Elite Competition, Co-option and the Iron Law of Oligarchy: Theory and a Tale of 14 Islands

Elite Competition, Co-option and the Iron Law of Oligarchy: Theory and a Tale of 14 Islands Jean Paul Carvalho Christian Dippel March 25th 2016 Abstract Under the iron law of oligarchy, elites are inescapably repressive. Existing models support this view when there is a unified elite disciplined only by popular revolt. We develop a model of oligarchy in which the elite is disaggregated into factions with distinct economic and political incentives. Elite competition can bend the iron law of oligarchy in various ways. Firstly, elite factions may have different economic incentives. Landowners prefer labor coercion, but merchant elites may not. Secondly, new elites may have political incentives to maintain legitimacy among the citizenry, regardless of their economic interests. We characterize conditions under which the old elite can co-opt new elites into supporting repressive policies. We illustrate the model s implications using detailed historical data on the plantocracies of 14 British Caribbean sugar islands in the 19th century, where the emancipation of slaves created new elite factions with distinct economic and political interests. We thank Melissa Dell, Alan Dye, Ruben Enikolopov, Fred Finan, Price Fishback, Gary Libecap, Adam Przeworski, Ken Shotts, Francesco Trebbi, Dan Trefler, Stephane Wolton, and Alex Whalley for comments on the early stages of this project. Jake Kantor, Vero Rogers-Thomas, and Tessa Seager provided excellent research assistance. Financial support from UCLA s Burkle Center, Center for Economic History and Price Center are gratefully acknowledged. University of California, Irvine. University of California, Los Angeles, and NBER.

If we want things to stay as they are, they will have to change. Giuseppe di Lampedusa, The Leopard In few places does the dead hand of the past lie as heavily on the present as in the Caribbean. Wallace (1977) 1 Introduction A rich political economy literature exists on the expansion of democracy: how elite-dominated oligarchic societies politically liberalize, either in the form of autocracies that transition into democracy or in the form of narrow-based democracies becoming more broad-based (Lipset (1959), Bourguignon and Verdier (2000), Acemoglu and Robinson (2000), Acemoglu and Robinson (2001), Lizzeri and Persico (2004), Brückner and Ciccone (2011), Besley, Persson, and Reynal-Querol (2012), Bidner, Francois, and Trebbi (2014)). Democratic expansion or contraction is determined by exogenous fluctuations in the power of the elite and non-elite citizenry. In an influential line of research, Acemoglu and Robinson (2000), Acemoglu and Robinson (2001), Acemoglu and Robinson (2006b) argue that democratic expansion occurs when the citizenry is temporarily empowered, and attempts to lock in this power by demanding permanent institutional changes. In the formal literature on oligarchy, elite survival and democratic transitions, society is divided into two monolithic blocs, a unified elite and a citizenry, the the elite disciplined primarily by the threat of revolt by the citizenry. 1 This has proved to be a simple and powerful workhorse model, but leaves open questions about the role of elite composition in repression and revolution. This is important because a great deal of social and political change has historically been driven by intra-elite competition. When the elite is composed of factions with distinct political and economic interests, under what conditions do these factions strike a bargain and impose repressive policies in a unified fashion? Under what conditions does intra-elite conflict break out and result in less repressive policies? How does the precise composition of the elite (e.g. landowners versus merchants) matter? 1 In the Meltzer and Richard (1981) framework, political competition is the threat of having a non-elite median voter set policies. The more unequal a society is, the further apart are the elite s and non-elite s policy bliss points. 1

Building on Acemoglu and Robinson (2008) and its antecedents, we develop a model of oligarchy subject to elite competition. As in this literature, the population is partitioned into an elite and citizenry whose interactions determine political and economic institutions. Our point of departure is the assumption of a unified elite. Concretely, we divide the elite into landowners and merchants. To analyze the impact of changing elite composition, we further disaggregate the elite into old and new factions. Specifically, we let the elite be composed of an old landowner elite, a new landowner elite and a new merchant elite. Economic institutions are determined by the interplay between these three elite factions. As we shall see, intra-elite competition is not independent of the threat of citizen revolt the usual mechanism for disciplining the elite but interacts with it in interesting ways. The different elite factions have distinct economic and political incentives. The landowning factions, both old and new, have an economic interest in repressive labor policies which lower wages in agriculture. The new merchant elite may have different economic interests and may need to be compensated by the old elite for supporting labor repression. In addition, political considerations may drive a wedge between even old and new landowning elites. By virtue of their proximity to the people, the new elite factions, both landowning and merchant, are deemed legitimate by the citizenry unless they support labor repressive policies. The old elite is always illegitimate. In the event of a citizen revolt, all illegitimate elite factions are expropriated. Hence both new elite factions will need to be compensated by the old elite for supporting labor repression and relinquishing their popular legitimacy. The old elite is more willing and able to purchase the new elite s support for labor repressive policies, when the agricultural sector is highly profitable relative to the non-agricultural sector, when labor repression increases agricultural profits considerably but does not reduce merchant profits, and when the likelihood of citizen revolt is low. This co-optive equilibrium conforms to the Iron Law of Oligarchy, which was proposed by Michels (1911) and which Acemoglu and Robinson (2006a) summarize as follows: The reason for persistence is therefore not persistence of the elites, but the persistence of incentives of whoever is in power to distort the system for their own benefit. 2 2 Michels said: Even when the discontent of the masses culminates in a successful attempt to deprive the bourgeoisie of power, this is effected only in appearance; always and necessarily there springs from the masses a new organized minority which raises itself to the rank of a governing class. 2

However, the Iron Law is a special case in which the old elite is able to align the incentives of the new elite factions to its own and create a de facto unified elite. The natural divergence in the political and economic interests of different elite factions means that the Iron Law can be bent in various ways. For example, a rise in the likelihood of popular revolt heightens the importance of legitimacy to the new elite, making them more expensive to co-opt. If the new elite becomes prohibitively costly to co-opt, a contest ensues in which the old elite supports a labor repressive policy and the new elite factions oppose it. If the labor repressive policy is defeated, then the relative income of the old landowning elite falls, making co-option even more difficult in the next period and rendering the merchant elite more likely to win the contest over policy. That is, a decline in political power of the old elite reduces their economic power, which leads to a further decline in their political power. In this way, the threat of popular revolt can trigger elite competition, produce endogenous changes in the power of the different elite factions and lead to less repressive policies. We extend the theory to allow elites the last resort option to invite a military coup or the intervention of a foreign power to halt the slide in their power. This last resort is of particular real world importance as there are many examples in which elite factions have invited such intervention to regulate intra-elite competition. For example, Thailand s May 2014 military coup has been characterized as the culmination of months of maneuvering by the Bangkok establishment to sideline the populist movement that has won every national election since 2001. The establishment was seeking to suspend democracy, at least in the short term, as it struggled to unseat a ruling party it has been unable to defeat at the polls (Fuller (2014)). In Sri Lanka, independence from the British in 1948 unleashed new political forces, including a left wing Buddhist nationalist movement. As the old Christian elite of the country faced extinction from power, several members of it orchestrated an ultimately abortive coup in January 1962. As we shall see, the option of inviting foreign intervention was particularly important for old elites in our empirical application. Adding this last resort also incorporates into the theory the flip-side of new elites higher legitimacy, namely a higher degree of accountability to the citizenry. This accountability arises naturally if new landowning elites benefit less from foreign intervention than old landowning elites, e.g. because old elites have better connections to the military or foreign power that allows them to protect their wealth and influence under the new regime. 3

We illustrate the implications of our theory in uniquely granular data on elites in the 14 British Caribbean plantation colonies after the end of slavery. There is a deep appeal in studying the 19th century Caribbean as a set of case studies for the dynamics of elite survival. One reason is that each of the 14 islands was small enough that researchers can can get a uniquely comprehensive picture of elites, both economic and political, and the dynamics of their changing identity and coherence. A second reason is that they experienced truly tumultuous exogenous shocks: In 1833, when London imposed the emancipation of slaves throughout the Empire, it turned over 90% of the British Caribbean s population into free citizens, giving rise to a new elite representing the freedmen. A third reason is that the islands oligarchic organization survived these challenges, and did so without elites resorting to the type of repression of freedmen associated with the postreconstruction U.S. South. Craton notes how throughout the 19th century, each major inquiry [by English Parliament] into the British West Indies noted with amazement that nothing had been changed since the last report (1988, p. 165). We present evidence that the Caribbean planter elites achieved this stability through a successful strategy of co-opting the new elites leaders. A fourth and final reason is that the persistence in the de facto organization of society was not achieved with co-option alone. When co-option became too difficult and the iron law fractured, elites chose as a last resort to make fundamental changes to the de jure institutions. The most salient such institutions were the locally elected parliaments ( assemblies ), which eleven of the 14 islands had on the eve of emancipation, and which constituted the main vehicle of elite power. All of them survived the first two decades after emancipation untouched, but then ten of the eleven voluntarily abolished themselves inside two decades from 1855 1876, effectively inviting the Crown to run the government as the colonial equivalent of a foreign power. Our empirical focus is to document the co-option of new elites and to analyze through the lens of the theory how the eventual fracturing of the iron law led to the response of de jure institutional changes. We take two different approaches in our empirical analysis. For the two largest colonies Barbados and Jamaica we were able to obtain the roll-call votes of the assemblies as well as a fairly complete picture of the identity of the assemblymen, allowing us to distinguish those representing old planter elites from new planter elites and new merchant elites. Aside from being economically the most important colonies, Barbados and Jamaica also illustrate two of the most important equilibrium paths in the theory. Barbados was an outlier among the Caribbean slave so- 4

cieties in that its entire land area was highly plantation-suitable land, meaning that old wealth was more secure and that both old and new wealth were tied to the plantation system. Consequently, it had less entry by new elites, new elites were primarily planters, and all new elites exhibited voting patterns that were no different from the planters, suggesting full co-option. In Barbados, thus, the iron law of oligarchy held sway, and it consequently remained the only one of the 14 islands that had not abolished its parliament by 1900. By contrast, more than half of Jamaica s land, while fertile, was unsuitable for plantations, giving rise after emancipation to significant economic interests that were not naturally aligned with the planters. Indeed, we see in the decade from 1844 a period of significant polarization in Jamaica s politics, in which new merchant elites formed a sizeable and distinct voting bloc. However, this was followed by a decade of renewed co-option after 1853, in which the non-planter bloc, despite increasing in size, was co-opted into weakening its opposition to the planters. Finally, Jamaica abolished its parliament as an immediate consequence of a violent revolt, the 1865 Morant Bay rebelion. Viewed through the lens of our theory, Morant Bay put pressure on Jamaica s new elites to re-gain legitimacy with the citizenry by ending co-option. This made co-option prohibitively expensive so that elites chose the last resort option to abolish their assembly and cede power to the Crown. 3 Our second empirical approach focuses on the full set of islands, where we do not have the roll-call data to observe elite alliances directly. Instead we focus on the relative power of old and new elites over time. Using a collection of 61 plantation surveys at different times between 1815 1891, laboriously assembled from 43 distinct sources, we partition assemblies into the seat shares of old and new elites, and show that the colony-specific timing of assemblies abolishing themselves is explained by a steep drop in the representation of old elites. To gain identification, we focus on an important Caribbean-wide law change in 1854, when the Caribbean Encumbered Estate Act (EEA) imposed a change in bankruptcy law that dramatically increased turnover in plantation ownership, replacing old (white) planter elites with new (creole and colored) planter elites. The EEA allows us to isolate the role of elite identity and composition in institutional change, because it occurred at a time when the economic circumstances of the plantation system, i.e. the terms of trade and the agricultural production mix, were very stable. In the plantation 3 The option of outside intervention in the theory can mean an intervention by a foreign military power or by a domestic military. In the Caribbean context, it meant the abolition of the assembly to be replaced by Crown Rule from London. 5

data, we show that the EEA was associated with a significant increase in the turnover of plantation ownership. In the assembly data, it was associated with a significant drop in old elites share of seats. Overall, the introduction of the EEA appears to have provided the immediate impetus for the wave of institutional changes that washed through the British Caribbean from the mid-1850s to the mid-1870s, by making co-option prohibitively expensive for old elites. We conclude with a qualitative description of the events surrounding the final vote to abolish the assembly in several of the islands, based on their Hansards or Assembly Minutes. This paper speaks to a well-established literature on the persistence of oligarchy and the eventual transition to democracy (Haggard and Kaufman (1997), Bourguignon and Verdier (2000), Przeworski, Alvarez, Cheibub, and Limongi (2000), Acemoglu and Robinson (2000), Acemoglu and Robinson (2001), Lizzeri and Persico (2004), Acemoglu and Robinson (2006b), Brückner and Ciccone (2011)) It extends this literature by modeling the persistence of oligarchy when the elites are not monolithic. By putting special emphasis on the possibility that elites can invite an intervention or a coup, it also ties the work in Acemoglu and Robinson (2008) on de jure and de facto institutions more closely to that literature. More broadly, this paper contributes to a literature on the nexus between endogenous institutional change, elite persistence, and economic growth (North and Weingast (1989), Acemoglu, Johnson, and Robinson (2005), Trefler and Puga (2014)). The paper is structured as follows. Section 2 introduces and analyzes a model of oligarchy under elite competition and co-option. Section 3 provides a historical account of the post-emancipation British Caribbean, and relates this account to our theory. In section 4, we analyze alliances between old planters, new planter elites and new merchant elites in Jamaica s and Barbados assemblies. In section 5, we study the erosion of old planter elites across the islands and the timing of the abolition of assemblies. Section 6 concludes. 2 The Model Society consists of a finite set of individuals N, with typical member i, partitioned into four groups: the old (landowning) elite E, the new landowning elite N 1, the new merchant elite N 2 and the citizenry C. The size of the elite factions are n E, n 1 and n 2 respectively, where n E +n 1 +n 2 = n. Each elite faction acts in a unified manner to maximize its own members welfare. The old elite 6

and new landowning elite share economic incentives, both benefitting from labor repression. The new merchant elite may have different economic incentives, but could still be co-opted by the old elite to support labor repressive policies. The citizenry is a non-strategic actor. With some likelihood, it revolts and expropriates the elite factions that supported labor repression. This gives both the new landowning and merchant elites political incentives to oppose repressive policies. Interactions occur over an infinite horizon. Time is discrete and denoted by t = 1, 2... Individuals have a common discount factor δ (0, 1). The state of the economy in a given period t is denoted by λ t {0, 1}, where λ t = 1 indicates that a labor repressive policy is in place. The income of each member of the landowning elite i E N 1 is state dependent and denoted by y 1 ( λ t ). Let y 1 ( 0 ) = y1 and y 1 ( 1 ) = y1 + 1, where 1 > 0 so that landowners benefit from labor repression. The income of each member of the merchant elite i N 2 is denoted by y 2 ( λ t ). Let y 2 ( 0 ) = y2 and y 1 ( 1 ) = y2 2, where 2 measures the degree to which the economic incentives of the landowning and merchant elites are aligned. We assume that 2 0, so that the merchant elite bears an economic cost from labor repressive policies. Let labor repression be in force in period 1: λ 1 = 1. Thenceforth, the state λ t is determined endogenously. The following interactions occur in each period t: At the beginning of the period, income is received based on the state λ t, i.e. the labor policy in place. Each member of the landowning elite i E N 1 receives income of y 1 ( λ t ) and each member of the merchant elite i N 2 receives income of y 2 ( λ t ). The old elite E then offers two contracts, one to the new landowning elite and one to the merchant elite. Under a given contract, the old elite pays the new landowning (resp. merchant) elite a sum of π1 t (resp. πt 2 ) per member if it supports implementation of the labor repressive policy in the next period t + 1. Otherwise, the old elite pays nothing. The old elite, by virtue of its participation in past labor repression is always considered illegitimate by the citizenry. Along with its economic interest in labor coercion, this means that the old elite always support the labor repressive policy. Support for labor repressive policy by new elite faction k is denoted by x t k = 1. Opposition to labor repression is denoted by xt k = 0. If a new elite faction supports labor repression, we refer to it as being co-opted. Subsequently, the citizenry revolt with probability r (0, 1). A new elite faction k {1, 2} is 7

consider illegitimate in period t only if it supports labor repression, x t k = 1. If revolt occurs, every member of an illegitimate elite faction is expelled from the elite in all future periods τ > t, being replaced by a member of the citizenry. In addition, there is also no labor repression in the next period: λ t+1 = 0 with probability one. If revolt does not occur, the next period s labor policy is determined by support for and opposition to labor repression by the various elite factions. If both new elite factions support the labor repressive policy, (x t 1, xt 2 ) = (1, 1), then it is implemented in the next period λt+1 = 1 with probability one. 4 If either faction of the new elite N 1 or N 2 opposes the labor repressive policy, a contest over next period s policy ensues. If exactly one new elite faction supports labor repression, then λ t+1 = 1 with probability p 1. If no new elite faction supports labor repression, then λ t+1 = 1 with probability p 2, where 0 < p 2 < p 1 < 1. After co-option payments are made, all members of the elite spend their entire current period income net of transfer payments on g units of a unique non-storable good with unit price. We assume that consumption utility is linear. We restrict attention to pure strategies. The action profile in a given period is denoted by a t = (π1 t, πt 2, xt 1, xt 2 ) and consists of a contract choice by the old elite and policy choices by the new elite. Let the history at time t be h t = (λ 0, λ 1, a 1,... λ t 1, a t 1, λ t ). A subgame perfect equilibrium consists of mapping from the space of histories to the space of action profiles a(h t ). A Markov Perfect equilibrium (MPE) requires in addition that a(h t ) = a( h t ), whenever λ t is the same in histories h t and h t. That is, MPE strategies are solely a function of the current state λ. The value function for a member of elite faction k is V k (λ) = g k ( a t, λ t) + δ ( 1 r(λ)1{x k (λ) = 1} )[ P (a t )V k (1) + ( 1 P (a t ) ) V k (0) ], (1) ( where consumption g k.) equals the income of faction k members net of co-option payments and P (a t ) is the probability that λ t+1 = 1. 4 As is common in the literature, voting institutions are not explicitly modeled, (e.g. Acemoglu and Robinson, 2008, 2006a). 8

2.1 Full Co-Option We shall now examine the conditions under which the Iron Law of Oligarchy holds with the old elite co-opting the new landowning and new merchant elites into supporting labor repression. Specifically, we construct an MPE in which both new elite factions support the labor repressive policy in all future periods, i.e. ( x t 1, 2) xt = (1, 1) for all t > 1. Recall that the old elite offers a pair of contracts, one to the new landowning elite and one to the new merchant elite. In an MPE, the contract specifies payments as a function of the current state λ. If new elite faction k supports labor repression in state λ, x k (λ) = 1, then the old elite pays each member of N k the sum π k (λ). The cost is divided evenly among the old elite N E. If new elite faction k, opposes labor repression in state λ, x k (λ) = 0, then members of N k receive nothing under the contract. The optimal way to induce support is for the old elite to set π k (λ) such that new elite faction k is indifferent between supporting and opposing labor repression in state λ, for k = 1, 2 and λ = 0, 1. The question then arises as to whether it is worthwhile for the old elite to commit to this contract and induce acceptance? In other words, is co-option too costly for the old elite? Suppose not. Then the contract ( π 1 (0), π 1 (1), π 2 (0), π 2 (1) ) is offered, both new elite factions support labor repression in each state and λ t = 1 with probability one for all t > 1. In this case, the payoff to each member of the new landowning elite i N 1, in state 0 and 1 respectively, can be written as V 1 (0) = y 1 (0) + π 1 (0) + δ(1 r)v 1 (1), (2) V 1 (1) = y 1 (1) + π 1 (1) + δ(1 r)v 1 (1). (3) Solving this system: V 1 (0) = y 1 (0) + π 1 (0) + δ(1 r) y 1(1) + π 1 (1) 1 δ(1 r) V 1 (1) = y 1(1) + π 1 (1) 1 δ(1 r). (5) (4) A one-shot deviation to opposing the labor repressive policy, x t 1 = 0, has countervailing political and economic effects. By reducing the probability that labor repression prevails in the next 9

period from one to p 1, opposition reduces the expected future income of new landowners. On the other hand, it enables the new landowning elite to maintain legitimacy among the citizenry. The payoff from a one-shot deviation to x t 1 = 0 when the state is λ is y 1 (λ) + δ [ p 1 ( 1 r ) V1 (1) + ( 1 p 1 (1 r) ) V 1 (0) ]. (6) Recall that ( π 1 (0), π 1 (1) ) is chosen to make the incentive compatibility constraints bind. Hence, V 1 (0) = y 1 (0) + δ [ p 1 ( 1 r ) V1 (1) + ( 1 p 1 (1 r) ) V 1 (0) ], (7) V 1 (1) = y 1 (1) + δ [ p 1 ( 1 r ) V1 (1) + ( 1 p 1 (1 r) ) V 1 (0) ]. (8) Solving this system: V 1 (0) = 1 δp ( ) 1 1 r y 1 (0) + δp ( ) 1 1 r y 1 (1), (9) 1 δ 1 δ V 1 (1) = δ( [ ]) 1 p 1 1 r y 1 (0) + 1 δ( [ ]) 1 p 1 1 r y 1 (1). (10) 1 δ 1 δ Recall that y 1 (0) = y 1 and y 1 (1) = y 1 + 1. Substituting into (9) and (10): V 1 (0) = y 1 1 δ + δp ( ) 1 1 r 1, 1 δ (11) V 1 (1) = V 1 (0) + 1. (12) These expressions for the value functions are implied by the incentive compatibility conditions, when they bind. The value functions must also satisfy (4) and (5). Substituting into (11) and (12) and rearranging yields expressions for the co-option payment to the new landowning elite: π 1 (λ) = π 1 = δ 1 δ r(y 1 + 1 ) δ ( ) 1 δ(1 r) [1 p1 (1 r)] 1, (13) 1 δ which is the same in each state λ = 0, 1. 5 Whenever non-negative, the co-option payment to the new landowning elite is given by (13). 5 This is due to the consumption sub-utility function being linear and the transition function P (a t ) not depending directly on the state. 10

The first term in this expression captures the political incentives of the new landowning elite, representing the cost to it of losing legitimacy and putting at risk its income in the event of citizen revolt. The greater the likelihood of revolt r, the larger is this cost and the more the new landowning elite must be compensated for supporting the labor repressive policy. The secondterm captures economic incentives, representing the benefit in terms of incremental income from agriculture 1 > 0 under labor repression. When the likelihood of revolt r is low, the economic benefit of supporting labor repression is large relative to the political cost and the new landowning elite is willing to support the labor repressive policy without compensation. This does not apply to the new merchant elite whose members bear both a political and economic cost from supporting labor repression. By the same process the co-option payment to each member of the new merchant elite i N 2 is given by π 2 = δ 1 δ r(y 2 + 2 ) + δ ( ) 1 δ(1 r) [1 p1 (1 r)] 2, (14) 1 δ in each state λ = 0, 1. The difference here is that π 2 is always positive. Now that we have identified the optimal contracts that induce support for labor repression by the new elite factions, we need to check whether the benefits of co-option for the old elite exceed the costs. In a full co-option equilibrium, the payoff to each member of the old elite i N E in state λ can be written as V E (λ) = y E (λ) n 1 n E π 1 n 2 n E π 2 + δ(1 r)v E (1). (15) A one-shot deviation to offering no payment to the new merchant elite N 2 in state λ yields the following continuation payoff y E (λ) n 1 n E π 1 + δ(1 r) [ p 1 V E (1) + (1 p 1 )V E (0) ]. (16) Hence there is no profitable one-shot deviation in state λ if and only if δ(1 r)(1 p 1 ) [ V E (1) V E (0) ] n 2 n E π 2. (17) 11

Notice that this condition is independent of the state. Computing V E (1) V E (0) from (15) and substituting expression (14) for π 2, yields the first incentive compatibility constraint for the old elite: δ(1 r)(1 p 1 ) 1 n 2 δ n E 1 δ r(y 2 + 2 ) + n ( ) 2 δ 1 δ(1 r) [1 p1 (1 r)] 2. (18) n E 1 δ Now consider a one-shot deviation to offering no payment to the new landowning elite N 1. The corresponding incentive compatibility constraint for the old elite is δ(1 r)(1 p 1 ) 1 n 1 δ n E 1 δ r(y 1 + 1 ) n ( ) 1 δ 1 δ(1 r) [1 p1 (1 r)] 1. (19) n E 1 δ Finally, consider a one-shot deviation to offering no payment to either new elite faction. The corresponding incentive compatibility constraint is δ(1 r)(1 p 2 ) 1 n [ 1 δ n E 1 δ r n1 (y 1 + 1 ) + n ] 2 (y 2 2 ) n E n E ( ) 1 δ(1 r) [1 p1 (1 r)] n 1 n E δ 1 δ [ n1 1 n ] 2 2. n E n E (20) A full co-option MPE is one in which x k (λ) = 1 for both new elite factions k = 1, 2 in both states λ = 0, 1. Conditions (18)-(20) are necessary and sufficient for the existence of such an equilibrium. The left-hand side of each one of these incentive compatibility constraints is the benefit from cooption in terms of a higher likelihood of labor repression and higher agricultural income in the next period. The right-hand side is the cost of co-option in terms of payments to the new elites. It is straightforward to characterize the conditions under which these constraints are satisfied. The results are stated in the following proposition. Proposition 1 There exists a full co-option Markov Perfect equilibrium if and only if, other things being equal, (i) The gain to landowning elites from labor repression 1 is sufficiently large, (ii) The cost to the new merchant elite from labor repression 2 is sufficiently small, 12

(iii) The old elite makes up a large proportion of the elite, n E /n is sufficiently large, (iv) The likelihood of revolt r is sufficiently small. Therefore, the Iron Law of Oligarchy is not guaranteed to hold when the elite is composed of different factions with distinct economic and political incentives. Several conditions must be satisfied. First, the gain to landowning elites from labor repression 1 must be large, otherwise the old elite will not be willing to be bear the cost of co-option. Second, the new merchant elite must be compensated for the cost to it from labor repression 2. Hence 2 must be sufficiently small for the old elite to be willing to co-opt the new elite. Third, when the old elite is relatively large it needs to co-opt fewer individuals and the cost of co-option is spread over a larger number of members. Hence it is more willing to pay the cost of co-option. Finally, when the likelihood of citizen revolt r is large, legitimacy is an overriding concern for the new elites making them prohibitively costly to co-opt. In addition, the expected benefit from co-option is small, as the old elite is likely to be expropriated during a revolt in any case. When these conditions fail, only partial or no co-option equilibria exist and the Iron Law of Oligarchy breaks down. 2.2 Partial Co-Option We shall now examine the conditions under which a partial co-option equilibrium exists, in which the old elite co-opts the new landowning elite but not the new merchant elite. (To be completed) 2.3 The Last Resort Option: Coups and Foreign Intervention By Proposition 1, both an expansion in the share of new elites and a higher likelihood of citizen revolt can cause co-option to break down and erode rents enjoyed by the old elite. To halt its slide, the old elite can invite an external entity (e.g. domestic military, colonial power) to intervene in order to reduce the likelihood of revolt and consolidate its position. (To be completed) 3 The British Caribbean Plantation Colonies After Emancipation Table 1 shows the 14 islands we study. They were founded in three waves. The oldest Antigua, Barbados, Jamaica, Montserrat, Nevis, St. Kitts, and the Virgin Islands were founded in the 13

1600s by small-scale British planters, much like those in the Chesapeake Bay. The second wave Dominica, Tobago, St. Vincent, and Grenada were annexed from France at the end of the Seven Years War in 1765, and were resettled by sugar planters from the other Caribbean islands The last three colonies Trinidad, St. Lucia, and Guyana were ceded to Britain between 1797 and 1803. Sugar was introduced into the Caribbean around 1700, and with it the great planters came to dominate (Taylor (2002, ch. 11)). By 1800, these great planters had become the wealthiest men in all of English America (Galloway (2005)). The planters political power found its expression in large part in their control of the local legislatures, the Assemblies, which the Caribbean s original smallhold settlers had pushed for in the 17th and 18th century to better control taxes and spending (Taylor (2002, p. 246)). These assemblies were powerful institutions, which in addition to their legislative functions, had extensive executive powers. Colonial Acts assigned all important administrative tasks to special boards, or commissions, upon which members of the assembly enjoyed either exclusive or majority control (Green (1991, p. 68)). Morrell and Parker assert that the assemblies seriously curtailed the powers of the governors in the colonies (1969, p. 435), and Lewis write of the Crown s incentives that to join with local white society meant a pleasant tour of duty, to fight them meant political conflict and social ostracism. Inevitably, the governor passed smoothly into the union, political and social, of government and vested interests (2004, p. 104). 6 From the peak of their powers in 1800, however, the tide was turning against the great planters. Slavery, which was critical to Caribbean wealth, came under increasing attack from the rising Abolitionist movement in London. In 1807, British parliament abolished the slave trade, dramatically increasing labor costs in the Caribbean plantations. In 1833, British parliament passed An Act for the Abolition of Slavery, which emancipated slaves throughout the British Empire in 1838. The old planter elites dominance of the assemblies continued in the early years after emancipation: A few merchants, lawyers, and medical practitioners secured seats in the Jamaica Assembly before 1840, but planters dominated colonial government in the thirties and forties. Barbados 6 Not only could they veto the budget including the governors salary, elites were not afraid of using the assemblies to publicly ostracize Crown administrators that over-stepped their boundaries: When Henry Barkly, the newly appointed Governor of Jamaica, dispatched a special message to the Assembly suggesting the postponement of a legislative action, he was publicly ignored and privately reproached. Had he not been new to the office and popular with assemblymen, he would have been declared in breach of privilege, and the Assembly would have suspended its business until Barkly had submitted an appropriate apology. (Green, 1991, p. 70) 14

Table 1: 14 Islands & Abbrev. & Year & Year of & Population & Pop-Share & Density & Colony & & Founded & Regime Change & 1836 & White, 1836 & 1836 & Antigua & ATG & 1632 & 1867 & 35,188 & 5.4 & 125 & Barbados & BRB & 1629 & never & 105,812 & 12.8 & 246 & Dominica & DOM & 1763 & 1862 & 16,207 & 3.9 & 21 & Grenada & GRD & 1763 & 1876 & 17,751 & 2.6 & 52 & Jamaica & JAM & 1655 & 1860 & 381,951 & 8.2 & 34 & Montserrat & MON & 1634 & 1861 & 6,647 & 4.3 & 65 & Nevis & NEV & 1623 & 1867 & 7,434 & 5.4 & 80 & St. Kitts & STK & 1628 & 1867 & 21,578 & 6.4 & 113 & St. Vincent & STV & 1763 & 1866 & 26,659 & 4.7 & 69 & Tobago & TOB & 1763 & 1873 & 11,456 & 2.3 & 38 & Virgin Islands & VIR & 1672 & 1855 & 7,471 & 12.4 & 49 & Guyana & GUY & 1803 & n/a & 66,561 & 0.7 & 6 & St. Lucia & SLU & 1803 & n/a & 17,005 & 11.3 & 27 & Trinidad & TRI & 1797 & n/a & 34,650 & 8 & 7 & Notes: The Data in this table are from the Colonial Blue Books and from Martin (1839). Year of regime change is not applicable for the 3 late colonies, which were founded as Crown Colonies. Year founded refers to the year a place became a British colony. The only non-island, Guyana, had very big hinterlands, so we use only its extent as it was surveyed in the 19th century. See online appendix figure 2. merchants petitioned that they were totally unrepresented in their Assembly. In 1837, twenty-two of twenty-five Antigua Assembly were planters (Green, 1991, pp.73). However, this dominance was soon challenged. Before emancipation it had been, throughout the Caribbean, distinctly the exception for a member of the legislature to be returned by more than 10 votes (Wrong (1923, p. 69)). However, this narrow franchise was driven by the small number of free people and not by particularly stringent franchise rules. In fact, the property holdings required to vote were actually quite low, having typically staying at their original 10-acre threshold from the Caribbean s smallhold days (Wrong (1923)). 7 Because the price of 10 acres of land was easily within the purview of what a smallholder could save up, the old planters; dominance of the assemblies was sure to erode over time. Indeed, smallholders established themselves quickly in the islands, Higman even writes of a spectacular growth in the extent of smallholding after 1838 (2001). While smallholding did not generate the wealth required to run for office, it did generate the wealth required to purchase 10 acres and thus acquire the right to vote, which provided the impetus for a new class of colored politicians in the islands. Holt (1991, p. 244) recounts how in Jamaica, in 1844, governor Elgin called for early elections to blunt the registration drive [of smallholders who had obtained 7 The franchise in the Caribbean, as elsewhere in the British colonies, was obtained primarily through land ownership. The franchise rules reported in the Blue Books show that there is very little variation in the 10-acre threshold across colonies. 15

10 acres of land]. Yet, when the new assemblymen convened, there were five new colored faces among them. Similarly, Lewis (2004, p.67) states that throughout the Caribbean a brief generation after Emancipation both black and coloured men had obtained positions of prominence. As a result, Green (1991, p.296) writes that although whites continued to dominate society in most colonies, the rigid barriers that had divided them from coloured people were eroded in the free period. In Dominica and Montserrat coloured men quickly assumed a dominant role in the legislature. They were a powerful element in Jamaica. [...] Increasingly, men of colour acquired plantation property. Both the emergence of new elites and their partial co-option by the old elites thus appear to have been salient features of the post-emancipation Caribbean. And the economic incentives of the new elites appear to have determined the degree of co-option across islands. In Tobago, where she gives many examples of successful black and colored planters, Craig-James argues that the planter oligarchy was, over time, no longer almost exclusively white. [...] The whites and coloreds in the dominant class had ties of kinship, friendship, interests and predicaments. The members of the dominant class including the most established of the black planters, attended the governor s balls, their wives and daughters pillars of the Church of England (2000, p199-201). But while some of the new elites were also planters, many were not: 8 In Dominica, Honychurch (1995, p.69-71) speaks of two unofficial parties, with the Mulatto ascendancy firmly in opposition to the white planters, and pressing for legislation promoting the welfare of the newly liberated citizens of the island. 9 Co-option was thus not the only possible outcome, and the iron law of oligarchy was just one possible equilibrium path. Our theory predicts that the co-option of new elites may has a particular life-cycle: Co-option can work for a while but then fail when the economic incentives of the new elites begin diverging from those of the old elites, when the new elites become too big and expensive to co-opt, or when there is an exogenous threat of revolt by the citizens which requires 8 [A] new petit bourgeois class emerged, consisting of merchants, successful estate owners without the ancestry and traditions of the older landed cls members of the professions, and an expanding managerial sector. This class was far more heterogeneous than the class it was gradually displacing in economic and political affairs. (Meditz, 1987, p.31) 9 This division appears to have been most pronounced in regards to public goods provision: Planters generally opposed all measures to expand education. Very likely the idea of spending money primarily for the benefit of the black majority did not appeal to most planters. Most of the white estate managers had no family or children, at least none they chose to recognize officially. The wealthier resident planters sent their children to a few select private academies on the island and to England (Holt (1991, p.196)). 16

the new elites to strengthen its legitimacy. In that case, a transition to broad-based democracy is one possible outcome. However, when elites have the outside option of inviting a foreign intervention, they may prefer to do that instead. And indeed, between 1855 and 1877, 10 of the 11 Caribbean parliaments pursued the colonial equivalent of inviting a foreign intervention: they invited the Queen to dissolve them and the Crown to write a new constitution for them, with a legislature appointed by London. 10 In some cases, this appears to have been supported by the new elites. Rogers for instance argues that fear of political displacement by the colored middle class was a primary reason for its cooperation in destroying representative government (1970, p. 316), and (Green, 1991, p.296) similarly argued that new colored elites felt just as threatened by mob violence as the whites. In other cases, it appears that old elites invited the Crown against the will of the new elites: Honychurch s account of Dominica s abolition of parliament is a cloak-and-dagger story of regime change that was pushed through overnight and against strong opposition from parts of the Assembly (1995, p.69-71). The Minutes of Grenada s Assembly see Appendix B report that on September 15th 1875, the day that An Act to Amend and Simplify the Legislature of the Island of Grenada came to a vote, an amendment was introduced by a faction of the Assembly to rename the bill into An Act to diminish the rights of the people in electing, from their own body, fit and proper representatives, and to lay oppressive taxation on every side. The Assembly minutes make no further comment on this proposed amendment, but Grenada s elites clearly did not uniformly support the abolition of their assembly. 4 A Tale of 2 Islands For the two largest colonies Barbados and Jamaica we were able to obtain the roll-call votes of the local parliaments as well as a fairly complete picture of the identity of the parliamentarian, allowing us to distinguish old planter elites from new elites, which we can in turn break into those with planting interests and those without. See Appendix B for details on these data. Aside from being economically the most important colonies, Barbados and Jamaica also illustrate two of the most important equilibrium paths in the model. Barbados was an outlier among the Caribbean slave 10 To illustrate: Britain (1879, p. 188) reports how Grenada s assembly at its meeting on February 9th 1876 addressed the Queen that it had passed a bill for its own extinction 17

societies in its geography. While all Caribbean islands shared their climatic conditions, there was large variation in geographic characteristics like elevation and soil. The typical Caribbean sugar colony was characterized by sugar-suitable coastal plains and a rugged interior that lay fallow during slavery. 11 Barbados was the only Caribbean sugar island that combined the advantages of limestone rather than volcanic soil with a high enough elevation to protect sugar from saltwater and storm surges. 12 In Barbados, the entire land area was highly sugar-suitable land, and over 95% of its land was under cultivation on the eve of emancipation, compared to under 50% elsewhere in the Caribbean (Martin (1839, p.32 102)). While Barbados was not particularly unique during slavery, it was unique after emancipation in that the only avenue to wealth and influence was through sugar cultivation. Jamaica by contrast typified the geography of the rest of the British West Indies but was unique in being by far the largest in scale. This scale meant that London abolitionists and Baptist ministers focused their efforts of empowering the emancipated slaves on Jamaica (Holt (1991, ch. 6)). This in turn meant that the old planters in Jamaica faced pronounced political opposition earlier than in the other colonies. Jamaica was also unique in that it ended up being the only Caribbean island were the assembly abolished itself as a direct response to the threat of revolt. Our theory s predictions about elite competition and co-option depend on whether the economic incentives of old and new elites are aligned. We therefore want to distinguish in the rollcall data between old and new elites, and between those new elites whose economic incentives aligned with the old elites, and those whose incentives did not. The colonial records that we use to obtain the names of assemblymen see the Appendix B do not provide any information on these assemblymen. We therefore laboriously assembled plantation ownership information from other sources. In total we were able to assemble 61 plantation surveys from 43 distinct sources. Each plantation survey gives us a snapshot of plantations in one colony for one year. Table 2 re- 11 The top panel of Online Appendix A Figure 1 shows that plantation lands exactly coincided with these coastal plains. 12 The Caribbean is divided into three island chains: The Greater Antilles are large islands with mountainous interiors and coastal plains. Of these, only Jamaica was a British colony, the others are Cuba, Haiti and the Dominican Republic. Most British Caribbean colonies Dominica, the British Virgin Islands, Grenada, Montserrat, Nevis, St. Kitts, St. Lucia, and St. Vincent belonged to the inner chain of the Lesser Antilles, which is volcanic and mountainous. The outer chain of Lesser Antilles Anguilla, Bahamas, Barbados, Turks and Caicos consists of flat limestone. This limestone was more suitable for sugar cultivation because it retained water better than the volcanic land on the inner chain (Richardson (1997, p. 147)) and because sugar does not like high elevations. The importance of elevation for sugar suitability is illustrated in the middle and bottom panels of Online Appendix A Figure 1. 18

ports what data we have. The data in Table 2 is organized in first differences for the purpose of studying plantation survival and ownership turnover in section 4. One aspect of the data in Table 2 is that for the pre-emancipation years, data sources are very consistent: This is because the data collection for the Slave Registries (around 1817) was mandated by the colonial office in London, and the Compensation Tables (for 1835) were collected by the treasury in London. 13 This means we can get a fairly consistent measure of who the old, i.e. pre-emancipation, planter elites are in all colonies. Distinguishing between new elites with and without plantation interests is harder, and in fact the data we have from the plantation surveys for post-emancipation years is too fragmented to do this for most colonies. For Barbados however, we have very regular reporting on plantation ownership for the period we study here, 1838 1865, and all plantation records come from the same consistently reported source. That means Barbados is the one place where we feel comfortable using the plantation surveys to distinguish new elites economic interests through the plantation records. For Jamaica, Table 2 shows that we cannot do this, but we are helped by the fact that Jamaica s prominent status in Caribbean history has generated a disproportionate amount of historical studies. Among these are two by Heuman (1981) and Holt (1991) that are focused on Jamaica s legislative politics and are explicitly biographical, describing assemblymen s ethnicity, occupation and politics in great detail. Using these sources, we are therefore able to distinguish assemblymen in Jamaica and Barbados in the 1838 1865 period into three groups: old planter elites, new elites with plantation interests, and new elites with no plantation interests, which we refer to as new merchant elites. How did these three elite groups interact in Jamaica and Barbados? Figures 1 and 2 visualize the roll-call voting network for two parliamentary sessions, 1844 1849 and 1854 1859. 14 The 1843 election marked the first active drive to register the black and colored smallholders and to get out their votes, and as a result, in the parliament of 1844 1849 black and colored politics formed a distinct faction in Jamaican politics (Holt (1991, ch. 7)). Black nodes are assemblymen representing old elites (planter families that appear in the plantation surveys prior to emancipation), grey nodes are new elites with plantation interests, and white nodes are new elites with no plantation 13 When British Parliament abolished slavery, it set aside money to compensate slave owners for their lost assets. The disbursal of that money was recorded in the Compensation Tables. 14 These visualizations were obtained in Gephi, using the Yifan Hu visualization algorithm. The underlying edgelist is set up so that two assemblymen (nodes) are connected by an edge if they agree on at two-thirds of the proposals on which they both voted. 19

Table 2: Overview of Plantation Data Used colony year t t+t plantation survival(%) ownership known t+t' ownership unchanged(%) Antigua 1817 1829 0.86 Y 1829 0.62 Antigua 1829 1835 1.00 Y 1843 0.69 Antigua 1835 1843 0.90 Antigua 1843 1851 0.99 Y 1851 0.78 Antigua 1851 1858 0.92 Y 1878 0.20 Antigua 1858 1878 0.86 Antigua 1878 1891 0.73 Y 1891 0.47 Barbados 1817 1835 0.81 Y 1835 0.51 Barbados 1835 1848 0.75 Y 1848 0.61 Barbados 1848 1854 0.92 Y 1854 0.79 Barbados 1854 1861 0.95 Y 1861 0.72 Barbados 1861 1865 0.98 Y 1865 0.87 Barbados 1865 1870 0.97 Y 1870 0.87 Barbados 1870 1898 0.85 Y 1898 0.49 Dominica 1817 1835 0.57 Y 1835 0.29 Dominica 1835 1878 0.71 Y 1878 0.12 Grenada 1817 1835 0.57 Y 1835 0.35 Grenada 1835 1849 0.77 Y 1867 0.24 Grenada 1849 1867 0.47 Guyana 1833 1838 0.67 Guyana 1838 1846 0.99 Guyana 1846 1860 0.51 Guyana 1860 1879 0.62 Y 1879 0.26 Jamaica 1829 1835 0.76 Y 1840 0.62 Jamaica 1835 1840 0.69 Montserrat 1835 1848 0.82 Y 1848 0.58 Montserrat 1848 1858 0.38 Nevis 1817 1835 0.81 Y 1835 0.24 Nevis 1835 1878 0.62 Y 1878 0.10 Nevis 1878 1897 0.78 Y 1897 0.53 St. Lucia 1852 1897 0.82 St. Kitts 1835 1847 0.81 Y 1850 0.40 St. Kitts 1847 1850 0.82 Y 1878 0.31 St. Kitts 1850 1878 0.85 St. Kitts 1878 1897 0.47 Y 1897 0.58 St. Vincent 1817 1827 0.62 Y 1827 0.42 St. Vincent 1827 1831 0.90 Y 1831 0.57 St. Vincent 1831 1835 0.43 Y 1835 0.00 Tobago 1819 1832 0.96 Y 1832 0.25 Tobago 1832 1835 1.00 Y 1835 0.49 Tobago 1835 1847 0.92 Y 1862 0.13 Tobago 1847 1862 0.91 Tobago 1862 1881 0.75 Y 1881 0.26 Tobago 1881 1894 0.93 Y 1894 0.44 Trinidad 1813 1835 0.30 Y 1835 0.21 Trinidad 1835 1882 0.50 Y 1882 0.18 Trinidad 1882 1888 0.87 Y 1888 0.77 Trinidad 1888 1894 0.83 Y 1894 0.78 Notes: One row in this table is one first difference. Column 2 is the base year and column 3 the next year. Not all plantation surveys included ownership information. As a result, first differences were re-defined for ownership information. For example, the plantation survey in Antigua in 1858 did not include ownership information. We can therefore ask if a plantation that existed in 1851 still existed in 1858, but we can only ask if a plantation owned by planter i in 1851, was still owned by planter i in 1878. 20

Figure 1: Voting Network in Jamaica s Parliament, 1844 1849 Session Figure 2: Voting Network in Jamaica s Parliament, 1854 1859 Session 21

interests. 15 As is apparent from this visualization, the 1844 1849 session was characterize by a large degree of homophily i.e. disagreement between groups but that disagreement was only between old elites and new merchant elites. New planter elites visually appear as part of the planter faction as opposed to the new elite faction. In 1844 1849, there were 37 assemblymen representing old planter elites, 10 representing new planter elites and 13 representing new merchant elites. Ten years later, Figure 2 shows that old planter elites no longer constituted a majority of assemblymen. In 1854 1859, there were 23 assemblymen representing old planter elites, 8 representing new planter elites and 28 representing new merchant elites. Faced with this large influx of new elites, Jamaica s planters had to co-opt to survive. While new planter elites largely voted with the old planters even in 1844 1849, ten years later, there was also noticeably more agreement by the new merchant elites, i.e. there is little visual separation between the black and white voting networks. Table 3 makes this point more precise by measuring voting agreement between the three blocs. 16 The top-panel describes Jamaica s parliament. The bottom-panel describes Barbados parliament. The table moves chronologically rightward across columns. In each panel, the first row reports the average agreement between two members of the old elite, the second row reports average agreement between member of the old planter elite and members of the new planter elite, and the third row reports the average agreement between members of the old planter elite and members of the new non-planter elite. The fourth and fifth row report t-statistics for the hypotheses that the first row differs, respectively, from the second and third. Several patterns can be observed in this table: One is that Jamaica s parliament was considerably more acrimonious from 1844 1853 than before and after. New merchant elites were around 15% ( 63%-48%, and 67%-50% ) less likely than another old-elite representative to vote with an old elite representative. During this most acrimonious period, even the new planter elites were less likely to vote with the old elites. The acrimony was less, but it was significant, with t-statistics of 3.47 and 4.57 respectively. This polarization abated considerably thereafter. While some homophily continued between new merchant elites and old elites, the voting disagreement was 15 From Heuman and Holt, we know that the majority of both types of new elites were colored or black, but we do not distinguish that here because we cannot measure it anywhere else. 16 Online Appendix A Figures 4 through 6 show the full sequence of Jamaica s post-emancipation parliamentary sessions until 1865, and Figures 7 through 8 show the same for Barbados. 22

Table 3: Homophily in Jamaica s and Barbados Parliaments 1838 1865 Sessions, Jamaica: (1838-1843) (1844-1849) (1849-1853) (1854-59) (1860-1862) (1863-1865) 1: Average Agreement (Old Elites & Old Elites) 57% 63% 67% 61% 63% 65% 2: Average Agreement (Old Elites & New Planter Elites) 57% 60% 60% 64% 64% 70% 3: Average Agreement (Old Elites & New Non-Planter Elites) 51% 48% 50% 55% 56% 60% t-statistic(1 = 2) 0.5 3.47 4.57-1.97-0.63-2.68 t-statistic(1 = 3) 5.97 15.54 15.69 5.4 5.03 3.57 Number of Connections (Old Elites & Old Elites) 772 640 538 210 185 264 Number of Connections (Old Elites & New Planter Elites) 234 340 205 166 97 94 Number of Connections (Old Elites & New Non-Planter Elites) 308 444 806 582 434 394 Sessions, Barbados: (1838-1849) (1849-1859) (1860-1865) 1: Average Agreement (Old Elites & Old Elites) 50% 50% 52% 2: Average Agreement (Old Elites & New Planter Elites) 51% 50% 54% 3: Average Agreement (Old Elites & New Non-Planter Elites) 51% 54% 52% t-statistic(1 = 2) -0.8 0.6-1.25 t-statistic(1 = 3) -0.28-1.09 0.16 Number of Connections (Old Elites & Old Elites) 1158 642 492 Number of Connections (Old Elites & New Planter Elites) 409 260 227 Number of Connections (Old Elites & New Non-Planter Elites) 90 25 45 Notes: TBA much less pronounced on average. For the new planter elites with planting interests, homophily disappeared altogether; in fact, they agreed more with the old elite representatives than the old elite agreed amongst themselves. Jamaica thus appears to have gone through a period of significant polarization between old and new elites from 1844 to 1853, after which there appears to have been full co-option of new elites with aligned incentives and partial co-option of new elites with divergent economic incentives. All this contrasts sharply with the situation in Barbados, where there is never any degree of homophily between new elites and old planter elites, regardless of whether the new elites were planters or not. 17 The evidence in Table 3 suggests that Jamaica s planter elites had re-established their control of the assembly in the 1853 1865 period. Why then did Jamaica s assembly abolish itself in January of 1866? In our theory, when revolt has occurred it is too late for new elites to regain legitimacy. The outside option of inviting a foreign intervention then becomes attractive not only for old elites but also for new elites that had been co-opted in previous periods. It is worth noting that if the outcome of a successful revolt is a transition to broad-based democracy, then even the new elites who had not been co-opted may prefer foreign intervention to maintain their elite status. 17 In Jamaica, the detailed narrative accounts in Heuman (1981) and Holt (1991) allow us to know with confidence that individuals we code as non-planters had no economic stake in the plantation system. In Barbados, however, we can only say whether a person is recorded on the plantation surveys or not. 23

However, on average they would be less supportive of the foreign intervention because they will be less negatively affected by the revolt. On October 11th 1865, Jamaica was rattled by the Morant Bay Rebelion, the largest and most violent uprising in post-emancipation Caribbean history. On that day, 22 civilians were killed and 34 wounded. On October 13th governor Eyre proclaimed Martial Law and British troops put the rebelion down. Under martial-law 354 freedmen were sentenced to death by courts martial and another 85 killed by soldiers and sailors in skirmishes. No disturbance in the West Indies since the days of Emancipation has caused half so much excitement. [...] It was the last occasion when regular troops have fired with ball or shot on rioters in the British West Indies. (Cundall (1906, p.12 15)). Morant Bay seriously panicked Jamaican elites and it is generally agreed that it was the immediate cause of Jamaica s abolition of parliament (Heuman (1994), Lewis (2004, p.96)). But how did each distinct elite group respond to Morant Bay? We uncovered a publication that included the signatures by prominent Jamaicans to a note of gratitude made to governor Eyre in late 1865 for his efforts at suppressing the Rebelion (Jamaica (1866)). 18 We digitized these lists and matched the signatories to Jamaica s assembly-data. 67% of old planter elites were on the signatory lists compared to only 45% of new merchant elites. Interestingly, only 20% of new planter elites signed the document. When we decomposed the new merchant elites by their past voting record, and calculated for each individual assemblymen in the new merchant elites their average voting agreement with the old elites, we found that this average agreement significantly negatively affected the probability of being a signatory. 19 It therefore appears as though it was exactly those new elites whose legitimacy with the citizenry would have been in question i.e. new planter elites and co-opted new merchant elites were reluctant to put their name on a document thanking the governor for putting down the rebelion. Those new elites whose legitimacy was not in question because they had no plantation interests and had more consistently voted against the 18 Another way to answer this would be to check who supported the final proposal in Jamaica s parliament to abolish parliament. Unfortunately, that last parliamentary session is conspicuously missing from the official records, not only from the North American and British collections, but even from Jamaica s National Archives, where the local librarians searched for this document in vein. 19 We calculated agreement in two ways: either as a raw average over all sessions in which an individual was an assemblymen, or normalizing the voting data to new merchant elite s average agreement with the old elites in a give session. The second method accounts for the fact that a relatively co-optable new merchant elite who was in the assembly only during the acrimonious sessions of 1844-1853 when agreement between new elites and old elites was generally low would otherwise show up as not very co-optable. The basic result comes through significantly with both measures and in both linear probability and probit regressions. 24

planters could afford to condone the governors actions in putting down the rebelion. 5 A Tale of 10 Islands Core to our analysis are the circumstances under which we see co-option of new elites as opposed to competition. In Barbados, co-option was easy because all wealth was generated by sugar so that new elites economic incentives were aligned with old elites and because the scarcity of land generated few new voters and therefore little pressure on new elites to pursue different policies. In Jamaica, co-option of new elites became evidently more difficult for ten years from around 1843. 1843, however, marked the nadir of the Caribbean sugar plantation system, at least until the mid- 1880s. Aided by stabilizing global sugar prizes, London abolitionists relaxing their supervision of local labor markets through so-called Special Magistrates, and the 1849 repeal of the British Navigation Act (which had prohibited Caribbean planters from selling their sugar to North American merchants), old Caribbean planter elites re-exerted the control. 20 As we showed, Jamaica did not abolish its parliament because its old elites lost control, it abolished its parliament because both old elites and co-opted new elites faced an immediate threat of revolt. But Morant Bay was an exceptional event in Caribbean history. Why then did elites in the other 9 colonies abolish their parliaments? Our theory predicts that old elites will prefer inviting an outside intervention if it becomes too difficult to co-opt the new elites, either because there are too many or because their preferences diverge too much from the old elites. What then gave the impetus for the wave of colonies abolishing their parliaments between 1855 and 1876? Our reading of Caribbean history is that it was a Caribbean-wide law change that led to a rapid churning of planter elites, replacing old planter families with new creole (i.e. locally born and raised) planters at a rate that made co-option too difficult. 21 Despite the plantation economy s recovery from its 1843 nadir, many old planters were slowly edging towards bankruptcy because of encumbrances that many British planter families had attached to their estates during sugar s heyday in the 18th century. Encumbrances were regular 20 For a detailed account of the recovery of the Caribbean sugar plantation economy after 1843, see Dippel, Greif, and Trefler (2015) who investigate the effects of the sugar economy on local Caribbean labor markets in the 14 British Caribbean sugar colonies. 21 The new planters were more accountable to the people. Being creole and in most cases colored meant that were less shielded from people s ill-will and that they did not have the option of moving their families to England in case of a revolt. 25

monetary commitments to the wider family in England that were to be paid from the plantations revenues. Because the legal tangle of encumbrance obligations was prohibitively expensive to resolve, many operationally profitable plantations were nominally bankrupt, piling up debt with their encumbrancers (Beachey (1978, ch.1), Cust (1859, pp.9 13)). In 1854, British Parliament passed the Caribbean Encumbered Estate Act (EEA), which made three innovations: First, it allowed any of the plantations many potential encumbrancers and creditors to initiate bankruptcy proceedings; second, it established a clear legal hierarchy of creditor claims, and third, it instituted a specialized court in London with offshoots in the colonies that ensured that these claims were processed in a timely manner (Cust (1859, pp.5-7, 13-15), Sewell (1861, pp.82, 89)). The EEA led to dramatic turnover in the ownership of sugar plantations throughout the Caribbean. According to Lowes, The act played a key role in the snowballing process of turnover so great a role, in fact, that one merchant reported that all but eight estates changed hands [in Antigua] between 1860 and 1897 (1994, p.21) Who were the new planters? Most of the sold estates ending up in the hands of local merchants and former estate managers, so-called attorneys (Lowes (1994), Craig-James (2000)). 22 The new planters were almost exclusively creole, i.e. born, raised and resident on the islands, and in large part colored (Lowes (1994), Craig-James (2000, p.199-201), Lewis (2004, p.67)). Brizan (1984, p.201 202) states that especially after 1850, the vacuum created by the exodus of White planters was now being filled by the rich Coloreds, and Craig-James (2000, p.199-215) gives many examples of successful black and colored planters in Tobago. In a contemporary account Froude lamented that the English of those islands are melting away. Families who have been for generations on the soil are selling their estates everywhere and are going off.... The white is relatively disappearing, the black is growing (Froude, In (1888, ch. XVII). These new planter elites may have supported policies similar to those of the old elites, 23 but were at the same time much more constrained in being held accountable. The introduction of the EEA in 1854 thus allows us to isolate the role of identity and elite composition in the theory, because it occurred at a 22 According to (Brizan, 1984, p.201), the rising class of merchants were becoming the owners of bankrupt estates. The EEA established priority among claimants of the merchant s lien, which gave the merchants that financed the dayto-day plantation operation first call to take over the estates (Crossman and Baden-Powell (1884), Cust (1859, pp.5-7, 13-15)). When these merchants were local they often took over the estates they had hitherto financed ((Brizan, 1984, p.201)). When the merchants were in London, they usually resold the estates to local interests (Lowes (1994, ch.1, pp.19-22)). 23 Henry Taylor, the colonial office s supervisor of West Indian affairs, viewed the new colored planter elites with suspicion: it was folly to imagine that a change in allotment of political power would result in better government. An oligarchy whether brown or black, would be no less odious than the present one, in (Rogers, 1970, p65). 26

Figure 3: Effect of EEA on Planter Turnover Notes: The left panel shows the linear effect of time elapsed for seven different ten-year bins on ownership persistence. One additional year in the 1855 164 window increases the probability of changed ownership by 2.5%. the right panel shows the same for plantation survival as the outcome. time when the economic circumstances of the plantation system the terms of trade and sugar s importance in the agricultural mix where very stable. If the EEA had sizeable effects on elite churning, we should see this in the plantation ownership data in Table 2. Because the years in which we observe plantations vary case by case, the data is set up in first differences whose start and end years are specific to each plantation survey. To deal with this, we regress a dummy for whether ownership remains unchanged over a given period, on the number of years that have passed, testing specifically for an increase in ownership turnover in the years after the EEA. The simplest specification is D(OwnerUnchanged) i,t t+t = αt + βt 2 + γt 1855 1864 + ɛ it, (21) Where α and β estimates the linear and quadratic effects of elapsed time between surveys on persistence. We reasonably expect α < 0 and β > 0, i.e. more time elapsed raises the probability of changed ownership, but this effect does not extrapolate linearly. Any increased turnover after the adoption of the EEA should get picked up by γ, where the variable T 1855 1864 takes on values from 0 to 10. As an alternative outcome, we ask whether the plantation itself survives from year t to t + T. Table 2 reports the raw data probabilities of firm survival and unchanged ownership in all plantation survey first differences. When we run this regression we estimate α = 0.035 with a t-statistic of 7.23, and γ = 0.01 with a t-statistic of 2.8. Each year that passes increases the probability of changed ownership by 3.5%, and in the ten-year window after the introduction of 27

the EEA, we see this effect increase to 4.5%. When we consider plantation survival as the outcome, we estimate α = 0.024, with a t-statistic of 3.3, and γ = 0.003 with a t-statistic of 0.4. In other words, elapsed time mattered but there was no additional effect of the EEA on plantations survival overall, only on ownership. We also replaced αt with a more flexible specification by slicing the entire time range covered by the plantation surveys into ten-year windows and estimating separate linear time trends for each. For example, for Barbados the next plantation survey after 1835 is 1848, so that in that first difference we regress D(OwnerUnchanged) i,1835 1848 on T 1825 1834 = 0, T 1835 1844 = 10, T 1845 1854 = 4, T 1855 1864 = 0 etc. Figure 3 plots the point estimates and 95% confidence bands for the effect of time in each of the ten-year windows. The left panel shows that only the effect of T 1845 1854 and T 1855 1864 are significant, but the latter has a much bigger point estimate of γ = 0.025 as opposed to γ = 0.014 for T 1845 1854. The right panel shows that for plantation survival overall, no time window is ever quite significant with a 95% confidence interval. While this evidence is coarse, it does support quite strongly the qualitative evidence that the EEA caused a dramatic increase in plantation ownership turnover. For Barbados and Jamaica in section 5, we could partition assemblymen into the three groups of our theory old planter elite, new planter elite, and new merchant elite. But we do not have the necessary data to do this for the other nine islands. Instead, we are limited to partitioning the assemblymen only into old elite and new elite, based on whether they appear in the pre-1854 plantation surveys or not. In Figure 4, we track each colony s seat share held by old elites over time, with the series ending in each colony when assembly is abolished. Each panel also includes a vertical line in 1854. All panels are scaled the same on both axes. We discussed Barbados and Jamaica in section 4: In Barbados, old elites were in a uniquely resilient position. This is reflected in the very high share of assembly seats held by old planter elites. This share was not noticeably affected by the EEA, and it never fell below 80% before the 1880s. As a consequence, Barbados was the only one of the 14 islands that ended the period with its assembly intact. For Jamaica, whose scale made it the focal point for abolitionists s efforts to empower the emancipated slaves, the voting networks in section 4 showed that political turmoil came earlier than elsewhere, and that the most acrimonious period of Jamaican politics was 1844-1853, followed by a period of relative calm in from 1854 to 1865. This is borne out also in Figure 4: 28

Antigua Figure 4: Share of Assembly-Seats of Planter Families Barbados Dominica Grenada.2.4.6.8 1 (mean) plantershare.6.7.8.9 1 (mean) plantershare (mean) plantershare.3.4.5.6 (mean) plantershare 0.2.4.6 18301840185018601870 year 1840 1860 1880 1900 year 1840 1845 1850 1855 1860 1865 year 18401850186018701880 year Jamaica Montserrat Nevis St Kitts.7.75.8.85.9.95 (mean) plantershare.6.7.8.9 1 (mean) plantershare (mean) plantershare.4.5.6.7.8 (mean) plantershare.3.4.5.6.7 18301840185018601870 year 1835 1840 1845 1850 1855 1860 year 18301840185018601870 year 1840 1845 1850 1855 1860 1865 year St Vincent Tobago Virgin Isl. (mean) plantershare 0.2.4.6.8 (mean) plantershare.2.3.4.5.6 (mean) plantershare 18301840185018601870 year 18401850186018701880 year 1835 1840 1845 1850 1855 year Notes: The blue time-line is the seat share of assemblymen we can match to any plantation survey before 1854. The first vertical line is the 1854 EEA. The second vertical line is the abolition of the assembly. 29

In Jamaica too, the post-eea period did not bring a significant drop in old elites representation in the assembly. In all of the other islands Antigua, Dominica, Grenada, Montserrat, Nevis, St. Kitts, St. Vincent and Tobago we see a clear erosion in the seat share of old elites after 1854 and in the lead-up to each colony s abolition of parliament. 24,25 In our theory, when a successful revolt occurs, old elites will have an incentive to invite a foreign intervention. New planter elites are likely to support this, and even new merchants may support it if the result of a successful revolt is democratic transition and if merchant elites fare better in oligarchy. This was the case of Jamaica. However, there were no revolts in the lead-up to the abolition of the assemblies in any of the other islands. Given that the elites did, however, abolish their assemblies, our theory suggests that it became too expensive for planter elites to co-opt the new elites. The right panel of Figure 3 does not suggest that plantations died, 26 but rather that old planter elites were replaced with new planter elites. It is possible that the new planter elites were less skilled at obtaining assembly-seats, and the plantation data is not good enough to allow us to distinguish whether the old planter elites were replaced primarily by new planter elites or new merchant elites in the assemblies in Figure 4. However, co-option would have gotten more expensive either way, just to varying degrees: new merchant elites had different economic incentives, new planter elites had aligned economic incentives but still needed to maintain legitimacy. Who, other than obviously the old planter elites, does the theory predict should support the abolition of parliament? New planter elites are likely to support the intervention, so long as their standing afterwards is not significantly worse than before, because the intervention will secure the status-quo policies while relieving the new planter elites of their need for maintaining legitimacy. Unlike in the Jamaican case of imminent or successful revolt leading to broad-based democracy, new merchant elites are likely to oppose the intervention because they do not favor the status quo policies that get locked in by the foreign intervention, and because they will no longer receive 24 In St. Vincent, the old elites seat share likely mis-measures the old elite s influence because St. Vincent was infamous for its degree of absentee ownership after emancipation (Smith and Forster (2013)). The absentee planters plantation managers and attorneys were probably still dominant enough in the assembly to constitute a majority. Lowenthal labeled St. Vincent the most regressive plantocracy of the Caribbean (Richardson, 1997, Foreword). 25 For the Virgin Islands, we did not have any plantation data, and can generally not say much, as its records were the worst by some margin. Dookhan (1975, p.201) quotes a contemporary source in the Virgin Islands like this: the peculiarity of the place is not so much the want of competent legislators as the universal consciousness of that want so that nobody will undertake to legislate 26 Indeed, Dippel et al. (2015) show that the sugar plantation remained very stable until the mid-1980s. 30

co-option rents after the intervention. Given this, what do we know about who supported the abolition of the assemblies when they happened? In Nevis, the Minutes of Assembly did not contain roll call votes, but there is a record that the meeting of June 14th 1866, when the abolition of the assembly was voted on, began with the reading of a petition by smallholders against the constitutional changes in the legislature, thus suggesting by default that it favored the planters. Because smallholder were the chief victims of planters labor repressive policies, it follows that planters must have supported the assembly s abolition in Nevis. There were three other colonies whose Minutes of Assembly includes records of the vote for the abolition of the assembly: In St. Kitts, the bill passed twelve to five, but the identity of the ayes and nos were not recorded. In Grenada, the bill also passed twelve to five. Of the four assemblymen whose names we could match to the 1867 plantation survey, all four voted in favor of the bill. In St Vincent, the bill passed ten to five. Of the two names we could match to the pre-emancipation plantation surveys, both voted in favor of the bill. Overall, while this evidence on who supported the final vote to abolish parliament is sparse, it is very consistent with our theory and our narrative of the post-emancipation Caribbean islands. 6 Discussion & Conclusion TBA References Acemoglu, D., S. Johnson, and J. Robinson (2005). The Rise of Europe: Atlantic Trade, Institutional Change, and Economic Growth. The American Economic Review 95(3), 546 579. Acemoglu, D. and J. Robinson (2000). Why did the West Extend the Franchise? Democracy, Inequality, and Growth in Historical Perspective. Quarterly Journal of Economics 115(4), 1167 1199. Acemoglu, D. and J. Robinson (2001). A Theory of Political Transitions. The American Economic Review 91(4), 938 963. Acemoglu, D. and J. Robinson (2006a). De facto Political Power and Institutional Persistence. American Economic Review Papers and Proceedings 96(2), 325 330. Acemoglu, D. and J. Robinson (2006b). Economic Origins of Dictatorship and Democracy. Cambridge University Press. 31

Acemoglu, D. and J. Robinson (2008). Persistence of Power, Elites, and Institutions. American Economic Review 98(1), 267 293. Beachey, R. (1978). The British West Indies Sugar Industry in the Late 19th Century. Greenwood Press. Besley, T., T. Persson, and M. Reynal-Querol (2012). Political Instability and Institutional Reform: Theory and Evidence. London School of Economics (unpublished). Bidner, C., P. Francois, and F. Trebbi (2014). A theory of minimalist democracy. Technical report, National Bureau of Economic Research. Bourguignon, F. and T. Verdier (2000). Oligarchy, Democracy, Inequality and Growth. Journal of Development Economics 62(2), 285 313. Britain, C. O. G. (1879). The Colonial Office List. HM Stationery Office. Brizan, G. (1984). Grenada, Island of conflict: from Amerindians to People s Revolution, 1498-1979. Zed Books London. Brückner, M. and A. Ciccone (2011). Rain and the Democratic Window of Opportunity. Econometrica 79(3), 923 947. Craig-James, S. (2000). The Social and Economic History of Tobago (1838 1990). Craton, M. (1988). Continuity not Change: the Incidence of Unrest Among Ex-Slaves in the British West Indies, 1838 1876. Slavery & Abolition 9(2), 144 170. Crossman, C. C. and G. E. M. Baden-Powell (1884). Report on the West Indian Incumbered Estate Act. London: Eyre and Spottiswoode. Cundall, F. (1906). Political and Social Disturbances in the West Indies: A Brief Account and Bibliography. Institute of Jamaica. Curtin, P. D. (1954, Spring). The British Sugar Duties and West Indian Prosperity. Journal of Economic History 14(2), 157 164. Cust, R. J. E. (1859). A Treatise on the west Indian encumbered estate acts. London: William Amer, Law Bookseller. Dippel, C., A. Greif, and D. Trefler (2015, February). The rents from trade and coercive institutions: Removing the sugar coating. Nber working paper no. 20958. Dookhan, I. (1975). A History of the British Virgin Islands, 1672 1970. Epping, England: Caribbean Universities Press. Froude, J. A. (1888). The English in the West Indies: Or, The Bow of Ulysses. London: Longmans, Green. 32

Fuller, T. (2014). Thai Army Declares Coup, Citing Need to Reform Nation. New York Times May 22nd. Galloway, J. (2005). The Sugar Cane Industry: An Historical Geography from its Origins to 1914. Cambridge Univ Pr. Green, W. (1991). British Slave Emancipation: the Sugar Colonies and the Great Experiment 1830-1865. Oxford Univ Pr. Haggard, S. and R. R. Kaufman (1997). The Political Economy of Democratic Transitions. Comparative Politics, 263 283. Heuman, G. J. (1981). Between black and white: race, politics, and the free coloreds in Jamaica, 1792-1865. Praeger Pub Text. Heuman, G. J. (1994). The Killing Time : The Morant Bay Rebellion in Jamaica. Macmillan. Higman, B. (2001). Jamaica Surveyed: Plantation Maps and Plans of the Eighteenth and Nineteenth Centuries. Univ of West Indies Pr. Higman, B. W. (2000). The Sugar Revolution. Economic History Review 53(2), 213 236. Holt, T. (1991). The Problem of Freedom: Race, Labor, and Politics in Jamaica and Britain, 1832 1938. The Johns Hopkins University Press. Honychurch, L. (1995). The Dominica Story: A History of the Island. Macmillan Pub Limited. Jamaica (1866). Addresses to His Excellency Edward John Eyre, Esquire, 1865, 1866. Kingston, Jamaica: M. DeCordova & Co. Printers. Lewis, G. (2004). The growth of the Modern West Indies. Ian Randle Publishers. Lipset, S. (1959). Some Social Requisites of Democracy: Economic Development and Political Legitimacy. American Political Science Review 53(1), 69 105. 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(2), 707 765. Lowes, S. (1994). The Peculiar Class: The Formation, Collapse, and Reformation of the Middle Class in Antigua, West Indies, 1834-1940. Columbia University. Martin, R. (1839). Statistics of the Colonies of the British Empire. Meltzer, A. and S. Richard (1981). A Rational Theory of the Size of Government. The Journal of Political Economy, 914 927. 33

Michels, R. (1911). Political parties: A sociological study of the oligarchical tendencies of modern democracy. Reprinted. New York: Free Press, 1962. Morrell, W. and W. Parker (1969). British Colonial Policy in the Mid-Victorian Age: South Africa, New Zealand, the West Indies. Clarendon P. North, D. C. and B. R. Weingast (1989). Constitutions and Commitment: the Evolution of Institutions Governing Public Choice in Seventeenth-Century England. The journal of economic history 49(04), 803 832. Przeworski, A., M. Alvarez, J. A. Cheibub, and F. Limongi (2000). Democracy and Development: Political Regimes and Economic Well-Being in the World, 1950-1990. New York: Cambridge University Press. Richardson, B. (1997). Economy and Environment in the Caribbean: Barbados and the Windwards in the Late 1800s. University Press of Florida. Rogers, H. A. (1970). The Fall of the Old Representative System in the Leeward and Windward Islands, 1854-1877. Ph. D. thesis, University of Southern California. Sewell, W. (1861). The Ordeal of Free Labor in the British West Indies. Harper & brothers. Smith, S. D. and M. Forster (2013). The curse of the caribbean? agency s impact on the efficiency of sugar estates in st.vincent and the grenadines, 1814-1829. (112). Taylor, A. (2002). The American Colonies, Volume 1. Penguin Group USA. Trefler, D. and D. Puga (2014). International trade and institutional change: Medieval venices response to globalization diego puga. Quarterly Journal of Economics. Wallace, E. (1977). The British Caribbean from the Decline of Colonialism to the End of Federation. University of Toronto Press. Wrong, H. (1923). Government of the West Indies. Clarendon Press. 34

Appendix A Appendix B Appendix B.1 Mathematical Appendix Data Appendix Assemblymen Data The names of assemblymen, as well as data on election cycles comes from the British Colonial Blue Books, annual statistical accounts that were sent to London from each individual colony to report on local conditions. 27 From 1836, the books Councils and Assemblies section reported the names of all local politicians, with election dates and the parish they represented. For each colony, The British National Archives maintain 6 data-series (Original correspondence, Entry Books, Acts, Sessional Papers, Gazettes, and Miscellenea), and the Blue Books form the bulk of the Miscellenea series. Appendix B.2 Roll Call Data Voting data can be cleaned from a colony s Minutes of Assembly, which forms part its Sessional Papers series. For most colonies, the minutes do not report actual voting data but in some including for Barbados and Jamaica they do. Appendix B.3 Plantations & Plantation Ownership Before emancipation, plantation data comes from two sources. From 1813 on, the Crown required colonies to register all slaves. Most colonies have three or four iterations of the slave registries, but each new iteration simply updated the initial information with births and deaths. The slave registries are fully uploaded on Ancestry, where we browsed them and manually compiled plantation listings. In the slave registries, one observation is a slave, but the owner is also lists. We systematically searched Ancestry s records for long series of slaves with identical owners, and then browsed the underlying images for the plantation and parish information that was not digitized by Ancestry. For most colonies, the next plantation survey came from the slave compensation table of moneys disbursed by London to slave owners to compensate them for the emancipation of their slaves. The full records have been compiled by a research team at University College London, and made publicly available. There are 30,308 claimants recorded, and their information can be viewed on consecutive url s running from http://www.ucl.ac.uk/lbs/claim/view/1 to http://www.ucl.ac.uk/lbs/claim/view/30308. These records indicate whenever the compensated individual was associated with a plantation and list the name of said plantation. After emancipation, most plantation surveys came from individual colony-specific almanacs. We had a number of research assistants search for these almanacs, primarily through interlibrary 27 For years before the 1890s, at most two copies exist of each Blue Book, one in the issuing colony s archives and one in the British National Archives, in London, where this data was hand-collected. 35

loans. For Barbados, we obtained photocopies of plantation surveys from the Barbados Museum and Historical Society. In total, we collected the 61 plantation surveys in Table 2 from 43 distinct sources. See Online Appendix A Table 1 for a detailed listing of all sources. Appendix B.4 Assemblymen Characteristics In Barbados, we only know whether an assemblyman, or one his relations, was listed as a plantation owner. In Jamaica, we have additional detailed knowledge on ethnicity and occupation from Heuman (1981) and Holt (1991). We coded up all the biographical information we could find in those two sources. Appendix C Some Details on the Colonial Constitutions Of the 10 colonies that abolished their parliaments, all of them eventually switched to full Crown rule, but nine of them first changed their constitution to a transitional semi-representative system with an elected minority. Historians agree that this was a transitional mechanism with the clear end-goal of securing a majority for the full switch toward Crown Colony rule (Craig-James (2000, p. 256), Brizan (1984, p. 204)). The Colonial Office List s description of Grenada s constitutional history reads, The constitution was remodeled by an Act on October 7th 1875 and a single legislative chamber was established, [which] consisted of 8 members elected by the people and 9 nominated by the Crown. This Assembly at its first meeting on February 9th 1876 addressed the Queen that it had passed a bill for its own extinction (Britain (1879, p. 188)). Since colonies only stayed in the transitional state for six years on average, there is little informational content in that period, we focus on the first constitutional change. 36

Not for Publication Online Appendix to Elite Competition, Co-option and the Iron Law of Oligarchy: Theory and a Tale of 14 Islands

Not for Publication Online Appendix Table 1: Data Sources for Plantation Surveys Antigua 1817 Slave Registries - Antigua Montserrat 1835 Compensation Tables Antigua 1829 Johnson (1830), A Descriptive Account of Antigua Montserrat 1848 House of Commons Papers 1847-48 (399), p.116-118 e Antigua 1835 Compensation Tables Montserrat 1858 House of Commons Papers 1857-58 [2403], p.99-101 c Antigua 1843 Hart, the 1843 Antigua Almanac & Registry Nevis 1817 Slave Registries Antigua 1851 The 1852 Antigua Almanac Nevis 1835 Compensation Tables Antigua 1858 House of Commons Papers 1857-58 [2403], p.74-77 c Nevis 1878 The 1879 Leeward Islands Almanac Antigua 1878 The 1879 Leeward Islands Almanac Nevis 1897 House of Commons Papers 1898 [C.8669], p.229-232 b Antigua 1891 Hall (1971), Five of the Leewards, Appendix A St. Lucia 1852 The 1852 St. Lucia Almanac Barbados 1817 Slave Registries - Barbados St. Lucia 1897 House of Commons Papers 1898 [C.8669], p.56-57 b Barbados 1835 Compensation Tables St. Kitts 1835 Compensation Tables Barbados 1848 Barbados Almanac for 1848 St. Kitts 1847 House of Commons Papers 1847-48 (245), p.121-124 d Barbados 1854 Barbados Almanac for 1854 St. Kitts 1850 The 1850 St. Christophers Almanac Barbados 1861 Barbados Almanac for 1861 St. Kitts 1878 The 1879 Leeward Islands Almanac Barbados 1865 Barbados Almanac for 1865 St. Kitts 1897 House of Commons Papers 1898 [C.8669], p.229-232 b Barbados 1870 Barbados Almanac for 1870 St. Vincent 1817 Slave Registries - St. Vincent Barbados 1898 Barbados Almanac for 1898 St. Vincent 1827 Shephard (1831), Historical Account of St.Vincent, T.6 Dominica 1817 Slave Registries - Dominica St. Vincent 1831 Slave Registries - St. Vincent Dominica 1835 Compensation Tables St. Vincent 1835 Compensation Tables Dominica 1878 The 1879 Leeward Islands Almanac Tobago 1819 Slave Registries - Tobago Grenada 1817 Slave Registries - Grenada Tobago 1832 Woodcock (1867), A History of Tobago, Appendix Grenada 1835 Compensation Tables Tobago 1835 Compensation Tables Grenada 1849 House of Commons Papers 1849 [1126], p.180-181 g Tobago 1847 House of Commons Papers 1847 [869], p.32-33 a Grenada 1867 The 1867 Grenada Almanack Tobago 1862 Woodcock (1867), A History of Tobago, Appendix Guyana 1833 House of Commons Papers 1833 (700), p.4-11 f Tobago 1881 Craig-James (2008), Tables 5.9-5.11 Guyana 1838 House of Commons Papers 1847 [869], p.94-98 a Tobago 1894 The Trinidad Almanack 1894 Guyana 1846 House of Commons Papers 1847 [869], p.94-98 a Trinidad 1813 Slave Registries - Trinidad Guyana 1860 The Guyana Almanack 1860 Trinidad 1835 Compensation Tables Guyana 1879 The Guyana Almanack 1879 Trinidad 1882 The Trinidad Almanack 1882 Jamaica 1829 The 1829 Jamaica Almanac Trinidad 1888 The Trinidad Almanack 1888 Jamaica 1835 Compensation Tables Trinidad 1894 The Trinidad Almanack 1894 Jamaica 1840 The 1840 Jamaica Almanac Notes: House of Commons Parliamentary Papers: (a) 1847 [869] The reports made for the year 1846 to the Secretary of State having the Department of the Colonies. Transmitted with the blue books for the year 1846. (b) 1898 [C.8669] West India Royal commission. Report of the West India Royal commission. Appendix C., vol. III., containing parts VI. to XIII. Proceedings, evidence, and documents relating to the Windward Islands, the Leeward Islands, and Jamaica. (c) 1857-58 [2403] The reports made for the year 1856 to the Secretary of State having the Department of the Colonies. Transmitted with the blue books for the year 1856. (d) 1847-48 (245) Seventh report from the Select Committee on Sugar and Coffee Planting; together with the minutes of evidence, and appendix. (e) 1847-48 (399) West India colonies and Mauritius. Returns to two addresses of the Honourable the House of Commons, dated respectively 8 & 31 May 1848. (f ) 1833 (700) Slave population. (Slave registries.) Return to an address to His Majesty, dated 29 July 1833. (g) 1849 [1126] The reports made for the year 1848 to the Secretary of State having the Department of the Colonies. Transmitted with the blue books for the year 1848. Online Appendix A Supporting Evidence For Guyana, the density reported in table 1 is based on an area we calculated as Guyana s historical border using Higman (2000, figure 1.8). We geo-coded this map and calculated Guyana s sugar suitability share based on these borders. The original map and our geo-coding of it is displayed in Figure 2. Figures 4 8 show the full sequence of voting networks in Jamaica and Barbados from 1838 through to 1865. i

Not for Publication Online Appendix Figure 1: Sugar Suitability, and the Plantation System in Jamaica Notes: The top panel shows the extent of sugar plantations in 1790 (black plus grey areas) and 1890 (black areas). Grey areas are thus plantations that ceased to exist and were therefore potentially available for freeholders and squatters. The map is digitized from Higman s (2001) figure 2.9. It is the only map of its kind for any of the West Indies colonies. Parish boundaries are also shown. The middle panel shows Jamaica s topography. The bottom panel shows the distribution the suitability of land for sugar cultivation in Jamaica, from black (most suitable) to white (unsuitable), taken from Dippel et al. (2015). ii

Not for Publication Online Appendix Figure 2: Guyana Notes: The left panel shows the historical boundaries of Guyana as dashed lines. The panel is from Higman (2000, figure 1.8). The right panel shows the results of geo-coding the map. Log Price of Sugar Online Appendix Figure 3: The Secular Decline in Sugar Prices (Figure 4 in Dippel et al. (2015)) 6.0 5.6 5.2 Cholera Preferential 4.8 Tariff 4.4 phaseout, Beet Bounty, 4.0 1846-54 1884-1903 1838 1853 1868 1883 1898 1913 Notes: This figure plots the log of the London price of sugar. Two events stand out. As part of the repeal of the Corn Laws and the move to Free Trade, Britain s preferential tariff on West Indies sugar was phased out over the period 1846 54 (Curtin, 1954). Second, France and Germany subsidized domestic production of beet sugar during 1884 1903, which further drove down sugar prices. iii

Not for Publication Online Appendix Figure 4: Voting Networks in Jamaica s Parliament, 1838 1843 (top) and 1844 1849 (bottom) Sessions iv

Not for Publication Online Appendix Figure 5: Voting Networks in Jamaica s Parliament, 1849 1853 (top) and 1854 1859 (bottom) Sessions v

Not for Publication Online Appendix Figure 6: Voting Networks in Jamaica s Parliament, 1860 1862 (top) and 1863 65 (bottom) Sessions vi