidence confirms the link between land and primogeniture, and

therefore between aristocratic power and primogeniture.

6.4 Feudal privileges

As previously mentioned, the organizational form of the landlords-peasants relationship has

evolved over time and varied across countries. In the present model the complexity of feu-

dal rights are captured by the tax parameter, which proxies for the variety of services and

other dues imposed on peasants (Pirenne (1963), Binswanger et al. (1995)). Geographical

differences were present, with Eastern Europe showing harsher conditions for peasants than

elsewhere (Labatut (1978)). The presence of feudal rents is identified by historians as a

growth-retarding factor, as confirmed by the performances of Russia as an extreme case33

(Dewald (1996)), but also of France if compared to England (Moore (1966)). Over time we

observe an evolution of feudal rents. Serfdom was slowly abolished after the 12th century,

but feudal dues persisted much longer. Following the initial crisis of feudalism many no-

bles, being unable to adapt to the new situation, became heavily indebted and eventually

ruined (Pirenne (1963)). However, the largest landowners were able to survive by initiating

a process of reorganization of production towards more efficiency. After the 17th century

with the decline of feudalism land did not provide a large income any longer, in part be-

cause of the tendency towards a progressive relaxation of the servile relationships, which

32The controversy on inheritance laws spread from Europe to colonial America. Virginia formally adopted

primogeniture, while the Massachusetts legislature adopted partible inheritance though it gave the elder

son a double share, an idea coming from Biblical sources. See Alston and Schapiro (1984) for a description

of inheritance laws in the colonies and De Long (2003) for a comparison with Europe.33A stricter enforcement of serfdom, disallowing migration out of agriculture, constituted an additional

growth-retarding factor for Russia.

36

intensified towards the end of the 18th century. The removal of feudal rights occurred first

in England, where the nobility reacted with a successful attempt to make the land more

productive. During the 17th and 18th centuries the methods of cultivation were modern-

ized, and enclosures of the landed properties was introduced.34 It was the French revolution

that notoriously completed the process of removal of residual feudal rights in agriculture,

with effects spreading with Napoleon from France to most of Continental Europe (despite

exceptions surviving in the latifundia of Spain and Southern Italy).

Therefore, consistently with the predictions of the model, feudal rents constituted a

growth-retarding factor, while the income of the nobility remained relatively high during the

first centuries of the modern age, despite regional and individual variations. In addition,

feudal power was abolished earlier in countries with more developed commerce and industry,

and therefore with smaller residual agricultural employment. At the same time the removal

of distortions made agriculture more efficient retarding its decline.

6.5 Estate costs

Turning to estate maintenance costs, over the centuries there was an evolution in their

nature even though they always consistently remained large (Pirenne (1963)). European

society at the close of the Dark Ages was characterized by pervasive military insecurity that

was caused on the one hand by the wars among the successors of Charlemagne, and on the

other by the incursions of Vikings, Magyars, and Moslems. This insecurity gave rise to an

extremely high demand for military protection (Volckart (2000)). The military revolution

of the 16th century dramatically increased the cost of war by changing its scale (Goldstone

(1991)). The costs of maintaining a court were another crucial component. Especially

after the 17th century, the mere standard of living also became extremely expensive for the

nobility which became more and more inclined to spend (Labatut (1978)). Religious reasons

were especially important in Spain, while elsewhere it was burdened with services due to the

State, and everywhere enormous sums were used for countryhouses and gardens. Finally, the

34Indeed O’Rourke and Williamson (2005) show that in England the ratio of wages to land rents started

rising only with the initial industrialization phase around 1850.

37

development of the market economy with the 17th century exposed the nobility to growing

needs of cash to sustain its living standards and social status.

In the face of these considerations, captured by our indivisibility constraint, for cen-

turies fragmentation of property was viewed as extremely dangerous for the preservation

of aristocratic wealth and power and, more generally, for political stability in aristocratic

systems.

6.6 Class structure and political power

Class structure in the early rural society was based on the distinction between lords and

peasants, with the clergy somehow in between. The nobility was formally involved in the

exercise of political power and the administration of the State. The first challenge against

aristocratic power was represented by rebellions starting in the 11th century within the cities

(Pirenne (1925)), led by enriched merchants demanding freedom from feudal control on so-

ciety. Later on, the civil war in England and the French revolution were two examples of

bourgeois revolutions under different circumstances, but characterized by a common deci-

sive factor, the formation of a social group with an independent economic basis which was

able to remove the obstacles that the agrarian heritage represented for the development of

democratic capitalism (Moore (1966)). This is consistent with our model’s assumptions and

results. More specifically, in England the rural elite played a leading role in the process

of promotion of modern organizational forms in agriculture, industrialization and political

evolution, and thus survived these developments with an osmosis between landed aristocracy

and capitalistic bourgeoisie that caused an early removal of feudal rights but at the same

time allowed the aristocracy to retain economic and political control. In France the lack of

independence of the nobility from the king and its reluctance to adapt led to a delayed but

more abrupt defeat of feudal power, guided by the capitalistic bourgeoisie that at the end

of the 18th century ended the ancien regime (Pirenne (1925)).

Overall the industrial revolution had little impact on the political role of the aristocracy

(Dewald (1996)). Still up until Waterloo the landed aristocracy kept its power, and up until

the middle of the 19th century the richest were still the landed aristocrats. Consistently

38

with the model’s predictions, the abolition of feudal rights and democratization did not

remove the nobility’s control of most of the economic, political and cultural life, at least

until the end of the same century. This supremacy, however, was challenged by the growing

middle class which, by mid-19th century, reached the political and economic leadership

in Western Europe (and, by the end of the same century, in most of Central Europe as

well). This middle class had made its appearance between the nobility and the workers

with the start of industrialization, and was further differentiated internally, including at

least two components, one closer to the lower aristocracy and the other to the working

class. In England, for instance, the early growth of capitalism caused the cadets to play a

particularly dynamic role within the middle class (Goody et al. (1976)). The revolutionary

potential of the disinherited offsprings of the aristocracy, pushed towards the army or the

clergy, is stressed also by Goldstone (1991) and Pirenne (1963), while on the other hand

Pirenne (1925) speculates that younger sons of peasants, who also were eradicated from the

land their families were attached to, may have originated within medieval cities the class of

merchants and artisans.

Finally, even though this model best captures the transition from feudalism to the tri-

umph of bourgeois power at the beginning of the 19th century, before formal suffrage legisla-

tion was introduced, our prediction of gradual but continuing democratization is confirmed

by subsequent political developments, with the process of formal extension of the franchise

starting in the 1830’s in England and France, a decade later in Germany (Bendix (1978)). By

the end of the same century mass democratization extended to the lower strata the political

voice that the bourgeoisie had gained earlier on.

7 Conclusion

This paper has established a connection between the transition of society from an aristo-

cratic political system into a democracy, the process of capital accumulation and sectoral

reallocation from agriculture into manufacturing, and the evolution of the system of inter-

generational transmission of property rights. The model replicates the historical stylized

facts of output growth, its sectoral composition and wealth distribution. It also provides a

39

framework to understand the evolution of class structure, the demise of landed property as

the sole source of power, and the emergence of the middle class.

Our model is based on the assumption of indivisibility of landed estates, which indeed

characterized Western Europe for several centuries. In the present paper, we do not yet

consider the eventual impact of democratization on this politico-economic constraint, which

will tend to lose its significance as a new polity, the modern nation-state, takes over the

tasks of external protection and internal administration. Once the political and economic

justification for primogeniture is removed, the breakup of great estates due to partition and

alienation will produce a fragmented ownership structure, promoting income and wealth

equalization. Eventually, the economy will evolve towards a standard factor-specific model

where the dynamics are determined by capital and the price of land, an altruistic bequest

motive induces partition of all the family’s wealth, and class stratification defined on the

specificity of wealth holdings disappears. The pressure of urbanization on land prices, the

expansion of capitalistic farming and the emergence of technological progress in manufac-

turing through innovation and education would constitute the next ingredients for future

investigation. Still, our present contribution can already help to capture the distinctive

agrarian roots of European civilization, which kept manifesting themselves until and even

after the advent of industrialization. The disproportional political weight of the agricultural

lobbies in the current European policy debate can only be understood within a framework

which explicitly considers the historical role of land and its rents.

APPENDIX

Proof of Lemma 1.

Note first of all that each individual’s indirect utility is monotonic in his income. Also

note that bLt > bWt has been established by Proposition 3, part (i). Finally, using expression

(4.12) for the equilibrium rent, Assumption 1 implies

[1 +τ(1− α)

α](1− τ)

1−αα > (

Σt

Γt)1−α (A.1)

40

By plugging the equilibrium expressions for factor prices (4.10)-(4.12) into yWt and yLt,

we can derive the following condition that ensures that yWt increases as the tax is abolished:

bWt <(1− α)(Σα

t − Γtα)

αNt(Γα−1t − Σα−1

t )(A.2)

while for yLt to decrease as the tax is abolished we need

bLt > (Γtα−1 − Σt

α−1)−1{Gt

[1 + τ(1−α)α](1− τ)

1−αα Γt

1−α − Σt1−α

(ΓtΣt)1−α− (1− α)(Γt

α − Σtα)

αNt}(A.3)

Define as βWt and βLt the levels of capital holding at which workers and landlords are

indifferent to tax policy, i.e., expressions (A.2) and (A.3) hold as equalities for bWt = βWt

and for bLt = βLt, respectively. Comparing (A.2) and (A.3) it follows that when (A.1) holds

then βLt < βWt. This implies that, for any bLt > bWt, conflicting tax preferences arise when

(A.2) and (A.3) hold, with workers preferring abolition and landlords retention.

Note that violation of either condition (i) or (ii) implies unanimity, with violation of

(i) implying that everybody loses from abolition and violation of (ii) implying that every-

body gains from it. If both are violated, we would have a “degenerate” form of conflicting

preferences, with workers losing and landlords gaining from abolition. ¤

Proof of Lemma 2.

From Lemma 1 we know that bWt < βWt is a condition for workers to prefer abolition.

But bWt > eb is also needed for them to vote. L + 1 workers need to vote to beat the L

members of the aristocracy. ¤

REFERENCES

Acemoglu, D. and Robinson, J., 2002, Economic Backwardness in Political Perspective,

NBER Working Paper No. 8831.

Acemoglu, D. and Robinson, J., 2000, Why Did the West Extend the Franchise? Democ-

racy, Inequality and Growth in Historical Perspective, Quarterly Journal of Economics

115, 1167-99.

41

Alston, L.J. and Schapiro, M.O., 1984, Inheritance Laws Across Colonies: Causes and

Consequences, Journal of Economic History 44, 277-287.

Becker, G.S., 1981, A Treatise on the Family, Harvard University Press, Cambridge.

Bendix, R.,1978, Kings or People: Power and the Mandate to Rule, University of California

Press, Berkeley.

Bergstrom, T.C., 1994, Primogeniture, Monogamy and Reproductive Success in a Stratified

Society, Working Paper 94-10, University of Michigan.

Bernheim, B.D. and Severinov, S., 2003, Bequests as Signals: An Explanation for the Equal

Division Puzzle, Journal of Political Economy 111, 733-764.

Bertocchi, G. and Spagat, M., 2001, The Politics of Co-optation, Journal of Comparative

Economics 29, 591-607.

Bertocchi, G. and Spagat, M., 2004, The Evolution of Modern Educational Systems: Tech-

nical vs. General Education, Distributional Conflict, and Growth, Journal of Develop-

ment Economics 73, 559-582.

Binswanger, H.P., Deininger, K. and Feder, G., 1995, Power, Distortions, Revolt and Re-

form in Agricultural Land Relations, in Behrman J. and Srinivasan, T.N., eds., Hand-

book of Development Economics, vol. 3B, 2659-2772, Elsevier, Amsterdam.

Boix, C. and Garicano, L., 2001, Democracy, Inequality and Country-Specific Wealth,

mimeo, University of Chicago.

Botticini, M. and Siow, A., 2003, Why Dowries?, American Economic Review 93, 1385-1398.

Bourguignon F. and Verdier, T., 2000, Oligarchy, Democracy, Inequality and Growth, Jour-

nal of Development Economics 62, 285-313.

Caselli, F. and Coleman, J.W.II, 2001, The U.S. Structural Transformation and Regional

Convergence: A Reinterpretation, Journal of Political Economy 109, 584-616.

42

Cervellati, M., Fortunato, P. and Sunde, U., 2005, Hobbes to Rousseau: Inequality, Insti-

tutions, and Development, IZA Discussion Paper, 1450.

Chu, C.Y.C., 1991, Primogeniture, Journal of Political Economy 99, 78-99.

Conning, J.H., 2002, Latifundia Economics, mimeo, Williams College.

Conning, J.H. and Robinson, J.A., 2005, Land Reform and the Political Organization of

Agriculture, Journal of Development Economics, forthcoming.

De Long, J.B., 2003, Bequests: An Historical Perspective, in Munnell A. and Sunden,

A. (eds.), Death and Dollars: The Role and Impact of Gifts and Estate, Brookings

Institution, Washington.

de Tocqueville, A., 1956, Democracy in America [1835], Penguin, New York.

Dewald, J., 1996, The European Nobility 1400-1800, Cambridge University Press, Cam-

bridge.

Doepke, M. and Zilibotti, F., 2005, Patience Capital and the Demise of the Aristocracy,

mimeo, UCLA and IIES.

Drazen, A. and Eckstein, Z., 1988, On the Organization of Rural Markets and the Process

of Economic Development, American Economic Review 78, 431-443.

Eaton, J., 1987, A Dynamic Specific-Factors Model of International Trade, Review of Eco-

nomic Studies 54, 325-338.

Ekelund, R.B. Jr., Hebert, R.F. and Tollison, R.D., 2002, An Economic Analysis of the

Protestant Reformation, Journal of Political Economy 110, 646-671.

Engerman, S. L., and Sokoloff, K. L. 1997, Factor Endowments, Institutions and Differential

Paths of Growth Among New World Economies: A View from Economic Historians

from the United States, in ed. by S. Haber, S. (ed.), How Latin America Fell Behind,

Stanford University Press, Stanford CA.

43

Faith, R.L. and Tollison, R.D., 2001, Inheritance, Equal Division and Rent Seeking, mimeo,

Arizona State University.

Falkinger, J. and Grossmann, V., 2004, Institutions and Development: The Interaction

Between Trade Regime and Political System, Journal of Economic Growth 10, 231-

272.

Galor, O. and Moav, O., 2006, Das Human Kapital: A Theory of the Demise of the Class

Structure, Review of Economic Studies 73, 85-117.

Galor, O. and Moav, O., 2004, From Physical to Human Capital Accumulation: Inequality

in the Process of Development, Review of Economic Studies 71, 1001-1026.

Galor, O., Moav, O. and Vollrath, D., 2002, Divergence and Overtaking: Land Abundance

as a Hurdle for Education Reform, Hebrew University, Department of Economics Work-

ing Paper No. 2002-11.

Galor, O. and Weil, D.N, 2000, Population, Technology and Growth: From Malthusian

Stagnation to the Demographic Transition and Beyond, American Economic Review

90, 806-828.

Goldstone, J.A., 1991, Revolution and Rebellion in the Early Modern World, University of

California Press, Berkeley.

Gollin, D., Parente, S. and Rogerson, R., 2002, The Role of Agriculture in Development,

American Economic Review 92, 160-164.

Goody, J., Thirsk, J. and Thompson, E.P. (eds.), 1976, Family and Inheritance: Rural

Society in Western Europe 1200-1800, Cambridge University Press, Cambridge.

Grossman, H.I., 1994, Production, Appropriation, and Land Reform, American Economic

Review 84, 705-712.

Habakkuk, H.J., 1955, Family Structure and Economic Change in Nineteenth-Century Eu-

rope, Journal of Economic History 15, 1-12.

44

Hansen, G.D. and Prescott, E.C., 2002, Malthus to Solow, American Economic Review 92,

1205-1218.

Horowitz, A.W., 1993, Time Paths of Land Reform: A Theoretical Model of Reform Dy-

namics, American Economic Review 83, 1003-1010.

Justman, M. and Gradstein, M., 1999, The Industrial Revolution, Political Transition,

and the Subsequent Decline in Inequality in 19-th Century Britain, Explorations in

Economic History 36, 109-127.

Kaldor, N., 1955, Alternative Theories of Distribution, Review of Economic Studies 23,

83-100.

Kuran, T., 2003, The Islamic Commercial Crisis: Institutional Roots of Economic Under-

development in the Middle East, Journal of Economic History 63, 414-46.

Kuznets, S., 1955, Economic Growth and Income Equality, American Economic Review 45,

1-28.

Labatut, J.-P., 1978, Les Noblesses Europeennes de la Fin du XVe a la Fin du XVIIIe

Siecle, Presses Universitaires de France, Paris.

Laitner, J., 2000, Structural Change and Economic Growth, Review of Economic Studies

67, 545-561.

Lewis, A.W., 1954, Economic Development with Unlimited Supplies of Labour, The Manchester

School 22, 139-191.

Lindert, P.H., 1986, Unequal English Wealth Since 1670, Journal of Political Economy 94,

1127-1162.

Lizzeri, A., and N. Persico, 2004, Why Did the Elites Extend the Suffrage? Democracy and

the Scope of Government, with an Application to Britain’s ‘Age of Reform’, Quarterly

Journal of Economics 119, 707-765.

45

Llavador, H. and Oxoby, R.J., 2005, Partisan Competition, Growth and the Franchise,

Quarterly Journal of Economics 120, 1155 - 1189

Maddison, A., 2001, The World Economy: A Millennial Perspective, OECD, Paris.

Marx, K., and Engels, F., 1975, Collected Works, Vol. 4 [1844-1845], International Pub-

lishers, New York.

Mookherjee, D. and Ray, D., 2002, Contractual Structure and Wealth Accumulation, Amer-

ican Economic Review 92, 818-849.

Moore, B.Jr., 1966, Social Origins of Dictatorship and Democracy: Lord and Peasant in

the Making of the Modern World, Beacon Press, Boston.

North, D.C., 1981, Structure and Change in Economic History, Norton, New York.

O’Rourke, K.H. and Williamson, J.G., 2005, From Malthus to Ohlin: Trade, Growth and

Distribution Since 1500, Journal of Economic Growth 10, 5-34.

Pirenne, H., 1925, Medieval Cities: Their Origins and the Revival of Trade, Princeton

University Press, Princeton.

Pirenne, H., 1963, Histoire Economique et Sociale du Moyen Age, Presses Universitaires de

France, Paris.

Platteau, J.-P. and Baland, J.-M., 2001, Impartible Inheritance Versus Equal Division: A

Comparative Perspective Centered on Europe and SubSaharan Africa, in de Janvry,

A., Gordillo, G., Sadoulet, E. and Platteau, J.-P. (eds.), Access to Land, Rural Poverty,

and Public Action, Oxford University Press, Oxford.

Smith, Adam, 1937, An Inquiry into the Nature and Causes of the Wealth of Nations [1776],

The Modern Library, New York.

Spring, E., 1993, Law, Land, and Family: Aristocratic Inheritance in England, 1300 to

1800, University of North Carolina Press, Chapel Hill.

46

Volckart, O., 2000, The Open Constitution and its Enemies: Competition, Rent Seeking,

and the Rise of the Modern State, Journal of Economic Behavior & Organization 42,

1—17.

Werner, K.K., 1998, Naissanse de la Noblesse: L’ Essor des Elites Politiques en Europe,

Librairie Artheme Fayard, Paris.

47