Scholarly article on topic 'The price of improvements: agrarian contracts and agrarian development in nineteenth-century eastern Spain1'

The price of improvements: agrarian contracts and agrarian development in nineteenth-century eastern Spain1 Academic research paper on "Law"

0
0
Share paper
Academic journal
The Economic History Review
OECD Field of science
Keywords
{""}

Academic research paper on topic "The price of improvements: agrarian contracts and agrarian development in nineteenth-century eastern Spain1"

ECONOMIC HISTORY REVIEW íf¡i

Economic History Review, 64, 2 (2011), pp. 598-620

The price of improvements: agrarian contracts and agrarian development in nineteenth-century eastern Spain1

By SAMUEL GARRIDO and SALVADOR CALATAYUD

Fixed-rent contracts do not free landlords from the need to supervise the land if it is of high value and fragile fertility, nor do they free them from the costs of monitoring farmers if they are poor peasants prone to fall into arrears. In such cases, however, compensation for improvements will encourage tenants to farm with care and act as a bond against non-payment of rent. This article studies the repercussions of these kinds of situations by analysing what happened in nineteenth-century Valencia, where being the owners of the improvements led to tenants eventually becoming the owners of the land.

The relationships between agrarian contracts and agrarian development in modern Europe have often been analysed by economic historians.2This article also undertakes such an analysis, but from a perspective that has not frequently been used in the existing literature on the field, namely, by examining the case of an agriculture that managed to combine dynamism with family farming, small proprietors, and the widespread use of poor peasants as fixed-rent tenants.

Agriculture in Valencia (eastern Spain) was already very market-oriented in the late eighteenth century, and from about 1850 onwards displayed a remarkable capacity to improve and grow.3 In 1928 about 55 per cent of all orange exports worldwide came from Valencia,4 where one hectare provided an output that was, by value, 4.3 times higher than the Spanish average.5 Because large landowners usually resorted to tenants, however, it has traditionally been claimed that a substantial part of the productive potential ofValencian agriculture went to waste.6 Recent research has shown that this was not so, but there are many issues that still need to be explained.7

1 The authors are grateful for all the help received from Eric Chaney, Jesús Millán, Enric Morellà, Alan Olmstead, Vicente Pinilla, Paul Preston, Ricardo Robledo, the Cañada Blanch Centre of the London School of Economics, and the Institute of Governmental Affairs of UC Davis. Research support was provided by the Generalitat Valenciana, the Spanish Ministry of Education, and the Spanish Ministry of Science and Innovation (SEJ2007-60845 and ECO2009-10739 Projects). The article has improved notably thanks to the comments and suggestions of the three anonymous referees and the editors.

2 Cohen and Galassi, 'Sharecropping and productivity'; Galassi and Cohen, 'Economics of tenancy'; Garrabou, Planas, and Saguer, 'Sharecropping'; Carmona and Simpson, ' "Rabassa morta" '; Liebowitz, 'Tenants'; Jonsson, 'Paradox'; Antoine, 'La légende noire'; Mokyr, Why Ireland starved, pp. 81-111; Robledo, La renta; Carmona and Simpson, El laberinto, pp. 117-45.

3 Garrabou, Un fals dilema; Millán, El poder; Calatayud, Capitalismo agrario.

4 International Institute of Agriculture, International yearbook, pp. 292-7; Grupo de Estudios de Historia Rural, Estadísticas históricas, pp. 153, 407, 1089, 1195.

5 Simpson, 'La producción', p. 77.

6 Palafox, 'Expansión', p. 335.

7 Catalayud, Millán, and Romeo, 'Leaseholders'; Catalayud, Millán, and Romeo, 'El rentismo'.

© Economic History Society 2010. Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.

MostValencian tenants were landless people with few economic resources. Since there were no credit institutions adapted to their needs and they had to pay high interest rates when they borrowed from moneylenders,8 it would seem more logical for landlords to use hired labour or sharecropping than fixed rent.9 Why was this not the case? How did landlords overcome the problems posed by their tenants' limited liability?10 Valencian farming required strong investments and there had to be some kind of mechanism that enabled farmers to have access to credits with good conditions, but how did it work?

After investigating the reasons why landlords opted for fixed rent to farm their best land, we will analyse the economics of fixed rent in Valencia, paying special attention to the fact that tenants were the owners of the improvements they carried out on their holdings. The compensation for improvements received by outgoing tenants played an important role in ensuring the smooth performance ofValencian agriculture. However, it was also one of the main reasons why, during the first half of the twentieth century, many farmers were able to become the owners of the land they were cultivating as tenants. Since the ownership of the improvements in turn depended on the tenant's past investment, this will allow us to discuss whether land rights were sometimes endogenous in Valencia.11

The agrarian area on which this article will focus was made up of a series of coastal plains that had a climate that was particularly well suited to growing fruit and vegetable crops, and also water (from scarce but very well-managed water resources) with which to irrigate them during its hot, dry summers. To the west of this coastal strip there was a transition zone running towards the inside of the Iberian Peninsula, which consisted of a mountainous territory that could not be irrigated and whose temperatures often dropped below 0°C in winter. There the land was largely used for olive trees, carob trees, vineyards, and cereals, which only received water from the scarce, irregular rainfall of the region.

Valencian irrigated land was characterized by the presence of a wide variety of crops, none of which clearly predominated over the others (table 1). Throughout the period covered by table 1, however, some districts became specialized and an increasingly large portion of their production was destined for exportation towards the industrialized countries of Europe: from the 1850s onwards vegetables became the main crop on the Huerta of Valencia and the Huerta of Gandía,12 while from the 1870s Alzira and Vila-real were dedicated above all to growing oranges.

8 The situation only changed to any significant extent in the early twentieth century, thanks to the appearance of a powerful cooperative movement promoted by the Catholic Church. Garrido, 'Why did most cooperatives fail?'; Martínez-Rodríguez and Martínez-Soto, 'Los pioneros'.

9 In principle, if a landowner uses wage workers, he accepts all the risks associated with farming. With fixed rent it is the tenant who runs all the risks, while sharecropping represents a halfway situation.

10 If the tenants are poor and lose their harvest, they will be unable to pay a fixed rent. It is therefore the landlords who are assuming the risk, because they will have to write off payment or postpone its settlement (which is the same as granting a loan to someone who cannot offer any kind of security). This will also amount to the creation of incentives for the tenant to overinvest in risky methods of production. Ray, Development economics, p. 439.

11 As in the case analysed by Bensley in 'Property rights'.

12 'Huerta' can be translated as market-garden. The Huerta ofValencia is the name of the district surrounding the city ofValencia.

Table 1. Crops grown in the Valencian region (in hectares)

1860 c. 1890 c. 1910 1922

Unirrigated land

Cereals and legumes 245,332 186,551 233,660 282,601

Vineyards 114,265 204,493 196,851 157,741

Olive trees 62,276 73,734 66,996 90,810

Fruit trees" 107,285 98,844 155,391 155,542

Other crops 9,140 8,783 12,041

Subtotal 529,158 572,762 661,681 698,735

Irrigated land

Cereals and legumesb 81,558 78,673 86,619 94,963

Rice 26,169 26,954 28,329 33,730

Market-gardens 12,238 26,897 43,051 59,617

Fruit trees' 2,141 10,772 32,069 41,482

Other cropsJ 23,358 46,212 53,444 13,185

Subtotal 145,464 189,508 243,512 242,977

Total 674,622 762,270 905,193 943,712

Notes: a Mainly carob trees.

b In fact a smaller surface area was dedicated to them. On irrigated land, cereals and legumes were usually cultivated in rotation with vegetables and industrial plants; in order to pay less tax, when this occurred farmers very often stated that the land was used exclusively for growing cereals. This is why the categories of'market-gardens' and 'other crops' cover a greater surface area than that indicated by the figures that are available, which are of very low quality. c Mainly oranges, which could not be grown on unirrigated land. d Mainly forage and industrial plants. Source: Garrabou, Unfals dilema, pp. 90, 161-78.

In the areas that specialized in vegetables, by the late nineteenth century farmers had developed very sophisticated land-saving techniques (which included the use of nurseries and greenhouses together with large quantities of fertilizers) and the land usually yielded three (and sometimes even four) crops a year.13 The use of the land was less intensive in the rest of the coastal area, but double cropping was still a common practice. As well as requiring highly skilled cultivators, this kind of farming consumed huge amounts of labour, because it involved a lot of work (albeit on an irregular basis) and did not lend itself easily to mechanization. On the subject of Valencian farming, one scholar wrote in 1922, 'the returns are directly proportional to the precision of the farming technique and to the amount of manual labour put in by the farmer'.14

This had two important repercussions. One was that hired labour was rarely used on an exclusive basis on irrigated land. Large landowners usually made use of tenant farmers, who, like the small and middle-sized landowners, worked the land mainly with a labour force provided by the family. It must be pointed out, however, that the predominance of family farming was compatible with the existence of a powerful labour market, because farms that consisted of more than three or four hectares had to resort to complementary hired labour (which in such cases could be monitored quite easily, since farmers and wage earners worked side by side).The second repercussion was that peasant families that possessed skills and were willing to work hard had many opportunities to climb the social ladder. Although they already existed, such opportunities increased when in the late nineteenth century Valencian agriculture became more integrated in the international market. From

13 Burriel, La Huerta de Valencia, pp. 527-54; Fontavella, La Huerta de Gandía, pp. 139-68.

14 Monforte, El problema agrario, p. 150.

then on the number of small landowners, as well as that of middle-class proprietors, grew notably.15

No data are available on the distribution of land ownership for the whole region, but figures do exist for many individual places. Generally speaking, the situation was similar to that of Gandía, where, in 1887,70 per cent of the landowners owned less than half a hectare and another 15 per cent possessed between half and one hectare. Altogether, landowners with up to one hectare owned 28 per cent of all the land. At the other extreme, landowners with more than five hectares accounted for 2.7 per cent of the total number and possessed 40 per cent of all the land.16 As a peasant family needed at least a hectare of irrigated land to be able to live on its crops, access to the estates of the larger landowners (as wage workers or tenants) was essential not only for landless peasants but also for about 80 per cent of farm owners.

Alston and Higgs have found that in the southern US the more valuable the land was, the more numerous wage workers were in relation to sharecroppers, and sharecroppers in relation to fixed-rent tenants.17 To explain why this was so, they base their argument on the fact that wage workers require more intensive supervision than sharecroppers and these in turn need more than fixed-rent tenants.18 The higher the value of the land, the more the landlord supervises it in order to prevent it from being misused. The greater the supervision of the land is, the lower the marginal cost of supervising the labour will be, and this is what, according to Alston and Higgs, accounts for the strong positive correlation between the value of the land and the degree of supervision required by each type of contract.

However, the same two variables correlated negatively in the case ofValencian farming (although orange groves were an important exception, as will be explained later). The most valuable land was that situated on the irrigated coastal plains. Circa 1880, about half of the total acreage there was cultivated by fixed-rent tenants, while the other half was owner-occupied, and sharecropping was almost completely non-existent. In the poorer inland areas, in contrast, 75 per cent of the land was owner-occupied and the rest was chiefly tilled by sharecroppers.19

The tendency of wealthy landowners in the unirrigated areas to use more wage workers than tenancy contracts must have been related to the fact that agriculture there was 'easier for the owner to run', for it needed a relatively small amount of labour and farm operations were not of a complex nature.20 With regard to the predominance of sharecropping over fixed rent when tenancy contracts were used, in table 2 it can be seen that variations in output were far bigger in the interior than on the coast. Yet the choice of sharecropping by landowners does not seem to have

15 Calatayud, Capitalismo agrario.

16 Garrido and Calatayud, 'La compra', p. 87.

17 Alston and Higgs, 'Contractual mix', p. 346.

18 However, this is not always the case. There have been reports of situations in which the sharecroppers make all the decisions with hardly any supervision; Allen and Lueck, 'Transaction costs', p. 79. In contrast, during the eighteenth and early nineteenth centuries Valencian landlords often gave their fixed-rent tenants very precise instructions; Catalayud et al., 'Leaseholders', pp. 153-7.

19 Reformas Sociales, pp. 122, 176, 308; Crisis agrícola y pecuaria, p. 526; Dirección General de los Registros, Memoria, vol. IV, pp. 66-8; Costa, 'Aparcerías'. There could be, however, important differences between one district and another: for example, 85% of the irrigated land in the district ofValencia was cultivated by tenants, but only 35% in the district of Vila-real.

20 Crisis agrícola y pecuaria, p. 525.

Table 2. Yields and output of selected crops on irrigated and unirrigated land in the

province of Castellón (1903-12)

Yield (qm/ha) Yield: coefficient of variation" Output

On unirrigated land

Wheat 8.6 29.7 33.1

Wine 15.9 45.3 15.0

Olives 6.2 93.2 11.0

Corn 7.6 38.3 2.0

On irrigated land

Wheat 24.9 17.5 10.4

Corn 26.4 17.8 3.0

Notes: a Standard deviation as % of mean.

b As % of total agricultural output, by value, on unirrigated and irrigated land. Qm: metric quintals.

Sources: Palacios, Servicio, pp. 9-12; Sarthou, Geografía, pp. 285-6.

been determined mainly by the desire to share the relatively high risk of losing the harvest: although variations in output were also very large in the neighbouring regions of Castile and Catalonia, fixed rent predominated in the former, while sharecropping was more common in the latter.21 The use of sharecropping was quite frequent, however, in the vineyards of both Castile and Catalonia, as in the other wine-producing areas of Spain, and a substantial part of the land under sharecropping in Valencia was devoted to vineyards.22 This raises the question of why the large Valencian landlords who made use of sharecropping in the interior always swapped over to fixed rent on their irrigated estates on the coast.

In studies on contemporary Asia, it has been found that the rent per hectare is usually lower under fixed rent than under shared tenancy. To explain why some landlords opt for the former, Hayami and Otsuka suggest that landlords for whom the cost of enforcing workers' effort is low tend to prefer shared tenancy, while those with high enforcement costs tend to choose fixed-rent tenancy, at the expense of a lower rent that compensates the tenants for a greater exposure to risk. As pointed out by these authors, this is consistent with the 'common observation' that small to medium-sized landlords, who know their tenants personally, usually opt for sharecroppers, whereas the large absentee landlords prefer fixed-rent tenants.23 In the irrigated areas ofValencia, however, practically all the landlords chose fixed rent, regardless of whether they were large landowners who did not reside within the community, resident lawyers with no experience of farming, or resident former peasants who had managed to become well-off and were capable of distinguishing between well and badly-done farm work.

21 On Castile, see Robledo, La renta; Carmona and Simpson, El laberinto, pp. 117-45. On Catalonia, see Garrabou et al., 'Sharecropping'.

22 On Valencia, see Piqueras, La vid, pp. 157-67. On Castile, see Carmona and Simpson, El laberinto, pp. 149-52. On Aragon, see Sabio, Viñedo, p. 174.

23 Hayami and Otsuka, Economics of contract choice, pp. 97-8.

If the presence of interlinked markets seems to be a very important requirement for the successful resolution of incentive and monitoring problems associated with sharecropping,24 such a requirement was fully satisfied in Valencia. The landlords in this region (as we shall see later) played a crucial role in the credit market, and it was a common practice for them to utilize their tenants' wives and daughters as wet nurses for their children and as domestic servants in their homes. This frequently led to the creation of strong personal ties and, for example, it was customary for leaseholders to attend their landlord's funeral. So, why did landlords use fixed rent and not sharecropping?

In 1879 the Valencia Board of Proprietors (Liga de Propietarios) stated that on irrigated land 'tenants found sharecropping repugnant',25 and one scholar who specialized in this topic insisted on this same idea, writing in 1922: 'Since the returns from irrigated land depend almost exclusively on the personal, manual labour of the farmer, sharecropping would hardly be a stimulus for the tenant, who prefers to pay a fixed price and to obtain all the proceeds from his efforts'.26 The argument is a version of what is usually called the 'Marshallian model':27 since the prosperity of the irrigated land in Valencia was based on the use of huge amounts of family labour, the peasant considered that entering into a share contract meant 'giving away' part of that work to the landlord. Yet it is surprising that all tenants chose fixed rent and none of them opted for sharecropping. The so-called 'screening model' predicts that the more able peasants will tend to opt for fixed rent, whereas those with less entrepreneurial or other skills will opt for sharecropping. For this to take place, however, the landlord must offer them a set of contracts to choose from, with the aim of screening the peasants according to their ability.28 Because there were only fixed-rent tenants on the irrigated lands ofValencia, it can be supposed that in fact the landlords did not offer any kind of choice and that the question really boils down to one of why landlords preferred fixed rent.

When landlords came to the conclusion that fixed-rent tenancy had to be abandoned, they did not replace it with sharecropping but instead with hired labour, which means that their preference for fixed rent cannot be a result of the existence of aversion to risk.29 Given the abundance of peasants who possessed small plots, with sharecropping the landlords would have had to face the risk of tenants' concentrating their efforts on cultivating their own land and using the parcels they farmed under share contracts as a complement.30 Even if landless peasants were chosen as tenants, this danger would only disappear if all the landlords opted for sharecropping, because multiple contractual arrangements were quite common and tenants would have continued to focus their efforts on the plots they farmed under fixed rent. However, exclusive use of sharecropping would result in a series of problems.

24 Bardhan, 'Interlocking factor markets'; Hayami and Otsuka, Economics of contract choice, pp. 70-84.

25 Las Provincias (Valencia), 18 Feb. 1879, p. 1.

26 Monforte, El problema agrario, pp. 5-6.

27 According to which, as the sharecropper receives only a fraction of his marginal product of labour, sharecropping leads to inefficient labour input decisions.

28 Singh, 'Theories of sharecropping.' pp. 56-66.

29 Calatayud et al., 'El rentismo', pp. 86-98; Millán, 'Triunfo y límites de la propiedad', p. 405; Sanmartín, 'El patrimoni', pp. 71-5; Garrido, Canem, pp. 110-13.

30 As has been observed in India today by Shaban, 'Testing'.

Hectares

Figure 1. The relation between the number of hectares and the number of parcels owned by 688 landowners in Vila-real in 1888 (each circle represents a landowner)

Sources and notes: Archive of the Irrigation Board (Comunidad de Regantes) of Vila-real, Vila-Real, 'Padrón de 1888'. See the explanation in n. 34.

Tenants usually worked the land of more than two landlords.31 At the same time it was not unusual to find landlords such as Florentino Márquez (who used 147 different tenants to work 106 hectares) and the Marquis of Jura Real (who used 130 different tenants to farm 73 hectares.32This resulted in a situation in which the land tilled by each peasant family was usually divided out over several plots that were often separated by considerable distances. In fact, it may well have been a social strategy that was essentially aimed at achieving just that. One drawback of the plots being so dispersed was that the farmers wasted a lot of time travelling from one place to another. However, it also had two well-known advantages. Firstly, it lowered the risk of losing the whole crop as a consequence of some meteorological disaster or pest. Secondly, the amount of land a peasant family could till was larger than when it was concentrated in a single plot.33 Figure 1 shows the relation between the number of hectares belonging to each landowner and the number of plots they were divided into. As can be seen, it suggests that having access to the second of the advantages we have just mentioned was something that was held in particular esteem by small and middle-sized farmers.34 Whatever the reason was for dispersing the plots, the important point to highlight is that if only sharecropping had been used some tenants would have attempted to sign contracts with many landlords, in order to increase the surface area they worked up to the point where the marginal product of the land was zero. Overall, in this way the sharecropper would obtain more output than by concentrating his

31 See tab. 3.

32 Municipal Archive of Vila-real, Vila-real (hereafter MAV), Padrón de 1859; Municipal Archive of Gandía, Gandía (hereafter MAG), Amillaramiento de 1887.

33 Irrigation determined the schedule of many of the jobs to be carried out on the land, which had to be completely stopped during the days immediately after the plot had been watered. Exactly when each plot was to be irrigated, however, only partly depended on the farmer's will, as it was necessary to wait one's turn. If a farmer had multiple plots it was easier to have each of them irrigated at different times, and consequently to have the family workforce employed all the time.

34 Figure 1 refers to the irrigated land in Vila-real in 1888, when it was divided among 2,968 owners. It has been drawn up from a sample of 688 proprietors. Initially, the more land they had, the greater the number of plots would be, but the number of plots diminishes again in the case of the larger landowners, who necessarily had to use wage workers or tenants. If, instead of optimizing the use of the family workforce, the primary aim of dispersing the plots had been to lower the risk of losing the crop, the result of the regression would have been a linear function (but R2 = 0.3822 for a linear function).

work and his capital on farming fewer plots, but at the expense of reducing the harvest obtained from each plot and, in consequence, the amount of crop received by each of the landlords.

Sharecropping did predominate, however, in other European regions where multiple contractual arrangements were also commonplace,35 but in the irrigated areas of Valencia the landlords would have had to incur especially high enforcement costs. In accordance with a proposal by Hayami and Otsuka, which seems to be full of common sense, agricultural production technology is a major determinant of contract enforceability: 'In general, the more complex and less standardised farm operations are, the more difficult it is for landlords to monitor tenants' work efforts'.36 In the irrigated lands ofValencia the skills that were used were very difficult to master, the peasants' know-how played a crucial role, and the work was usually applied on a discontinuous basis. As has already been mentioned above and in contrast to what Alston and Higgs suggest, here the higher the value of the land was, the less common it was to make use of contracts requiring more supervision of labour. The value of the land, however, depended on the types of crops the land could be used for, and these crop characteristics are what in fact seem to be acting as an explanatory variable.37

In the late nineteenth century one large Valencian landowner, the Count of Montornés, weighed up the advantages and disadvantages of each type of agrarian contract, put his conclusions into practice on one of his own estates, and then described his experience. The estate consisted of 1,000 hectares, 120 of which were irrigated land, 400 were unirrigated land, and the rest was a pine forest. Cultivation was the responsibility of 70 families who had to live as tenants in houses distributed throughout the estate. There was a school, a chapel (where the peasants were obliged to attend services), a slaughterhouse, a bakery, and a company store, among other facilities. There was also a Provident Society that was partially funded by the Count himself. This information is important because it shows that the relationships that the proprietor sought to develop with the peasants was one of a paternalistic nature and that he was able to exercise perfect control over the behaviour of the dwellers in the property.

The experiment consisted of making each peasant a wage worker, sharecropper, and fixed-rent tenant at the same time, depending on 'the classes of land and crops'. They were used as hired labourers to work in the forest, on the upkeep of infrastructures and in the making of oil and wine in the olive oil mill and the winery of the estate. Sharecropping was utilized in the case of the vineyards, olive trees, and carob trees; that is to say, 'in the large extensions [of unirrigated land] devoted to regular crops ... to be harvested at normal times of the year'. Lastly, each peasant received between one and two hectares of irrigated land under fixed-rent tenancy. The peasant was as closely bound to the landlord's rules and regulations when he acted as a sharecropper as when he was a wage worker, so that the sharecropper was deprived of practically any capacity to take the initiative.

35 Galassi and Cohen, 'Economics of tenancy', p. 589.

36 Hayami and Otsuka, Economics of contract choice, p. 99.

37 Depending on the zone, the same crop could have completely different characteristics. For example, on the irrigated land wheat provided per-hectare yields comparable to those ofNorth-Atlantic Europe, and it was usually included in rotations together with vegetables, which made it possible to achieve more than one harvest a year. On the unirrigated land, its yields were very low and fallow was used.

However, the Count placed great emphasis on the fact that when the cultivator acted as a fixed-rent tenant he was 'absolutely free' to make decisions regarding farming.38 The conditions under which the Count of Montornes conducted his experiment were exceptional, but that is precisely what makes the case so interesting. The monitoring mechanisms used in the colony were so severe that none of the tenants could ever hope to shirk or cheat and get away with it, but the landlord still came to the conclusion that on irrigated land it was a better idea to use fixed rent instead of sharecropping.

Fixed-rent contracts freed owners from the need to watch over tenants, but they did not free them from the need to watch over their land so that it was kept in good condition and so that, if possible, its selling price increased as a result of having been improved. Indeed, Valencian landlords accorded a great deal of importance to being able to use landed property to obtain cash (either by sale or by mortgage) should they need to do so.39 The tenant could cause depreciation of the farm if he depleted the soil or allowed the irrigation channels and buildings to deteriorate. On the other hand, he could improve it by building greenhouses, adding sand to clayey soils, or levelling the ground better to make irrigation more efficient. There were undoubtedly fewer problems attached to supervision carried out to prevent misuse of the land than to supervision to ensure its improvement. Both aspects were very much connected to the length of the contract, but in a complex way.

The question of whether tenants are more likely to carry out improvements with long leases than with short ones has been discussed on numerous occasions.40 However, the crux of the matter does not lie in the length of the contract, but rather in the incentives the tenant is offered to make improvements. Long contracts are in themselves an incentive, but often an insufficient one. According to what is known in game theory as 'an eye for an eye' strategy, it can be supposed that if neither landlords nor tenants had known when the relationship between them was going to end, and if both sides were able to take some kind of retaliatory action if the other side stopped collaborating, then tenants would have tended to carry out improvements from the outset and landlords would have given them stability so that they could enjoy their investment.41 With long contracts, initially tenants will be willing to make improvements, but since they know when the relationship will end, they will tend to impoverish the land as that moment draws near—unless the landlord takes the decision at that time to boost incentives.42 With

38 The quotations are from Conde de Montornés, 'Ensayo de explotación agrícola y social', pp. 136, 138. Information has also been drawn from Conde de Montornés, 'Ensayo de explotación agrícola-social'. Both works were written for the Ninth International Congress on Agriculture (Madrid, 1911), whose Organizing Commission was presided over by the Count.

39 Calatayud et al., 'El rentismo', p. 96; Pons and Serna, La ciudad, pp. 333-62.

40 For example, Solow, Land, pp. 77-85, and Mokyr, Why Ireland starved, pp. 81-111 (for Ireland); Whyte, Agriculture, pp. 137-72 (for Scotland); Chambers and Mingay, Agricultural revolution, p. 46 (for England).

41 Axelrod, Evolution, pp. 109-23. If certain conditions are fulfilled, a person tends to collaborate the first time he comes into contact with the other, and then imitates the behaviour adopted previously by the other party.

42 To encourage good stewardship, in some parts of England the landlords paid for the improvements carried out by the tenant during the two years prior to the expiry of the contract; Pigou, Economics of welfare, p. 178.

short leases that are not accompanied from the outset by a strong incentive, tenants are likely to impoverish the land right from the start and it is almost certain that they will never improve it.

Landowners in the Valencian region offered their tenants short contracts (85 per cent of the agreements drawn up by deed had a duration of six years or less).43 However, they were usually extended verbally from year to year. This allowed landlords to recover the land almost straight away whenever they wanted to, and at the same time it was a way of encouraging the tenant not to farm the soil excessively or in a negligent manner. However, as a rule, it was not enough to get the tenant to invest labour and capital in improving the leasehold. The only person who could decide when to stop renewing the annual contract was the landlord, and it was much easier for a landlord to take retaliatory action (that is, eviction) against a tenant who failed to fulfil his duties than vice versa. Furthermore, the improvements could have negative repercussions on the cultivator, because they might lead the landlord to behave in a non-cooperative manner. This is what happened to one character in La barraca (The cabin), a novel by Vicente Blasco-Ibanez about the Huerta of Valencia published in 1898. When his landlord made him choose between accepting a substantial increase in the rent or leaving, he 'protested and even cried as he recalled the merits of his family, who had given their all in those fields to make them the best'; but the landlord was unyielding: 'So they were the best, were they? . . . Then you must pay more'.44

It is sometimes argued that landlords would not want to confiscate their tenants' improvements, because future tenants would no longer carry out any improve-ments.45 However, some landlords did display predatory behaviour, and a fair landlord might be replaced by a predatory one if the farm were sold or inherited. In consequence, two requirements were essential for improvements to be made: the tenant should be offered some kind of reward in exchange, and he should have a solid guarantee that he was in fact going to receive that reward. Repaying the tenant for the money and the effort he has spent in ameliorations is usually thought to be one of the most effective procedures to fulfil the first requisite.46 This is what happened in the Valencian region, although compensation for improvements was not compulsory under Spanish law.47 When a cultivator left his farm, two valuers chosen by the parties calculated the amount the tenant was to be reimbursed for 'any increase in the value of the farm that could be attributed to his endeavours'.48 Since it was easier for a tenant to take reprisals against a fellow peasant than to retaliate against a landlord, the second requisite was satisfied by the fact that the

43 Pons, 'Els contractes', p. 183.

44 Blasco-Ibanez, La barraca, p. 29. Cases like this were uncommon in the irrigated lands ofValencia, but they did occur. Monforte, Elproblema agrario, pp. 12-13, and Modesto, 'Crisis', pp. 102, 217, provide some examples.

45 Guinnane and Miller, 'Limits', p. 594; Liebowitz, 'Tenants', p. 438. In contrast, Mokyr, Why Ireland starved, pp. 85-6, considers the landlord's increasing the rent a logical step after the leaseholder had carried out improvements—'the question is, however, by how much the rent is raised'.

46 Pigou, Economics of welfare, pp. 174-83.

47 From medieval times to 1889, the law stated that improvements should be compensated, but allowed landlords to contract out. Thus, in many Spanish regions, such as Castile, the usual practice was not to reimburse the tenant. From 1889 onwards the Spanish tenant no longer enjoyed any legal right to compensation. Spain therefore followed a course that ran in the opposite direction to that undertaken in England and Wales, where tenants' right to be compensated was first recognized by law in 1875, and compensation became compulsory as of 1883.

48 Monforte, El problema agrario, p. 14.

Figure 2. Joint debt of the 19 fixed-rent tenants of Vicente Calatayud in Beniarjo, as a percentage of the total annual income (current income plus arrears) expected by the landlord

Source: Archive of the Kingdom ofValencia, Fondo Calatayud-Enriquez de Navarra, 27, 3-15.

outgoing cultivator was not usually reimbursed by the landlord, but instead by the incoming tenant. Moreover, in some districts the peasant community was sufficiently united to be able to take reprisals of a collective nature against any tenant that sought to act in an opportunistic manner. Such reprisals 'have often stained the fields ofValencia with blood',49 and La barraca (the novel by Blasco-Ibanez mentioned above) tells the story of how the community ostracizes a peasant family that, driven by poverty, dares to work the land of a predatory landlord. It is interesting to note that the peasants in the Belgian and French regions where improvements were settled on a much stricter basis also acted in a similar way.50 Thanks to compensation for improvements, tenants were more motivated to work well and invest, while landlords saved on supervision costs and lessened the danger of their property losing value. Furthermore, improvements acted as a bond against non-payment of rent, as in the case of the Irish tenant-right, analysed by Guinnane and Miller.51

Although Valencian landlords did not rack-rent their tenants (so that during the good years tenants were able to keep a substantial part of the output for them-selves),52 under no circumstances were they usually excused from paying the rent. However, landlords did tolerate arrears (which were in fact very frequent) and would usually allow a long period of time to elapse before taking steps to evict the farmer who failed to pay. 'It is customary', wrote the British Vice-Consul in Valencia in 1869, 'to allow tenants to redeem their arrears'.53 Figure 2 illustrates one of these situations. It shows the arrears tolerated by one landowner who had

49 Ibid., p. 16.

50 See Rowntree, Land and labour, for Belgium. For France, see Prothero, 'Tenant-Right'; Depressed condition (P.P. 1882, C.3375-V), pp. 78-82.The Irish tenant-right was also paid from tenant to tenant, although according to Guinnane and Miller, 'Bonds', p. 119, it cannot always be linked to payment for improvements.

51 Guinnane and Miller, 'Bonds'.

52 Calatayud et al., 'El rentismo', pp. 97-8.

53 Reports (P.P. 1871, C.271), p. 54.

PRICE OF IMPROVEMENTS 609

Table 3. Landless peasants with fixed-rent contracts on irrigated lands in Valencia

Tenants who were landless peasants"

Rented hectares tilled by landless tenantsb

Average number of different landlords a landless tenant had

Alboraia (1828) Vila-real (1859) Alboraia (1860) Alzira (1861)

48.4 79.3 60.2 63.7

72.0 32.9 75.6

55.1 65.9 53.4

n.a. n.a.

Gandía (1887) Castellón (1910)

2.6 2.4

Notes: a As % of the total number of tenants. b As % of the total number of rented hectares.

Sources: Felix, L'estructura, p. 10; MAV, Reparto de guardería rural de 1859; Municipal Archive of Alboraia, Alboraia (hereafter MAA), Padrón de 1860; Romero, Propiedad, p. 280; MAG, Amillaramiento de 1887; Municipal Archive of Castellón, Castellón, Reparto de guardería rural de 1910.

19 tenants to farm 7.9 hectares. Over the 15 years that are considered here, the landlord always used the same 19 tenants, and all of them slipped into arrears at one time or another.

Two lawyers who were very familiar with such matters stated that landlords acted in this way because they were sure 'to get paid, although not on time, the rent they were owed, because the incoming tenant had to pay any amounts left outstanding by the outgoing tenant'.54 As can be gathered from the numerous references that appear to this respect in the written contracts, what really happened was that, if a tenant was evicted for not paying the rent, the landlord made up for the debt by confiscating his improvement rights, which were then resold to the incoming tenant.55This allowed landlords to give competent tenants they were interested in keeping on a certain amount of leeway with regard to arrears, and it also lowered the chances of poor harvests being followed by a period of deficient farming that reduced the selling price of the land.

Furthermore, the fact that landlords allowed arrears also served to provide the farmers with financial resources—something that was of vital importance in an agriculture that required heavy investment but was frequently run by poor tenants. As table 3 shows, it was quite common for landlords to have to resort to landless peasants as tenants. Valencian historians have traditionally interpreted the frequency with which they incurred arrears as meaning that they were in an almost permanent state of financial difficulty. Sometimes, however, the tenant deferred payment of the rent so as to be able to carry out an investment.56 On other occasions he was temporarily short of cash because he was holding back the harvest until its selling price rose. An incoming tenant could borrow from mon-

54 Burriel and Oller, 'El problema de la tierra', p. 118.TheVice-Consul in Valencia had said 60 years before that 'there is not sufficient land for the number of labourers, which is the reason that men are so easily found ready to pay premiums to buy out tenants, and pay the arrears of others, in order to lease the land'; Reports (P.P. 1871, C.271), p. 54.

55 Several different variants of this procedure also existed. For example, in one contract signed in 1870 the incoming tenant took responsibility for the arrears of his predecessor, but in return he was given ownership of the improvements. Another contract signed for four years in 1884 was renewed by parol for the next two years, but in 1890 it was rewritten to include the fact that the tenant transferred his improvement rights to the landlord to settle a debt of 1,001 pesetas. MAG, DA-383, 1870, fo. 246; and A9-51, 1890, fo. 154.

56 Calatayud et al., 'Leaseholders', p. 160.

eylenders in order to obtain the money needed to compensate his predecessor for improvements and then refinance his debt with the arrears in the rent payments that his landlord allowed him to accumulate. In all these cases, then, the tenants utilized the improvement rights as collateral to obtain an interest-free loan from their landlords.57

The international literature on agrarian contracts has paid little attention to the question of compensation for improvements, which seems to indicate that relatively few places followed this practice to the same extent as in the Valencian region. This raises an inevitable question: if in Valencian farming compensation for improvements was apparently a source of advantages for both landlords and tenants, why were landlords everywhere not willing to pay this compensation?58

The answer is almost certainly that landlords were aware that, in exchange for these advantages, paying compensation entailed three threats: tenants would tend to gain stability and it would become increasingly difficult to change tenant farmer; the rent would tend to increase more slowly than the price of the land; and tenants might end up becoming owners of the land. With regard to this last possibility, when the issue of compensation for improvements was discussed in the French parliament in 1850 when member of parliament M. Morellet said that, since a tenant with the right to claim for compensation became 'almost an owner', compensation for improvements was 'a direct and powerful blow to property rights': 'This amounts to acknowledging that the farmer can do with the landlord's property as he wishes, transforming it and modifying at his will, and if the landlord fails to pay the added value, he will have his property expropriated by his own farmer'.59 This kind of problem became very apparent in the Huertas ofValencia and Gandía, the two districts that from the mid-nineteenth century onwards specialized in market-gardening.

Growing vegetables was already quite important in these districts from at least the mid-eighteenth century onwards. However, prior to the 1850s vegetables were just one of the items (together with wheat, corn, and beans) in a rotation in which the main product was hemp. When hemp fell into decline, vegetables became the core crop of the rotation. In the Huerta of Valencia this process was helped by strong demand from the nearby city. The Huerta of Gandía enjoyed a warmer microclimate that allowed tomatoes or peppers to ripen a few weeks earlier than in other coastal districts, and this meant that they could be placed on the market when prices were higher. Both places had the infrastructures required to export perishable goods to Great Britain.

The fertility of the market-gardens was very fragile and could quickly deteriorate if farming was not carried out in a suitable manner. Furthermore, specializing in

57 Castellanos and Fernández, 'Amos', p. 224, quote a contract from 1873 in which the tenant undertakes to pay an annual interest of 5% until he has paid off the arrears left by his predecessor. This is a rather exceptional situation that illustrates the fact that landlords granted interest-free loans while they tolerated arrears.

58 For example, from the numerous testimonies collected by the Royal Commission of 1894-7 it appears that, in the late nineteenth century, a substantial number of English landlords were reluctant to pay compensation for improvements; Agricultural Depression (P.P. 1897, C.8540), pp. 90-8.

59 'C'est déclarer que le fermier dispose de la chose du maître, qu'il la transforme et la modifie à son gré, et que si le maître ne rembourse par le montant de la plus-value, il sera exproprié de sa chose pour son propre fermier', quoted by Congost, 'Sagrada propiedad', pp. 66-7.

Figure 3. Percentage of leased land where the tenant changed annually in Gandía (100 = total land area where tenants changed between 1864/5 and 1891/2)

Sources and notes: MAG, Apéndices al amillaramiento trasladando los colonatos (1864-92). In I860, 75% of the irrigated land in Gandía was tilled under leases. The land where the tenant changed in 1864/5 accounted for 6% of the total amount of leased land in 1860.

them meant cultivating a great variety of products that required a wide range of technically complex tasks. As one British Consul explained in 1888:

The Huerta of Gandía is really more dedicated to market-gardening than to agriculture, as usually so understood. Only where irrigation is difficult or expensive is the land dedicated to other cultivation than that of vegetables. These are produced on a very large scale for export, and, thanks to the combination of heat and moisture, as many as three crops in succession are often grown on the same land during the year. Thus, tomatoes may be followed by melons, and these again by Indian corn. The early crops most grown are tomatoes, pimientos (Capsicum annuum), and French beans. The two former are first planted in hot-beds, carefully covered up, and afterwards planted out under the lee of ridges, with an additional shelter of rice-straw.60

Since a market-garden was almost impossible to run if the farmer did not have a high degree of autonomy, from the mid-nineteenth century onwards landlords stopped giving instructions about how much fertilizer and farm work had to be applied, and about the crop rotations tenants had to use. They also stopped investing in improving their land. They no longer legally formalized the contracts, automatic annual renewal became increasingly popular, and changes of tenant ended up being rather exceptional.61 This can be seen in figure 3: as specialization in market-gardening became more pronounced in Gandía during the second half of the nineteenth century, it became less and less common for tenants to leave the farm. Figure 4 shows that 37 per cent of the plots subject to fixed-rent contracts

60 Foreign Office, Annual Ser., no. 275, Diplomatic and Consular Reports (P.P. 1888, C.5252-52), p. 2.

61 Modesto, 'Crisis', pp. 158-74; idem, Aús i costum, pp. 115-29.

■ Duration stated in the written contracts □ Actual length of time the tenant remains on the land

Figure 4. Initial intended duration and actual duration of fixed-rent contracts in Gandía (1862-92)

Source and notes: MAG, Notary records and Apéndices al amillaramiento trasladando los colonatos. See n. 62.

in 1862 continued to have the same tenant for at least the next 30 years;62 furthermore, most of the changes that did take place were due to an old tenant being replaced by a member of his family.

Because the investments they made to improve the land were very large, the amount that had to be paid to outgoing market-gardeners by their landlords was also very high (it was not unusual for it to account for half the value the holding would have reached on the market if it had been sold free of tenant).63 Contemporaries considered this to be the main reason for the enormous stability enjoyed by tenants in the Huertas ofValencia and Gandía:

When a landowner tries to evict a tenant on expiry of the lease, the improvements this tenant has carried out are evaluated by experts and, because more importance is almost always given to the value of the improvements than to the value the holding would have without them, the owner has no choice but to leave the tenant to tend his farm.64

Nevertheless, there was a second factor involved—the collective action of cultivators in order to ensure that the compensation for improvements would always be paid. Tenants did not consider that a long occupation of the holding granted them any property right over the land. If they were requested to leave, they left without offering any kind of resistance provided that their successors or the landlord made it worth their while, that is to say, when the incoming tenant or the

62 In Gandía there is a source of information that does not appear to exist anywhere else in Spain: a series of registers where, between 1862 and 1892, local council clerks noted the changes of tenant that took place in the plots that were farmed under a fixed-rent agreement. We selected a sample of 150 plots (which were actually 162 in all, due to the fact that some of them were later divided out among more than one cultivator) and we analysed how many times they changed tenant during the next 30 years. Figure 4 is based on the results from that sample and on a review of all the lease contracts drawn up by the notaries in the town in 1850, 1860, 1870, 1880, 1890, and 1900.

63 Monforte, El problema agrario, p. 15; Burriel and Oller, 'El problema de la tierra', p. 118; Garrido-Juan, El

arrendamiento, pp. 46-7, 60.

64 Morán, 'El colonato', p. 39.

landowner bought the improvements from them.65 However, if they were ejected without receiving any compensation, 'there is a general, tacit agreement among tenants not to become leaseholders of land from which another [tenant] has been evicted, and it is often said that it will end up overgrown with weeds'.66

In 1878 the landlords got together and attempted to redirect the situation in their favour. In response, tenants declared themselves on a 'farmers' strike' (huelga de labriegos) and refused en masse to pay the rent.67 The revolt ended after 76 people were deported to the island of Minorca, but it strengthened the tenants' position, since they gained even more stability. The Valencian custom prevented a farmer from having to pay a higher rent as a consequence of the improvements that he himself had carried out,68 but moreover landlords found themselves deprived 'of the right to dispose freely of their properties, thus making it impossible to raise the rent'.69 Indeed, while families of tenants stayed rooted on 'their' farms, the rent could remain unchanged in nominal terms for 30, 40, or even more years.70 Many witnesses claimed that: 'Property rights are not being respected'.71

In the late nineteenth century orange trees became a more profitable crop than vegetables.72 For landlords they offered the advantage that, unlike vegetables, they were not particularly difficult to cultivate using hired labour. The shift into oranges, however, was often very complicated for the landowners of the Huertas who had previously specialized in market-gardening. Fixed-rent contracts and liberal compensation for improvements had boosted agrarian development. Yet compensation for improvements had also created a path dependence from which it was not easy to escape. Although past investments only granted the tenant property rights over the effects of those investments (and not over the land), in practice landlords had to 'repurchase' their land if they wanted to be free to use it as they pleased. It is to be expected that whenever paying liberal compensation for improvements is compulsory, the same thing will tend to happen. Therefore it is of interest to establish whether this resulted in a loss of capacity of Valencian agriculture to adjust land-use in response to the changes in relative prices. This is the subject of the next section.

65 Orellana, El colonato, pp. 14-15.

66 Bernabé, Discurso, p. 13 (original emphasis).

67 Cucó, Republicans, pp. 15-143; Millán, 'Triunfo y límites de la propiedad', pp. 396-409.

68 'The capital represented by the improvements is not usually included in the sum paid by the tenant when he pays the rent'; Monforte, El problema agrario, p. 16. In his well-known discussion of the inefficiencies involved in landlord-tenant relations, Pigou claimed that the law should grant the outgoing tenant the right to be compensated for improvements. He went on to argue that, even in such a case, the tenant 'knows that the rent may be raised against him on the strength of his improvements, and his compensation claim does not come into force unless he goes to the extreme length of giving up the farm'. To overcome this 'imperfection', he concluded that 'not merely compensation for tenants vacating their holdings, but . . . legal prohibition of renting tenants' improvements is required'; Pigou, Economics of welfare, pp. 179-80, 181. What Pigou is calling for is the system used in the Valencian region without the need for any kind of legal prohibition.

69 According to the Land Registrar (Registrador de la Propiedad) of the city ofValencia. Dirección General de los Registros, Memoria, IV, p. 64.

70 Tirado, 'Gran propiedad', pp. 116-31; idem, 'Salvador Castillo', pp. 230-5; Sebastià and Piqueras, 'Per-vivencias feudales', pp. 237-74.

71 Orellana, El colonato, p. 14. Similar ideas are expressed in Bernabé, Discurso, pp. 10-15, and Aparisi, Consideración especial, pp. 31-51.

72 Palafox, 'Estructura de la exportación'; idem, 'Expansión', p. 328; S. Garrido, 'The "Anticalifornia". Family farming, prices, and quality product in the Spanish citrus industry (1840-1936)', Asociación Española de Historia Económica, working paper 0701 (2007), pp. 6-7.

Table 4. The orange in the Vakncian Region

I II III IV

1872 3,546

1878 8,225 2.0

1900 22,125 2.2 4.9

1910 37,310 78.7 2.6 6.1

1922 36,868 78.9 7.6

1926 40,398 77.4 13.4

1929 54,853 82.6 11.3

1932 62,525 83.8 9.6 23.2

1935 64,136 83.5 17.7

Notes: I: Hectares devoted to oranges trees in the Valencian Region. II: I as % of Spanish total.

III: Contribution of the orange to the Spanish agrarian product (as %). IV: Share of the orange in total Spanish exports (in value, as %). Source: S. Garrido, ' "Anticalifornia"(see above, n. 72), pp. 11-12.

Oranges were gradually becoming more important in the Valencian region from the 1830s onwards, but they did not really take off until the 1870s, as can be seen in table 4. It must be taken into account that in one part of the Valencian irrigated areas it was not possible to cultivate oranges (or growing them was a very troublesome business) either because the soil was excessively clayey or owing to the climate.

Creating an orange grove required heavy investment, but then orange trees needed less work per hectare than the crop rotations carried out on irrigated land. Since the attention and care required by a grove were quite standardized, they were not difficult to monitor.73 If hired labour was used and a worker mistreated a tree, the landowner (or his foreman) would immediately find out about it. Oranges were much easier to market than vegetables, as they were always bought, while still on the tree, by traders who were responsible for picking, storing, and exporting them.74 In contrast to what happened in the market-gardens (where organic fertilization continued to be essential to enrich the soil), in the orange groves there was an early switch to mineral fertilizers—and therefore to a capital input, instead of the very labour-intensive methods used to collect manure and organic material. All this meant that, in the vast majority of cases, oranges were cultivated by wage workers (or by means of family work when they were owner-farmed by small-scale growers).75

Since a plot of land with full bearing orange trees could be sold for a great deal more than neighbouring plots given over to other crops, in this particular case

73 This is the reason why some absentee landowners who attempted to work their orange groves using agrarian contracts opted for sharecroppers instead of fixed-rent tenants; Lassala, Memoria, p. 8; Garrabou, Unfals dilema, p. 143.

74 Unlike vegetables, which were harvested on a very irregular basis by the farmer, who even sold them directly to consumers (using the womenfolk in the family) when they were destined for nearby urban markets. This provides a supplementary argument explaining why fixed-rent tenants were used in market-gardens instead of sharecroppers: sharecropping would have left landowners with large quantities of highly perishable vegetables and the need to sell them.

75 Maylin, Memoria sobre el arbolado, p. 30; Bou, Estudios, p. 204; Font de Mora, El naranjo, pp. 19-23; Garrido, Canem, pp. 110-13, 117-26.

Alston and Higgs's explanation about the relationship between the value of the land and the use of hired labour does seem to work.76 However, this is because, as has been stated earlier, the exogenous variable is not the value of the land itself, but instead the characteristics of the crops being grown on that land: land that had been planted with orange trees had the highest value precisely because it was planted with orange trees, and the landowner thus kept a close watch over his trees. Supervising the work carried out on orange groves therefore had a low marginal cost, which made the use of hired labour an attractive proposition.

For orange trees to be planted, however, the land and its improvements usually had to stop belonging to different people. This was always accomplished by means of a sale, but sometimes it was the tenant who sold the improvements to the landlord and other times it was the landlord who sold the land to the tenant.

In the districts that had not previously specialized in market-gardening, taking the land in hand and planting orange trees was a simple affair for landlords who wanted to do so, as the compensation for improvements they had to pay was a relatively small amount. In the nineteenth century, however, some absentee landlords preferred to have their groves worked by tenants.77 Although in those cases it was the landowner who met all the expenses necessary to plant the trees and look after them before they started bearing fruit (the aim being to prevent the amount of compensation for improvements from soaring), such groves would frequently be purchased by their farmers. The case of Florentino Márquez illustrates the timeline of the process. Márquez, a notary public in the city of Valencia, bought 124 hectares in Vila-real in the mid-nineteenth century, planted orange trees, and cultivated them by means of tenants. In the early twentieth century his descendants began to sell the groves, and following the First World War they did so at a much faster rate. They owned 91 hectares in 1920, 42 in 1930, 22 in 1940, and none at all in 1950. In order to sell they had always used the same procedure: the 49 hectares that they sold in the 1920s, for example, were purchased by the 79 tenants that worked them, 39 of whom were landless.78

In the districts that specialized in market-gardening, the heavy compensation they had to pay for improvements dissuaded landlords from taking the land in hand to plant orange trees. If a tenant planted trees at his own expense without the landlord's permission the Valencian custom did not give him the right to be paid compensation. Landlords rarely granted such authorization, because it amounted to renouncing once and for all any control over their land. The tenants never planted orange trees without authorization because they were afraid the landlords would evict them and appropriate the trees when the grove started to yield fruit.

As a consequence of these problems, on the eve of the First World War orange trees covered only 20 per cent of the irrigated land in the Huerta of Gandía (at the same time they covered 85 per cent in the district ofVila-real and 65 per cent in

76 In 1912, in Alzira one hectare of irrigated land that was not planted with orange trees was worth an average of 18,000 pesetas, which rose to 30,000 when it contained a full bearing grove. In Gandía one hectare of market-garden on prime quality land sold for 36,000 pesetas, and for 54,000 when it contained full bearing orange trees; Costa, El arbolado, p. 92. Orange trees began to bear fruit five years after being planted, and became full bearing trees at an age of approximately 15 years.

77 Bou, Estudios, p. 205; Calatayud et al., 'El rentismo', pp. 92-3; Modesto, 'Crisis', pp. 261-70.

78 MAV, Padrón de 1859, Amillaramiento de 1888, and Repartos de Guardería Rural (1898-1950). Calatayud, Capitalismo agrario, pp. 103-23, and Garrido, Cánem, pp. 96-104, analyse several similar cases.

Table 5. Owner-operated and tenanted land in Alboraia (Huerta of Valencia)

Number of cultivators % of the total cultivated area that they tilled

Î860 Î889 Î937 Î860 1889 Î937

Owner-occupiers 29 бб 272 1.9% 4.4% 29.G%

Owner-occupiers who 112 144 344 O: б.7% O: 7.0% O: 33.G%

were also tenants T: 22.3% T: 24.3% T: 21.1%

Landless tenants 429 459 242 б9.1% 64.3% 1б.9%

Total 57G бб9 858 1GG.G% 100.0% 1GG.G%

Notes: O: Percentage tilled as owner-occupiers. T: Percentage tilled as tenants.

Sources: MAA, Padrón de I860, Padrón de 1889; Declaración jurada que presenta el campesino de las tierras que trabaja en este término municipal (1937).

Alzira). When the orange underwent a period of exceptional prosperity in the 1920s, however, orange trees did spread in Gandía at a particularly fast rate. In 1926 they covered 51 per cent of the irrigated land, and 80 per cent in 1936.79 Three procedures had been followed to accomplish this: in order to get rid of them, some landowners paid their tenants an amount for compensation that was equivalent to half the value of the estate; others 'gave' the tenants half the land, in order to leave the other half free; and some landlords sold the whole farm to the tenant for half its market value.80 This last option was the one chosen by many landlords who lived in Valencia, Madrid, and other large cities. In the late nineteenth century absentee rentiers who owned more than five hectares of irrigated land accounted for less than 2 per cent of the total number of owners of the Huerta of Gandía, but they controlled almost 25 per cent of its acreage.81 By 1936 most of that surface area had already been bought by its former tenants (and planted with orange trees).

In the Huerta ofValencia the same process displayed slightly different characteristics. The clayey nature of the soil and the high consumption of vegetables in the neighbouring urban centre meant that cultivating oranges did not appear to be quite so attractive as in Gandía. Thus, the tenants were less anxious to buy and in the late 1930s about 40 per cent of the farmland in the district continued to be tilled under fixed-rent contracts (table 5 shows the case of the village of Alboraia). For the same reasons, landlords showed little interest in taking their land in hand and, in the long term, sales from landlords to tenants ended up affecting a much larger area than was the case in Gandía: between the late 1880s and the late 1960s tenants acquired about 65 per cent of the cultivated area.82

In 1986 the application of many aspects of the Valencian custom became compulsory by law following a decision taken by the Valencian Autonomous Parliament, which enacted a law aimed at regulating the surviving 'historical leases'.83The law was passed by a parliament in which the left-wing parties (which

79 Conferencia nacional, p. 120. Fontavella, La Huerta de Gandía, p. 172.

80 Garrido-Juan, El arrendamiento, pp. 46-7.

81 MAG, Amillaramiento de 1887.

82 In the late 1880s about 15% of the agrarian surface area of the Huerta ofValencia was owner-operated, compared to 95.5% in 1972; Instituto Nacional de Estadística, Censo agrario, pp. 44-9.

83 Valencian Historical Leases Act (Ley de Arrendamientos Históricos Valencianos) 1986. See Rúa, Arrendamientos, and Amat, Arrendamientos.

were governing Valencia for the first time since the 1936-9 Spanish Civil War) were in the majority. In fact, the law had an almost purely symbolic value, because in 1986 there were very few leases of that kind left. In 1950 only 11.4 per cent of the cultivated area of the Valencian region was tilled by tenants, in 1962 the figure had dropped to 3.8 per cent, and it had fallen again to 2.4 per cent in 1968.84

No single-factor explanation seems to be enough to demonstrate why landlords opted for a particular type of agrarian contract, and Alston and Higgs have made it clear that any decision a landlord took in this respect forced him to choose between incompatible objectives. The Valencian landowners who utilized fixed-rent contracts intended to fulfil at least six aims simultaneously: (a) to receive the highest possible rent on a regular basis (but taking great care to ensure that the rent was not so high as to cripple the tenant, because in this case there would be negative effects on the tilling of the land); (b) to be able to evict unwanted tenants quickly and economically; (c) to ensure that tenants with proven skills and diligence did not give up the farm; (d) to have some kind of guarantee that tenants with few economic resources would be able to pay the rent, and to open channels of credit so that this type of farmer could invest in the holding; (e) to prevent tenants from impoverishing the land and devaluing it; and (f) to increase the sale value of the estate as a result of the attention dispensed by those tilling it.

In order to achieve (b), landlords offered their tenants short contracts. To accomplish (c), they renewed the contract year after year if they were satisfied with the work carried out by the tenant. To accomplish (d), (e), and (f), on leaving the farm all the tenants received compensation for improvements. These payments for improvements were what apparently made it possible to combine the six objectives. Nevertheless, the final consequences were often of little benefit to landlords.

The contemporary witnesses who denounced a lack of respect for property rights were wrong. What really happened is that there were two different rights: those of the landlord over the land and those of the tenant over the improvements. They were mutually independent and the borderline between the two was clearly established by custom. Nevertheless, as a consequence of the farmers being the owners of the improvements, in the second half of the nineteenth century landlords lost the capacity to change farmer and to keep the rent up to date. This meant that in the twentieth century many landlords ended up selling their property rights over the land to the tenants at a low price.

University Jaume I; University of Valencia

Date submitted 29 December 2007

Revised version submitted 1 September 2009

Accepted 14 October 2009

DOI: 10.1111/j.1468-0289.2009.00521.x

84 Sevilla and Gamiz, 'Estructura', pp. 48-9.

Footnote references

Allen, D. W. and Lueck, D., 'Transaction costs and the design of cropshare contracts', Rand Journal of Economics, 24 (1993), pp. 78-100.

Alston, L. J. and Higgs, R., 'Contractual mix in southern agriculture since the civil war: facts, hypotheses, and tests', Journal of Economic History, 42 (1982), pp. 327-53.

Amat, P., Arrendamientos históricos valencianos (Valencia, 2004).

Antoine, A., 'La légende noire du métayage dans l'ouest de la France (XVIIIe-XXe siècle)', in G. Béaur, M. Arnoux, and A. Varet-Vitu, eds., Exploiter la terre. Les contrats agraires de l'Antiquité à nos jours (Rennes, 2003), pp. 457-69.

Aparisi, J. M., Consideración especial de la locación enValencia (Valencia, 1932).

Axelrod, R., The evolution of cooperation (New York, 1984).

Bardhan, P. K., 'Interlocking factor markets and agrarian development', Oxford Economic Papers, 32 (1980), pp. 82-98.

Bensley, T., 'Property rights and investment incentives: theory and evidence from Ghana', Journal of Political Economy, 103 (1995), pp. 903-37.

Bernabé, J. A., Discurso leído en la solemne inauguración del curso de la Universidad Literaria de Valencia (Valencia, 1907).

Blasco-Ibáñez, V., La barraca (Barcelona, 1976 [1st edn. 1898]).

Bou, F., Estudios sobre el naranjo (Castellón, 1879).

Burriel, E., La Huerta deValencia (Valencia, 1971).

Burriel, F. and Oller, M., 'El problema de la tierra en Valencia', in Crónica del I Congreso Diocesano de Acción Católica (Valencia, 1929), pp. 109-24.

Calatayud, S., Capitalismo agrario y propiedad campesina. La Ribera del Xúquer, 1860-1930 (Valencia, 1989).

Calatayud, S., Millán, J., and Romeo, M. C., 'El rentismo nobiliario en la agricultura valenciana del siglo XIX', Revista De Historia Económica, XVIII (2000), pp. 79-107.

Calatayud, S., Millán, J., and Romeo, M. C., 'Leaseholders in capitalist Arcadia: bourgeois hegemony and peasant opportunities in the Valencian countryside during the nineteenth century', Rural History, 17 (2006), pp. 149-66.

Carmona, J. and Simpson, J., 'The "rabassa morta" in Catalan viticulture: the rise and decline of a long-term sharecropping contract, 1670s-1920s', Journal of Economic History, 59 (1999), pp. 290-315.

Carmona, J. and Simpson, J., El laberinto de la agricultura española (Zaragoza, 2003).

Castellanos, F. and Fernández, J., 'Amos i masovers: les relacions d'arrendament aValència (1870-1885)', Afers, 6, 11/12 (1991), pp. 215-25.

Chambers, J. D. and Mingay, G. E., The agricultural revolution, 1750-1880 (1966).

Cohen, J. S. and Galassi, F., 'Sharecropping and productivity: "feudal residues" in Italian agriculture, 1911',

Economic History Review, 2nd ser., XLIII (1990), pp. 646-56.

Conde de Montornés, 'Ensayo de explotación agrícola y social', Los Mercados, 547 (1911), pp. 1-7.

Conde de Montornés, 'Ensayo de explotación agrícola-social', in ••. ••, ed., IX Congreso Internacional de Agricultura (Madrid, 1912), pp. 135-9.

Conferencia nacional naranjera (Madrid, 1926).

Congost, R., 'Sagrada propiedad imperfecta. Otra visión de la revolución liberal española', Historia Agraria, 20 (2000), pp. 61-93.

Costa, J., El arbolado y la patria (Madrid, 1912).

Costa, J., 'Aparcerías y arrendamientos en la provincia de Alicante', in G. Cano, ed., Estructuras y regímenes de tenencia de la tierra en España (Madrid, 1987), pp. 211-36.

Crisis agrícola y pecuaria. Información escrita. Tomo tercero (Madrid, 1887).

Cucó, A., Republicans i camperols revoltats (Valencia, 1975).

Dirección General de los Registros, Memoria y Estados formados por los Registradores de la Propiedad, 4 vols. (Madrid, 1889).

Felix, E., L'estructura de la propietat agrària a Alboraia (1828-1889) (Alboraia, 1986).

Font de Mora, R., El naranjo (Madrid, 1935).

Fontavella, V., La Huerta de Gandía (Zaragoza, 1952).

Galassi, F. and Cohen, J. S., 'The economics of tenancy in early twentieth-century southern Italy', Economic History Review, XLVII (1994), pp. 585-600.

Garrabou, R., Un fals dilema. Modernitat o endarreriment de l'agricultura valenciana, 1850-1900 (Valencia, 1986).

Garrabou, R., Planas, J., and Saguer, E., 'Sharecropping and the management of large rural estates in Catalonia, 1850-1950', Journal of Peasant Studies, 28, 3 (2001), pp. 89-108.

Garrido, S., Cànem gentil. L'evolució de les estructures agràries a la Plana de Castelló (Castellón, 2004).

Garrido, S., 'Why did most cooperatives fail? Spanish agricultural cooperation in the early twentieth century', Rural History, 18, 2 (2007), pp. 1-18.

Garrido, S. and Calatayud, S., 'La compra silenciosa. Arrendamientos, estabilidad y mejoras en la agricultura valenciana de regadío', Investigaciones De Historia Económica, 8 (2007), pp. 77-108.

Garrido-Juan, R., El arrendamiento consuetudinario valenciano (Valencia, 1943).

Grupo de Estudios de Historia Rural, Estadísticas históricas de la producción agraria española, 1859-1935 (Madrid, 1991).

Guinnane, T. and Miller, R., 'Bonds without bondsmen: tenant-right in nineteenth-century Ireland', Journal of Economic History, 56 (1996), pp. 113-42.

Guinnane, T. and Miller, R., 'The limits to land reform: the land acts in Ireland, 1870-1909', Economic Development and Cultural Change, 45 (1997), pp. 591-612.

Hayami, Y. and Otsuka, K., The economics of contract choice. An agrarian perspective (Oxford, 1993).

Instituto Nacional de Estadística, Censo agrario de España (Madrid, 1972).

International Institute of Agriculture, International yearbook of agricultural statistics, 1928-29 (Rome, 1930).

Jonsson, U. L. F., 'The paradox of share tenancy under capitalism: a comparative perspective on late nineteenth-and twentieth-century French and Italian sharecropping', Rural History, 3, 2 (1992), pp. 191-217.

Lassala, V., Memoria sobre la producción y el comercio de la naranja en España (Valencia, 1873).

Liebowitz, J. J., 'Tenants, sharecroppers, and the French agricultural depression of the late nineteenth century', Journal of Interdisciplinary History, XIX (1989), pp. 429-45.

Martínez-Rodríguez, S. and Martínez-Soto, A., 'Los pioneros del cooperativismo agrario de crédito español (1880-1920)', Ciriec-España, 63 (2008), pp. 89-112.

Maylín, A., Memoria sobre el arbolado cultivado y espontáneo en la provincia de Castellón (Madrid, 1904).

Millán, J., El poder de la tierra. La sociedad agraria del Bajo Segura en la época del liberalismo (Alicante, 1999).

Millán, J., 'Triunfo y límites de la propiedad en el arrendamiento valenciano', in S. De Díos, J. Infante, R. Robledo, and E. Torijano, eds., Historia de la propiedad (Madrid, 2004), pp. 374-410.

Modesto, J. R., Aúsi costum de bon llaurador. L'arrendament de terres a l'Horta deValencia (1780-1860) (Catarroja, 1998).

Modesto, J. R., 'Crisis y transformaciones en la gran propiedad rentista valenciana' (unpub. Ph.D. thesis, Univ. of Valencia, 2004).

Mokyr, J., Why Ireland starved (1985).

Monforte, A., El problema agrario levantino (Valencia, 1922).

Morán, F., 'El colonato en España', in Semana Social de España. Segundo Curso. Valencia (Zaragoza, 1908), pp. 25-91.

Orellana, P., El colonato enValencia (Valencia, 1886).

Palacios, F., Servicio Agronómico Nacional. Sección de Castellón (Castellón, 1913).

Palafox, J., 'Estructura de la exportación y distribución de beneficios. La naranja en el País Valenciano (1920-30)', Revista De Historia Económica, 1 (1983), pp. 339-51.

Palafox, J., 'Expansión de las exportaciones, crecimiento, diversificación industrial e industrialización', in J. Azagra, E. Mateu, and J. Vidal, eds., De la sociedad tradicional a la economía moderna (Alicante, 1996), pp. 321-45.

Pigou, A. C., The economics ofwelfare (1932).

Piqueras, La vid y el vino en el País Valenciano (Valencia, 1981).

Pons, A., 'Els contractes d'arrendament al País Valencia', Estudis d'Historia Agraria, 8 (1990), pp. 175-99.

Pons, A. and Serna, J., La ciudad extensa (Valencia, 1992).

Prothero, R. E., 'Tenant right and agrarian outrage in France', Contemporary Review, 50 (1886), pp. 832-50.

Ray, D., Development economics (Princeton, 1998).

Reformas Sociales. Tomo III. Información oral y escrita practicada en virtud de la Real Orden de 5 de diciembre de 1883. Valencia (Madrid, 1891).

Robledo, R., La renta de la tierra en Castilla laVieja y León (1836-1913) (Madrid, 1984).

Romero, J., Propiedad agraria y sociedad rural en la España Mediterránea (Madrid, 1983).

Rowntree, B. S., Land and labour: lessons from Belgium (1910).

Rúa, J. L., Arrendamientos históricos valencianos. Doctrina jurisprudencial de la Sala de lo Civil y Penal del Tribunal Superior de Justicia de la Comunidad Valenciana (Valencia, 2002).

Sabio, A., Viñedo y vino en el campo de Cariñena (Zaragoza, 1995).

Sanmartín, A., 'El patrimoni castellonenc del baró de la Pobla', Recerques, 41 (2001), pp. 59-75.

Sarthou, C., Geografía general del Reino deValencia. Provincia de Castellón (Barcelona, 1914).

Sebastia, E. and Piqueras, J. A., Pervivencias feudales y revolución democrática (Valencia, 1987).

Sevilla, E. and Gámiz, A., 'Estructura espacial de las formas de tenencia de la tierra en España', Revista De Estudios Agrosociales, 74 (1971), pp. 7-75.

Shaban, R. A., 'Testing between competing models of sharecropping', Journal of Political Economy, 95 (1987), pp. 893-920.

Simpson, J., 'La producción y la productividad agraria españolas, 1890-1936', Revista De Historia Económica, XII (1994), pp. 43-84.

Singh, N., 'Theories of sharecropping', in P. K. Bardhan, ed., The economic theory of agrarian institutions (Oxford, 1989), pp. 33-72.

Solow, B., The land question and the Irish economy, 1870-1903 (Cambridge, Mass., 1971).

Tirado, C., 'Gran propiedad y arrendatarios en l'Horta deValencia', Estudis d'Historia Contemporania Del País Valencia, 8 (1990), pp. 95-131.

Tirado, C., 'Salvador Castillo i Madroño: estudi d'un cas d'acumulació territorial a la segona meitat del segle

XIX', Afers, 6, 11/12 (1991), pp. 227-38. Whyte, I., Agriculture and society in seventeenth-century Scotland (Edinburgh, 1979).

Official publications

Reports from Her Majesty's representatives respecting the tenure of land in the several countries of Europe, III (P.P. 1871, C.271).

Royal Commission on Depressed Condition of Agricultural Interest. Reports of the Assistant Commissioners (P.P. 1882, C.3375-V).

Foreign Office, Annual Ser., no. 275, Diplomatic and Consular Reports on Trade and Finance. Spain. Report for the

year 1887 on the agriculture of the Barcelona District (P.P. 1888, C.5252-52) Royal Commission on Agricultural Depression (P.P. 1897, C.8540).