Contents>> Vol. 4, No. 1
Intergenerational Land Transfer in Rural Cambodia since the Late 1980s: Special Attention to the Effect of Labor Migration
* 矢倉研二郎, Faculty of Economics, Hannan University, 5-4-33 Amamihigashi Matsubara, Osaka 580-8502, Japan
Using primary data collected from three villages in Prey Veng and Pousat provinces, this study describes how land has been transferred from parents to children in rural Cambodia since the late 1980s. While equal division among all children is the most favored practice—thus further farm fragmentation is anticipated in the near future—parents with very small land endowment are unable to divide land equally among all their children and some children are unable to receive land from their parents. The expansion of migration opportunity has not caused fundamental changes in land transfer practices, but the premarital migration experience of children is negatively associated with land transfer, especially when children settle in the migration destination or marry a person from another province whom they met at the migration destination and move to their partner’s place of origin. The data indicate that, taking advantage of labor migration experience, children of land-poor parents choose to leave their home province and make a living without land. However, landless children are in a disadvantaged economic situation because they are also less likely to receive non-land assets from their parents and farmland from their spouse’s parents.
Keywords: land transfer, migration, family, marriage, Cambodia
Farmland under family farming is generally transferred from parents to their children in societies where the private ownership of land is established. While the way in which land is transferred intergenerationally varies according to societies and times, inheritance is basically classified into partible and impartible. In the former, parents’ land is divided among their children, while in the latter, land is given to only one heir. Scholars have argued that land inheritance rules exert an important influence on rural society and economy (Platteau and Baland 2001). How land is transferred between generations naturally affects agricultural production since it determines farm size (Ram et al. 1999) and brings about changes in land distribution and rural class structure (Khera 1973; Wiegandt 1977). It has also been found to shape family structure and social interaction among kin (Goldschmidt and Kunkel 1971; Khera 1972) and determine the rate and the timing of marriage (Emigh 1997) and birth rate (Habakkuk 1955). Furthermore, some scholars argue that land inheritance rules can influence the development of the manufacturing industry since it affects labor mobility from rural areas (ibid.; Wegge 1999).
Although the empirical data on intergenerational land transfer1) in Cambodia is scarce, partible transfer seems to have been common practice. For example, Ebihara (1971) found that in a village in Kandal province in the late 1950s, farmland was divided among children in a basically equal manner (ibid.). Land became collective property under the Democratic Kampuchea (Khmer Rouge) regime and subsequently the People’s Republic of Kampuchea in the early 1980s, but was distributed among households after the dissolution of collective farming groups (krŏm samokki) in the 1980s. Since then, parents distribute land to their children. In fact, some studies show that children who got married after the land distribution of the 1980s received land from their parents (Amakawa 2001; Yagura 2005; Kobayashi 2007). However, these studies do not examine whether partible (and equitable) land transfer is still a common practice in rural Cambodia because they do not use data from the parents’ side.
In this regard, we must note that socio-economic changes experienced in rural Cambodia since the 1980s have had a considerable influence on land transfer practice. First, population is increasing rapidly while unexploited land is diminishing—in other words, land is becoming a scarce resource. In fact, while there remains untapped land in some regions, such as the northwestern and the northeastern provinces, there seems to be little land left for agricultural use in regions with a high population density such as the provinces surrounding Phnom Penh.2) On the other hand, the baby-boomers of the 1980s have started to get married in the first decade of the twenty-first century. According to custom, parents should give land to all these married children, yet this may not be possible where land is becoming scarce. Second, labor migration from rural areas is increasing, which can also be regarded as a response to land scarcity. Since the middle of the 1990s, the migration of rural youth to urban areas, especially Phnom Penh, has increased with the development of sectors such as the garment industry, construction, and commerce in urban areas. Increased opportunities to earn a living outside rural areas make farmland less necessary.
Both these changes ought to promote a transition from partible to impartible transfer of land. The change to impartible land transfer then raises the question of who will be deprived of access to land and how these landless children should make a living.
On the other hand, increasing employment opportunities in non-agricultural sectors may actually make it possible to sustain partible land transfer because the non-agricultural income can compensate for a small agricultural income (Platteau and Baland 2001). This means that the direction of the effects of labor migration on land transfer practice cannot be assumed a priori.
The objective of this study is to describe how land has been transferred from parents to children in rural Cambodia since the late 1980s. Special attention is placed on whether the practice of partible and equitable land transfer is giving way to unequal and impartible land transfer, and whether and how the increasing labor migration from rural areas affect the changes in land transfer practice. Because there is no nation-wide survey on intergenerational land transfer in Cambodia, this study uses data collected by the author in three villages of Prey Veng and Pousat provinces in 2009.
Examining changes in land transfer practice and the effect of rural out-migration is important because it provides us with clues to understand whether and how land distribution in Cambodian villages has changed in recent years as well as predict changes in land distribution in the near future.3) Land distribution itself is a fundamental factor determining income distribution, prevalence of poverty, and production organization, as well as the efficiency of agriculture. But studying land distribution is beyond the scope of this study and will be left for further research.4) This paper is limited to the presentation of implications of changes in land distribution or stratum structure within a village, from the findings of an empirical analysis of intergenerational land transfer.
The organization of this paper is as follows. In the next section, after presenting factors determining intergenerational land transfer practice in general, I discuss expected changes in Cambodia since the 1980s. Section III describes sample villages and Section IV reveals the normative aspects of the equal division of land by presenting parents’ plan of land transfer. Section V describes the actual situation of intergenerational land transfer in the villages surveyed, with a focus on the extent to which equitable and partible land transfer is maintained and its relationship with parents’ land endowment. In addition, I estimate the degree to which children are unlikely to receive land from their parents based on the current land endowment of their parents. Section VI investigates the effect of the premarital migration experience of children on land transfer from their parents after their marriage. Section VII sheds light on the economic situation of married children who have not received land from their parents, by examining land transfer from their spouse’s parents as well as transfer of non-land assets from their own parents. In the last section, after summarizing the findings of the previous sections, I discuss possible effects of the recent trends in intergenerational land transfer on the stratum structure of rural society in Cambodia.
II Factors Determining the Mode of Intergenerational Land Transfer
As discussed above, there are two basic modes of intergenerational land transfer: partible and impartible. As long as the land is an indispensable asset and a major means of making a living in an agricultural society, partible transfer seems to be the natural choice and impartible transfer is adopted only in exceptional cases.
Based on previous studies, Platteau and Baland (2001) argued that impartible transfer is likely to be selected for the following reasons or under the following circumstances: (1) to save administrative costs of tax collection (in cases where tax is levied on land);5) (2) agricultural production has economies of scale; (3) to maintain patriarchal relations by not dividing symbolic assets; (4) there are rules to restrict membership at community level (for restrictive use of commons); (5) farm size becomes too small to be economically viable if land is divided among children (ibid.).
These circumstances are practically absent in Cambodian (Khmer) rural society and therefore partible transfer is more likely to be selected. First, tax has not been levied on the farmland of peasants. Second, traditional rice farming without mechanization is unlikely to benefit from economies of scale. Third, the Khmer family, as indicated by previous studies, is not characterized by patriarchal relationships (Ebihara 1971; Ledgerwood 1995; Takahashi 2001). Fourth, rules to restrict community membership are not known. Finally, population density is relatively low and unexploited land existed until recently.
Among these points, the first four are still relevant. Only the last point is less valid in rural Cambodia today, as the population continues to rise while the frontier has been diminishing since the 1980s. When farm size becomes too small, parents may stop giving land to all their children and cases of impartible transfer may increase.
On the other hand, labor migration from rural to urban areas or between rural areas has increased since the late 1990s in Cambodia. The increase in labor out-migration in recent years can facilitate the transition from partible transfer or equal division of land to impartible land transfer in two ways. First, as migration opportunities provide rural youth with the means to make a living without land, parents may not give land to migrating children and may decide to give only to children who stay in their native village. Second, the increase in youth labor migration can also affect the selection of a marital partner and the place of residence after marriage. At migration destinations, migrants meet other migrants of the opposite sex from diverse locations, who might emerge as prospective spouses. Therefore, an increasing number of youths might marry a person from a remote place (such as another province) and settle in the place of origin of their spouse, far from their home village. Parents might then not give land to these children because they cannot cultivate land in their home village.
As for the selection of marital partner, as Yagura (2012) indicates, migrating children who have little expectation of receiving land from their parents may consciously search for their partner at the migration destination because they have little incentive to return to their home village.
On the other hand, increasing opportunities of migration may preserve partible land transfer. As argued by Platteau and Baland (2001), abundant off-farm employment opportunities may actually sustain partible inheritance rule because small-scale farmers are able to make a living by combining off-farm income. This argument also holds true of migration opportunities. Furthermore, recent improvements in land productivity in Cambodian rice farming may contribute to sustaining partible land transfer by enhancing the viability of small-scale farmers.6)
Though changes in the economic environment can cause a shift in intergenerational land transfer practice, empirical studies also show that longstanding practices, especially the equal division of land, tend to persist even in the face of major socio-economic changes (Salamon 1980; Takeuchi 2000; Platteau and Baland 2001). Where the equal division of land is a longstanding custom, children may regard receiving land from their parents as their right; therefore changing this custom could entail much friction among family members.7) This fact cannot be ignored in analyzing land transfer practice in rural Cambodia, where the equal division of land has been common practice.
III Villages Surveyed
III-1 Outline of the Survey
Data used in this paper was collected through a field survey in August 2009. Using a questionnaire, members of the research group, consisting of Cambodian postgraduate and undergraduate students, college graduates as well as the author, conducted interviews in Khmer with heads of sample households or their spouses.
The survey was conducted in three villages, which we shall call S, K, and Y. Village S is located in Ba Phnom district, Prey Veng province. K is also part of Prey Veng province, but is located in Peam Ro district. Village Y is located in Bakan district, Pousat province (see Maps 1, 2, and 3).
Before settling on the villages, I first chose as the survey area two provinces that have experienced a relatively large scale of population outflow in recent years.8) This condition is very important for studying the effect of out-migration on intergenerational land transfer. In addition, the two provinces have a different population density, which enables us to examine the effect of farm size. Farm size was expected to be larger in Pousat than in Prey Veng, where the population density is higher.
After choosing these two provinces, I conducted a preliminary survey in various parts of each province to select villages for the main survey. The three villages were eventually selected for the following reasons. First, a certain proportion of villagers in all three villages have experienced labor migration. Second, the three villages have different agricultural conditions. While all three have only rice fields (no upland fields), farmers in S and Y can grow rice only in the wet season while farmers in K can grow rice in both wet and dry seasons, thanks to a lake near the village that provides water even in dry seasons. The rice fields for the wet and dry seasons are, however, different (double cropping is not possible).
The difference between S and Y lies in farm size: on average, farmers in Y have much larger lands. We can examine the effect of land scarcity by comparing S and Y, and the effect of dry season farming (that is, the availability of employment in farming throughout the year) in K.
The number of households surveyed was 187 in S, 208 in K, and 179 in Y. We interviewed all the households in S and Y, except those whose heads were absent during the period of our survey. In K, we only visited the households in the southwestern part of the village so that the number of households surveyed would be about the same in all three villages. As such, statistics derived from whole sample have little bias without adjustment (that is, weighting according to the number of sample households in each village).
In the survey we collected information on land transfer to household heads and their spouse from the two sets of parents, as well as land transfer from the same household heads and their spouse, as parents, to their children. Information on the transfer to their children is used in this paper because it includes data on whether and how land is divided among children. In the following sections, “parents” refers to the head (and his/her spouse) of sample households, and “children” refers to children of the household head, unless otherwise noted.
According to villagers in the three villages, S and K existed even before the French colonial period. As for Y, its establishment dates back to at least the 1920s because a villager who was in his nineties was born in Y.
The continuity of village society before and after the era of Democratic Kampuchea (DK) seems strong in S and K. Under the DK era, some of villagers in S were assigned to reclaim land in Peam Chor district in the province. They were not relocated but commuted there from S. People in K were moved to the western part of the same commune to reclaim land just before the collapse of the regime, but they returned en masse once the regime fell. Villagers of Y were all relocated to Veal Veng district (the western part of Pousat province), but after the regime collapsed, most of them returned to Y, though some fled to the Thai border area. On the other hand, some people who had been relocated from Takeo and Kampot to the western part of the country settled in Y after the regime collapsed and did not return to their original provinces. Furthermore, some villagers had moved to Pousat from Takeo and Kampot on their own initiative even before 1975.
After the collapse of the DK regime, krŏm samokki were formed in all three villages.9) Farming was conducted under krŏm samokki for only one season (1979) in Y and for three seasons from 1979–81 in S and K. After the last farming season under krŏm samokki, farmland was divided among households.
Methods of land distribution differed according to villages. In S, each household received an area of wet season rice field based on the number of household members and their work capacity. Villagers were classified into three categories based on their work capacity: kâmlăng muoy (those who can plow rice fields), kâmlăng pi (those who can uproot and transplant rice seedlings), and kâmlăng bey (those who can only tend cattle; basically children). Allocation of rice field per person for each of these categories is 0.25, 0.20, and 0.15 hectares, respectively.
In K, each household received 0.13 hectares of wet season rice field and 0.10 hectares of dry season rice field per person, without taking into account members’ work capacity. As mentioned earlier, in K, wet season rice and dry season rice are grown in different fields. Nevertheless, some villagers told the author during the interview that villagers also reclaimed dry season rice fields by themselves.
In Y, rice fields cultivated by each krŏm were divided among the member households of that krŏm, and each household received a uniform area of land within a krŏm, irrespective of the number of household members. However, the area of rice field per household varied from one krŏm to another because each krŏm also reclaimed rice fields independently in 1979. This also indicates that a vast area of unexploited land in Y still remained in the early 1980s.
As for residential land (dey phoum), people in S and K basically occupied plots they had used before the DK era, though some households in S received residential land from the local authority. In Y, all households received a uniform size of residential land (30 meters by 80 meters) from the local authority.
Comparison of the Population Census in 1998 with that in 2008 reveals demographic changes in the three villages. During this decade, both S and K experienced a large increase in the number of households (from 170 to 207 in S and from 295 to 369 in K), but the population of S increased only slightly (from 751 to 778) and the population of K even decreased (from 1,475 to 1,418). In Y, while almost no change was found in the number of households (from 182 to 186), the population decreased by as much as 10 percent (from 1,039 to 933). Population in these statistics do not include family members who were absent at the time of the census due to, for example, labor migration. Therefore, the population decrease during the decade indicates an increase in out-migration from these villages. In addition, it is highly likely that the fertility rate in these villages decreased in this decade, as is observed in national- and provincial-level data (Cambodia, National Institute of Statistics 2010a). As a result, average household size became smaller in all three villages. Even so, the number of households increased in S and K, probably because baby-boomers of the early 1980s married and established their own household during this period.
Table 1 shows the distribution of farmland at the time of the survey.10) In S and Y, almost all the farmland is wet season rice field, while in K, around half of the farmland is dry season rice field. Smallness of farm size in S is evident from the table. On the other hand, the larger the average farm size, the higher the landholding inequality among households.
Besides receiving land from the government as mentioned above, some acquired land through reclamation (especially in Y), and younger generations received land from their parents. Parents mostly give land to their child at the time of the child’s marriage or within a few years after the marriage. Land acquisition through purchase is also prevalent. Among households in which parents gave land to their children after 1979, 32 percent had purchased farmland before and 15 percent had sold the land since then. Land rental markets also exist. Among 574 sample households, 26 households rent out farmland and 56 households rent farmland in the last farming season.
Based on the estimate of village chiefs, yield of wet season rice (unhusked) is 1.5–2.0 ton/ha in S, 2.0 ton/ha in K, and 3.5 ton/ha in Y. Yield of dry season rice in K is 2.0–2.5 ton/ha. The village chief of S conjectured that rice yield in his village had been increasing in the last decade, while the chiefs of the other two villages did not think that the rice yield in their villages had increased.
Although farming in these villages still relies very much on manual labor and draft animals, farm machines such as power tillers (kŏ yon), threshers, and harvesters have been gradually introduced in recent years in the villages surveyed, especially in K and Y.
III-5 Labor Migration
Table 2 shows the incidence of labor migration in the preceding 12 months by family members of the sample households. Labor migration refers to working outside the village, excluding cases in which workers commute from their own house in the village. In the table (and hereinafter), “husband” and “wife” refer to the household head and his/her spouse. In cases where household heads are currently unmarried (widows, widowers, and the divorced), they are also referred to as “husband” or “wife” according to their sex.
Migration data of husbands and wives in Table 2 indicates that: husbands are more likely to migrate than wives; migration is more prevalent among younger generations than among older generations (except in village S); and prevalence of labor migration is higher where land endowment is smaller. Table 2 also shows that labor migration is more widespread among unmarried children than among their parents. Interestingly enough, inter-village differences in migration prevalence are much smaller among unmarried children. This suggests that labor migration has become widespread among rural youth irrespective of the land endowment of their villages.
Phnom Penh is the top destination of labor migration, accounting for around 80 percent of migrants’ destinations. Migrants from Y, however, are more likely to go to Thailand, probably because of its geographical closeness. A large portion of male migrants works as construction workers or motorbike taxi drivers, and female migrants work at garment factories.
Table 3 allows us to see whether labor migration has been increasing in recent years. It shows the premarital migration rate, or proportion of husbands and wives who experienced labor migration before marriage, by age group. In this table, only the data of husbands and wives who come from the villages surveyed are used so as to detect changes in the prevalence of labor migration from the three villages. Though this data is biased in the sense that it does not include those who have not returned to the village after emigrating from the village, it does inform on the basic trend of labor migration in the last decades.
As is evident from the table, the premarital migration rate is much higher among husbands and wives in their twenties and thirties. In particular, around half of the husbands in their twenties in all three villages experienced labor migration when they were single. While few of the wives in their forties or older have migrated before marriage, 20 to 40 percent of wives in their twenties have experienced labor migration. In S, the premarital migration rate is relatively high even among middle-aged husbands. In contrast, in Y, no husbands in their thirties or older experienced labor migration before marriage, but nearly half of the husbands in their twenties migrated when they were single.
The data indicate that labor migration used to be characteristic among males of villages with small land endowment like S, but it has become widespread in recent years among young people, irrespective of their gender and land endowment.
IV Parents’ Land Transfer Plan
IV-1 Inclination toward Partible and Equal Land Transfer
Before examining the actual situation of intergenerational land transfer in the three villages, this section shows parents’ plans for distributing land to their children.
In the interview, we asked household heads (or their spouse) who owned farmland at the time of the interview but had yet to give farmland to all their children, whether they planned to distribute land to all their children. Among 489 respondents who had more than one child, 84.5 percent answered “yes,” 7.4 percent “no,” and the rest (8.2 percent) responded “do not know.” There is no significant difference in the distribution of responses between the villages. Then, we asked those who answered “yes” to the question above whether they planned to give farmland to their all children in equal portions. Among 408 respondents, 93.4 percent answered “yes,” only 5.6 percent responded negatively, and 1 percent responded “do not know.” No significant difference was found between villages in this category as well. These responses indicate that most of the couples plan to divide their land equally among all their children.
We proceeded to ask those who indicated the equal division of land among children their reasons for doing so. The top answer was: “we should treat our children equally,” cited by 70 percent of the respondents. “In order not to cause dispute or envy among children” was the next most popular reason, cited by 13.8 percent of the respondents. These answers indicate that fear of dispute underlies parents’ inclination toward equal division, although I did not hear of actual cases of dispute related to land transfer. In addition, both parents and children seemed to regard receiving land equally from parents as children’s rights.
On the other hand, there are some differences according to generation. The older generations tended to answer “no” to the question of partible land transfer and equal land division. Among household heads in their sixties, 21 percent responded negatively to the question of giving land to all their children. To the question of equal division among children, 12 percent of those in their fifties and 10 percent of those in their sixties answered “no.” Such a tendency seems to reflect the difficult situation of some older parents who have only a small piece of land left and for whom buying land to give to their children (a practice adopted by affluent parents) is also difficult. On the other hand, as high a rate as 16 percent of household heads in their twenties answered “do not know” to the question of partible land transfer. This would reflect the fact that they need to decide on this question only some 20 years later and thus are still very uncertain about what to do.
IV-2 Relationship with Land Endowment
Next we examine the relationship between land transfer plan and farm size. How much land parents can give to their children depends very much on how much land parents own at present and the number of children yet to receive land. As an indicator of the availability of land to children, we define “transferrable land endowment” (TLE, hereinafter) as the area of farmland owned by parents at the time of the survey, divided by the number of their children aged below 40 (including both married and unmarried children) yet to receive land from them. The relationship between TLE and the land transfer plan is presented in Table 4. The denominator of TLE does not include children aged 40 or older because the older children are mostly married and have established their own households by the early 1980s, and hence would have received land from the local authority. Parents generally did not give land to these children.
As is evident from Table 4, parents with small land endowment (especially those with TLE<0.2 hectares) are more likely to plan to give land only to some of their children. Nevertheless, it is also noteworthy that around 75 percent of those with TLE<0.2 answered that they would give land to all their children. On the other hand, inclination toward equal division is just as strong among those who planned to give land to all their children regardless of their land endowment. This again indicates the normative aspect of the equal division of land.
The data presented above reveal that most parents want to divide land equally among their children because they feel that their children should be treated equally. This inclination is widely shared, even among those who have very limited land endowment, which suggests that further farm fragmentation is anticipated in the three villages in the near future.
V Land Transfer in Practice
V-1 Basic Situation of Land Transfer
Parents’ plans of land transfer presented in the previous section may merely reflect their wishes; whether they can do as they desire would depend on their socio-economic environment as well as their children’s life course. In this section, actual transfer of land from parents (household heads and their spouse) to their married children aged below 40 is examined. We only take into account land transfer to married children because giving land to unmarried children is very rare in the three villages (and probably in Cambodian villages in general). Children aged 40 or older are excluded for the same reason mentioned in the previous section. Our analysis focuses on the transfer of farmland because almost all children who received residential land from their parents also received farmland.11) In other words, residential land is not given to children as a substitute for farmland. For this reason, the term “land” refers to farmland (rice field) hereinafter unless otherwise mentioned.
Note that the land transfer situation presented here is not the final one in the sense that parents may give land in future to children who had yet to receive land at the time of the survey, especially children who had got married very recently.
Table 5 shows the proportion of married children aged below 40 who received land from their parents. As we can see, 79 percent of the children received land from their parents. The proportion of children receiving land is unrelated to land endowment of each village, as the proportion in K is lower than in S. Nevertheless, the average area of land given to children (the average derived from those who received land) is clearly proportional to the average farm size of the village; thus the differences between villages are very large.
Table 5 also shows that land is principally divided among children regardless of gender and birth order, which is consistent with the argument by Ebihara (1971) that there is little difference in land transfer according to gender. Though the eldest child tends to receive more land than average, the probability of receiving land is not especially high. The probability of receiving land is a little lower for the youngest child, but this is probably caused by the fact that the youngest children are more likely to live with their parents or their spouse’s parents.12) The proportion of those who received land is lower for those who married in recent years. This may reflect the fact that in some cases land is transferred from parents to children a few years after their marriage.
To check whether there is any change in land transfer practice, Fig. 1 shows the proportion of those who received land from their parents and the average size of farmland received according to the age group of children. We can observe that in all the three villages younger children are less likely to receive land.13) However, we may not conclude that the practice of partible transfer has been weakened because, as mentioned above, some younger children might not have received land yet because they had only just got married. Though the average size of land received tends to be smaller for the younger generation in Y, the average size of land received for each age group is falls within a very narrow range: around 0.2 hectares in S, 0.4 hectares in K, and 0.8 hectares in Y.
V-2 Is Land Divided Equally among All Children?
In this subsection, in order to grasp the prevalence of partible and equitable land transfer, we check whether parents who have more than one married child aged below 40 gave land to all these married children. The data is presented in Table 6. As shown in the fifth column of the table, nearly two-thirds of parents gave land to all their married children aged under 40. The difference between villages is large, as only 53.2 percent of parents in K did so. The difference according to age is also significant. The proportion of those who have not given land to any of their married children under 40 is higher among the younger parents. Younger parents are unable to give land to their children probably because they themselves tend to have limited land; or because their children have only very recently got married (and thus will receive land from their parents in the near future).
The figures in the sixth and seventh columns of Table 6 reveal the prevalence of equal division. The sixth column shows that among those parents who gave land to all their married children under 40 (corresponding to 102 households), 73.5 percent distributed the land equally. The prevalence of equal division over the whole sample amounts to 47 percent (seventh column). We can also observe that equal division is less common for younger parents.
Overall, the data presented in Table 6 indicate that equal division of land is not implemented as much as parents might have wished, though the observed difference in the size of land between siblings may reflect an adjustment to the quality of land in some cases.
As in the case of parents’ plan of land transfer, their land endowment is assumed to affect the actual land transfer.14) But it is not parents’ current land endowment but the size of land owned before giving land to their children that determines whether and how they distribute this land. In addition, the number of children also matters. Therefore, we have constructed an indicator “original land endowment” (OLE), defined as the size of farmland owned by parents at the moment they give farmland to their child for the first time after having received land from the local authority in the 1980s, divided by the number of children aged below 40 (including both married and unmarried children). For parents who have not given land to any of their children, the size of farmland they currently own is used as the numerator. We examine the relationship between parents’ land endowment and the actual land transfer using this OLE.
As shown in Table 7, among married children under 40, the proportion of those who received land from their parents tends to increase with parents’ land endowment—except in village K. Table 7 also shows a more obvious positive relationship between OLE and the average size of land received by children, though the link is very weak in S.
The effect of parents’ land endowment on their land transfer practice is presented in the lower part of Table 6 (only for parents who have more than one married child aged below 40), which shows that a larger OLE is associated with higher probability of giving land to all their married children (the fifth column). Among those who gave land to all their married children, around 80 percent gave in equal size if OLE is less than 0.8 hectares. Interestingly, parents with OLE of 0.8 hectares or more are less likely to practice equal division. A possible reason for this is that even if they give a disproportionately larger piece of land to some of their children, they can still afford to give large areas of land to their other offspring because of their rich land endowment situation.
Though Table 6 shows that parents with small land endowment are less likely to give land to all their married children, it is noteworthy that half of the parents with OLE as small as less than 0.2 hectares, still gave land to all their children, and that 80 percent of them divided this land equally. This again suggests the durability of the practice of equal division. Let us now turn to some examples of parents who gave farmland to all their married children even though they owned very little land.
Household No. 150 in Y is a headed by a woman. The 60-year-old household head, whose husband is deceased, has five married children. She gave 0.18 hectares of rice field to each child and, as a result, was landless at the time of the survey. The eldest child (son, 29 years old), the second child (daughter, 26 years old), and the fourth child (son, 23 years old) each married a local introduced by their parents. Their partners also received land from their own parents. These children now live in the same commune or the same district and earn a living by farming without migration. The third child (daughter, 24 years old) married a man from Battambang province and also received farmland from her husband’s parents. This daughter lives with her mother but had also worked in Thailand with her husband as an agricultural wage laborer in the last 12 months. The fifth child (daughter, 21 years old) married a man she met from another district in the same province. The couple also worked in Thailand as agricultural wage laborers, using the house of the husband’s parents as their basic place of residence. The third and the fifth children migrated in the last 12 months as well, but stayed abroad for only one month, probably because of their seasonal agricultural wage laborer job at the destination. Therefore they also need farmland to make a living, and that is probably why their mother gave them land, even though the size of the land distributed is very small. Or, we could infer that equal division is possible thanks to migration opportunities, such that children can make a living even though land given from their parents is very small.
The second example is household No. 43 in S. It is a family of five composed of a couple and their three children. The eldest child (son, 22 years old) had worked in Phnom Penh as a construction worker for 1 month in the last 12 months, and the other two children (a daughter, 15 years old and a son, 10 years old) still go to school. Even though none of their children are married, the parents have already decided to give 0.15 hectares of farmland and one cow15) to each child. This case demonstrates the normative aspect of the equal division of land—parents decide to divide land equally among their children even when it is not clear at all what their children might become and where they might live in the future.
On the other hand, especially when land endowment is small, some parents decide not to give land (at least for the time being) to children who can make a living without land from their parents. Household No. 150 in K is an example. The parents currently own only 0.48 hectares of land. Among their three children (all sons), the first two are married. The parents gave land, albeit small (0.12 hectares), only to the eldest son, aged 38. This son married a woman from the same district when he was 22, and lives in his wife’s village (information on whether his wife received land from her parents is not available). His major occupation is farming, though he had also worked in Phnom Penh as a construction worker for 1 month in the last 12 months. The second son (28 years old) married a childhood friend from K. Though he has not received land from his parents, his wife received land from her parents. The couple resided in K at the time of the survey and makes a living by farming without migrating. Their mother (as a respondent of the interview) answered that she would give farmland to all her children in future. However, it is unlikely that land will be given to the second son in the near future because eight years had already passed since he got married—it is unusual in the villages surveyed to give land to children many years after their marriage. Judging from the fact that the second son and his wife can support themselves without labor migration, his wife seems to have received enough farmland from her parents. This is probably why the parents have not given land to the second son so far.
In other cases, parents give land to a limited number of their children even though they have enough land to divide among all their children. A couple in K (household No. 200) have four children, all of whom are married. But the couple gave all their farmland (2.5 hectares) to the eldest child (daughter) and are now landless. This means that their other three children will never receive land from them. The eldest daughter married a relative from the same commune as K. She and her husband currently live in K and make a living by cultivating the land she received from her parents (her husband did not receive land from his parents); they do not engage in labor migration. The second and third children (both sons) found partners from the same district as K and each lives in the village of his wife. While they cultivate land received from their wife’s parents, they also go to Phnom Penh seasonally to work as construction laborers. The youngest child (daughter) had worked at a garment factory in Phnom Penh before marrying a relative from the same commune as K. She now lives in the village of her husband and makes a living by farming (with land her husband received from his parents) and running a grocery shop. While one possible reason these three children did not receive any land from their parents is that they inherited land from their spouse’s parents, we cannot ignore the possibility that their parents gave them a large sum of money as a substitute for land.16) This example suggests that children would accept impartible land transfer if they receive a sufficient amount of other assets as a substitute for land. But this also means that ordinary parents who do not have major assets other than land, must divide land among all their children, even if it causes farm fragmentation, if they want to maintain equality among children. In fact, of those children who have not received land from their parents, only a small proportion has received non-land assets, as we shall see in Section VII-2. It is also rare for parents to give financial assets to their children as a substitute for farmland, as in the example above.17)
V-3 Children’s Prospect of Receiving Land in Future
As mentioned above, married children yet to receive land at the time of the survey stand the chance to receive land in future as long as parents still own enough land. In other words, the prospect of receiving land in future is small if parents’ current land endowment is small. Thus the distribution of parents’ current land endowment is an indicator of how many children are unlikely to receive land from their parents.
In this subsection, we estimate the proportion of children who are unlikely to receive land from their parents in future, among children yet to receive land at the time of the survey, by the following method.
We assume for simplicity that parents will not buy or sell land in future. Parents own land of size T (hectares) and have N children to whom they have not given land. Let t (hectares) be the size of land parents will give to each child, or “land transfer size.” So the number of children to whom parents can give land in future is formed by T/t. If N<T/t, parents can give land to all of the N children, but if N>T/t, then N–T/t of their children cannot receive land. By calculating N–T/t for all households with positive N, we can estimate the proportion of children who are unlikely to receive farmland from their parents in future, among children yet to receive farmland at the time of the survey (including both married and unmarried children but limited to those aged below 40).
Because this is a simulation, t can take any value, but we run our estimation on t=0.2, 0.4, and 0.8 hectares. This is because these values are close to the average size of land that children received from their parents in S, K, and Y respectively (see Table 5). The proportion of children unlikely to receive land estimated based on t=0.2 can be regarded as the minimum estimate because there are very few cases in which land transfer size is less than 0.2 hectares in the villages surveyed. Even in land-scarce village S, such cases account for only 9 percent of land transfer. This suggests that parents avoid dividing land to such a small scale from the standpoint of economic viability.
The simulation results are presented in Table 8. The important point to note is whether the proportion of children not receiving farmland from their parents will increase if the current level of land transfer size is maintained. For village S, under the current transfer size of around 0.2 hectares, the simulated proportion is 16 percent, which is slightly lower than the proportion of those who have not received land from their parents among married children under 40 (19 percent). Therefore, as long as parents maintain the current level of transfer size, though it is already the minimum level, the proportion of children not receiving land will not increase in the near future. Village K is in the same situation. Under the current transfer size of around 0.4 hectares, 28 percent of children will not be able to receive land from their parents. Though the value is high, it is comparable to the current level (29 percent). If land transfer size is reduced to 0.2 hectares, the proportion will even decrease to 16 percent.
The situation in Y is different. If parents give their children land up to 0.8 hectares, as they have done so far, as much as 44 percent of children will be unable to receive land, which is much higher than the proportion of married children under 40 currently not receiving farmland from their parents (17 percent). Though the proportion of children not receiving land will be smaller if the transfer size is reduced to as small as 0.2 hectares, the simulated value is still 23 percent, which is higher than even S and K. The high value in Y reflects the fact that a larger proportion of households (parents) in Y are landless.
These results indicate that the proportion of children who do not receive land from their parents could increase to a greater extent in land-rich villages rather than in land-poor villages. In the case of Y, the negative prediction is a result of significant inequality in landholding (there are many landless households). It also suggests that this landholding inequality may increase further in the near future.
VI Does Migration Experience Matter?
VI-1 Relationship between Migration and Land Transfer
As presented in the previous section, a number of married children have not received farmland from their parents. In this section we examine whether children’s lack of access to their parents’ land is related to their migration experience. To investigate the effect of migration rigorously, we pay attention to children’s labor migration experience before marriage because their migration decision after marriage is highly influenced by whether they receive land from their parents and how much.
Table 9 shows the difference in proportion, among married children aged under 40 who received land from their parents, between those who migrated before marriage and those who did not. According to the table, the proportion is around 30-percentage point lower for those who migrated before marriage, irrespective of the village and the gender of the child. One possible reason is that those with migration experience are on average younger than those without, as migration has become more widespread recently. Being young, many of the children have just got married and have not received land from their parents yet (although they will a few years later). However, even discounting the age factor, we still observe a negative effect of premarital migration experience on land transfer (the lower part of Table 9).
We should also take note of the effect of land endowment. Children of parents with small land endowment would be more likely to migrate before marriage to supplement family income; such children would be less likely to receive land from their parents simply because their parents do not have much land. However, Table 10 shows that, even after taking into account parents’ land endowment measured by OLE, in S and K villages, children with premarital migration experience are still less likely to receive land from their parents. In Y, on the other hand, migration experience does not cause a significant difference. The exact reason is not known, but it might be because labor migration from Y is a more recent phenomena and the rate of migration in Y is lower than in S and K. The difference between Y and the other two villages suggests that premarital migration experience begins to exert an influence on parents’ land transfer decision when migration becomes very common among young people.
Why does migration experience deter land transfer? One factor that links land transfer to migration experience is children’s postmarital place of residence. As argued in Section II, some of the children who migrated when they were single continue to live at the migration destination even after they get married. In addition, if migrating children marry migrants from other regions whom they met at their migration destination, they might settle in the home province of their spouse. Parents would then withhold land from children who leave their home province because they cannot (and do not need to) cultivate land in the village.
Therefore, in the following subsection, we examine whether children’s premarital migration experience leads to provincial exogamy (marriage with a person from a province other than the home province) and residency in another province. We define place of residence by whether children live in their home province or another province, because it is generally very difficult to cultivate land in the home village while living in another province. This is not the case for children living in other districts of the same province, as districts are close by in some cases.18) In fact some children living in others districts of the home province cultivate their land in their village of origin. Distinction in the place of residence by province leads to distinction between provincial exogamy and provincial endogamy (marriage with a person from their home province) simply because provincial exogamy would be more likely to result in residency in another province (= the place of origin of their spouse).
VI-2 Premarital Migration Experience, Marriage, and Postmarital Residency
Fig. 2 illustrates the relationship between premarital migration experience and the prevalence of provincial exogamy, with controlling for parents’ land endowment (OLE). This figure indicates that, in all three villages, those with migration experience before marriage are more likely to marry a person from a province other than their home province. Interestingly, parents’ land endowment is also associated negatively with the prevalence of provincial exogamy, except for those without premarital migration experience in S and K.
Among the cases of provincial exogamy, some couples were matched through inter-provincial kin ties. Such matching is not unusual, especially in Y, because there are many villagers who originate from Takeo or Kampot province, as mentioned above. People living in neighboring provinces may also come to know each other from interactions in daily life.19)
Nevertheless, the reason why those who have premarital migration experience are more likely to marry with a person from another province is that they met their partner at the migration destination. We asked parents how their children aged below 40 got to know their partner and found that out of 158 children with premarital migration experience, 56 percent married a person they met at the migration destination, 75 percent of whom are from other provinces. By comparison, among 415 children with no premarital migration experience, there are 33 cases of provincial exogamy, but in 24 of them (73 percent), the couples met at the time of engagement (in marriages arranged by parents), or they already knew each other as they are relatives.
Next, we examine the relationship between migration experience and the current place of residence of married children. Children’s residency is not an unambiguous concept if we take account of labor migration. In this paper, even if a child works in a remote place (such as Phnom Penh) for most of the year, his/her home village is regarded as the place of residence as long as he/she has a house in the village or is still considered as a member of the household by the household head (=the parent).
Fig. 3 shows the relationship between premarital migration experience, the origin of the spouse, and the current place of residence of married children aged under 40, with controlling for parents’ land endowment. The data is the total of the three villages because some categories have too few samples if we were to divide the data by village. This figure indicates that even if children experienced migration before marriage, only a very small fraction of them live in another province after marriage if they are married to a person from their home province (“Provincial endogamy & Migrated”). It also demonstrates that while children married to a person from another province (“Provincial exogamy”) are more likely to live in another province, such a tendency is stronger for those who have premarital migration experience.20) Furthermore, out of 29 children who married a person from another province after experiencing migration and who currently live in another province, 23 live in the home province of their partner.
In summary, children’s premarital migration experiences promote provincial exogamy by providing young migrants with opportunities to meet migrants from various provinces. This provincial exogamy in turn increases the number of children who do not return to their home village but move to the place of origin of their spouse in another province.
VI-3 Effect of Migration Experience on Land Transfer
Finally, we examine whether and how provincial exogamy and residency in another province promoted by premarital migration experience affects land transfer from parents to their children. Before presenting the compiled data, we give examples of children who apparently do not receive land because of their premarital migration experience, so as to understand concretely how the migration experience can affect intergenerational land transfer.
The eldest son (37 years old) of a couple in Y (household No. 164) is an example of a child who did not receive land probably because he can make a living at the migration destination. The couple has nine children, six of whom are married (one of them is divorced). They gave 0.27 hectares of land to all six of these married offspring, except the eldest son, who was also the only one out of the six to have migrated before marriage. He was employed in waste materials collection in Phnom Penh,21) and at the age of 27, he married a woman who also did the same job in Phnom Penh. In fact, both husband and wife are from the same district and have known each other since they were children. Neither husband nor wife has received farmland from their parents and both continue to make a living by collecting disused articles in Phnom Penh. Apparently, this is not because his parents do not have much land; in fact they still have 1.5 hectares of rice field. When asked about their land transfer plans to their children in the future, the parents responded that they would not give land to children who lived in a remote place. In other words, the eldest son would receive land if he returned to the village, which is unlikely to happen as he can make a living, albeit a difficult one, in Phnom Penh.
The second example (household No. 122 in S) is a case in which children married a person from another province whom they met at the migration destination, then moved to the place of origin of their partner after marriage. The couple has nine children, five of whom are married. Among these five married children, only the first child, a son, has received land (0.18 hectares) from his parents, with the remaining 0.8 hectares of land being cultivated by the parents. The eldest son migrated to Phnom Penh to work but married a childhood friend from S (who also migrated to Phnom Penh). Though his wife also received land from her parents and the couple lives with the husband’s parents (household No. 122), both of them also migrated to Phnom Penh to work. Of the other four married children, three sons migrated to Phnom Penh or Siem Reap when they were single and all married someone from another province whom they met at the migration destination. The three sons have not received land from their parents as mentioned above and currently live in Kampong Speu, Kampong Thom, and Siem Reap provinces respectively, which are probably the place of origin of their wives. Though information on whether their wives received land from their parents is unavailable, the three sons all engage in farming in their place of residence. Another married child (daughter) had worked at a garment factory in Phnom Penh before marriage and settled down with a man from another province whom she met in Phnom Penh. Her husband had worked at an NGO in Phnom Penh and has also not received land from his parents. The couple continues to work at the same jobs in Phnom Penh as before their marriage.
These examples demonstrate that premarital migration experience of children lead to settlement in another province such as the migration destination or the place of origin of their spouse, and that such children are less likely to receive land from their parents. However, there is another possibility that children move to another province precisely because they cannot receive land from their parents. In other words, in some cases, children decide where they settle subsequent to marriage, based on whether they will receive land or not from their parents. With this possibility in mind, I merely confirm here that the lack of land transfer from parents to children is closely associated with children’s moving out of the home province, which is promoted by their premarital migration experience.
To ascertain whether the above examples are exceptions or the norm, we examine the data of our sample as a whole. Fig. 4 illustrates the relationship between migration experience of married children aged under 40, the place of origin of their partner, and land transfer from parents, with controlling for land endowment of parents. The following is demonstrated. First, marrying a person from another province (provincial exogamy) does not have a great effect on the probability of land transfer for children with no premarital migration experience, but it does have a large impact for children with migration experience, especially among those whose parents’ land endowment is small. Second, while the negative effect of migration experience is found even among children married with a person from their home province, its effect is much greater among children married with a person from another province. In conclusion, provincial exogamy after migration experience is negatively associated with land transfer from parents, especially when parents’ land endowment is small.
The data presented in Fig. 4 does not, however, take account of the effect of the place of residence. As argued above, if provincial exogamy tends to lead to residency in other provinces, the effect of the marital partner’s place of origin demonstrated by Fig. 4 may be superfluous and merely reflect the effect of the place of residence.
To check whether this is the case, we prepared Fig. 5, showing the relationship between place of origin of spouse, the current place of residence, and land transfer from parents for married children under the age of 40 with premarital migration experience. Examining the findings, especially the category of OLE<0.6 because the samples in the OLE≥0.6 categories are scarce,22) we notice the following points. First, while children living in other provinces are less likely to receive land from their parents, among these children, there is no considerable difference depending on the place of origin of spouse. In other words, irrespective of the place of origin of spouse, those who live in other provinces are less likely to receive land from their parents. This indicates that it is the place of residence rather than the place of origin of spouse that greatly affects the land transfer decision.
Living in another province need not necessarily deter land transfer if the renting out of land were possible. In fact, land rental exists in the surveyed villages, even if it is not widely prevalent, as 11 percent of sample households rent farmland from other households. In addition, migration is one of the major reasons prompting landowners to rent out their land, accounting for 11 out of 66 cases of rental of farmland. Therefore, the observed negative association between residency in another province and land transfer by parents suggests that the renting-out of land is not a preferred option for farmers in the surveyed villages for some reasons, though examining the validity of this supposition is beyond the scope of this study.
Furthermore, as mentioned earlier, it is also highly likely that some of the children who moved out to another province did so precisely because they did not expect to receive land from their parents. In such a case, the possibility of land rental in their home village is irrelevant to their decision.
Second, among children who live in their home province, those who married a person from another province are less likely to receive land from their parents compared with those who married a person from their home province. If all land-endowment groups are combined, while 77 percent of the latter received land, only 38 percent of the former have been given land. There are 18 such children in the sample. It seems that some of them have yet to receive land because they got married very recently, but the majority of them are children of land-poor parents: among the 18 children, 6 are children of landless parents and the other 6 are children of parents with OLE of less than 0.2 hectares (though not zero). Household No. 1 of K is an example. The parents of this household own only 0.3 hectares of farmland, though they also rent land from another household. Among their five children, the first three (all daughters) are married but have not received any land from their parents. These three daughters had migrated before marriage and had met their spouses, who were also from other provinces, at the migration destination. The eldest daughter worked at a garment factory in Phnom Penh. While working in the capital, she met her future husband, who was a monk at that time, and married him in 2007. The second and the third daughters went to Koh Kong province to work in fishery-related jobs. The second daughter still lives in Koh Kong, after getting married in 2004 to someone she met there and who is probably from Koh Kong. The third daughter got married in 2007 to a man who had left his own province to work in fishery in Koh Kong. Their husbands have not received land from their parents either. Currently the eldest and the third daughters, as well as their husbands, live with their parents in K without migrating to other places, and help the parents with farming (and supposedly also engaging in agricultural wage labor in the village).
What does this mean? One possible inference is that children who cannot expect to receive land from their parents, who have small land endowment, try to find their marital partner at their migration destination. This is supported by the findings of Yagura (2012) that among young unmarried migrant workers in Phnom Penh, those who do not expect to receive land from their parents are more likely to hope to marry someone they meet in Phnom Penh, rather than someone from their place of origin. This finding is also consistent with the finding of this paper that among children with premarital migration experience, the smaller the land endowment of parents, the higher the prevalence of provincial exogamy (Fig. 2).
Yagura (ibid.) proposes two possible reasons why the availability of farmland from parents affects marital partner selection. First, those who do not have farmland feel handicapped in their search for a potential partner in their locality because they are regarded as less desirable a marital partner (at least from the economic standpoint); therefore they have to find their partner at their migration destination. This choice can produce an economically better outcome if the partner comes from a more affluent region than their own place of origin; in such a case, their living conditions can be improved by relocating to the partner’s place of origin.
Second, if a child does not have farmland in his/her home village and thus does not engage in farming, marrying a local is of lesser importance, economically speaking. Farming often requires support from family members (parents and sibling), thus if one engages in farming in one’s place of origin, it is to one’s advantage to marry with a local because support from the spouse’s family is also easily available.23) There is no such advantage in marrying with a local if one does not engage in farming.
VII Is Lack of Land Transfer Compensated for?
VII-1 Land Transfer from Spouse’s Parents
Even if a child does not receive land from his/her own parents, he/she can still expect that his/her spouse be given land from his/her parents because both sons and daughters usually receive farmland from their parents, as is evident from Table 5. One might also suppose that parents would give more land to their child if the spouse of their child does not receive land from his/her parents.
What happens in reality is the opposite. Among married children aged under 40, the proportion of those who received land from the spouse’s parents24) is 89 percent for those receiving land from their own parents, but only 52 percent for those who do not receive land from their own parents (the difference is statistically significant at 1 percent level). Such differences are present even if the sample is divided according to village, age, gender, place of residence, and year of the marriage of children.
This gap is not caused by a bias in the distribution of parents’ occupations. If parents engaging in non-farm activities as their major occupation tend to marry their children off to children from other non-farming family, parents of both the groom and the bride would not give land to their child as they do not have land, or land is not important for the newly-wed couple who will engage in non-farm activities like their parents. In reality, however, such couples account for only 7 percent of couples who have not received land from either the husband’s or wife’s parents. Rather, the positive association between land transfer from the husband’s parents and the wife’s parents implies that children who are unable to receive land from their parents are less likely to be chosen as a partner by those who can expect land from their parents. In short, “positive assortative matching” takes place in the three villages.25)
In fact, married children aged under 40 who have not received land from either their own parents or their spouse’s parents constitute a very small minority: they account for only 9.8 percent of married children under 40, while 70.9 percent received land from both their own parents and their spouse’s parents (the rest received land either from their own parents or their spouse’s parents). But these compositions differ according to premarital migration experience and the place of origin of spouse, as presented in Table 11. This table reveals the following two points. First, even if they receive land from their own parents, children married to a person from another province are less likely to receive land from their spouse’s parents, regardless of their premarital migration experience (93.0 vs. 59.3; 87.1 vs. 72.0). This is probably because provincial exogamy tends to lead to settlement in another province on the part of their spouse.
Second, of children with premarital migration experience and a spouse originating from another province (most of them meet their partner at the migration destination), only 38 percent received land from their spouse’s parents if they have not received land from their own parents. This makes sense if the children marry a person they had met at the migration destination, for the very reason that they did not expect to receive land from their parents, as argued above. Because the situation of their partner is similar (otherwise he/she would not consider searching for a mate at the migration destination), couples formed at migration destinations are less likely to receive land from either the husband’s parents or the wife’s parents.
In summary, the data presented above clearly indicate that for children not receiving land from their parents, their spouses are also unlikely to receive land from their parents, and that such a tendency is especially evident among couples who meet at migration destinations.
VII-2 Transfer of Non-land Assets
Children who do not receive land from their parents need not face great difficulty in making a living if they were to receive other assets in lieu. In the villages surveyed, major non-land assets given by parents to their children include draft animal (cattle and buffalo) and financial assets (cash and gold). Though we do not have information regarding the transfer of such non-land assets from the spouse’s parents, we collected data on whether and how household heads and their spouse (as parents) gave non-land assets to their children.
Paradoxically, however, the data demonstrate that most children who did not receive land from their parents did not get non-land assets either. While the proportion of children receiving draft animals is 60 percent for those receiving land, the proportion is only 15 percent for those not receiving land. Similarly, while 29 percent of those who received land also received financial assets, only 18 percent of those not receiving land were given financial assets.
Migration experience is also associated with the transfer of non-land assets. As shown in Table 12, the likelihood of receiving non-land assets is especially small for those who have premarital migration experience and who did not receive land from their parents. This is probably because parents of these children are poor in the first place (that is why their children must migrate to support their family) and thus do not have many assets to bestow on their children. Whatever the reason, the data suggest that a large proportion of children not receiving land from their parents, especially those with premarital migration experience, cannot aspire to receive other assets from their parents and are therefore forced to start their married life under difficult economic conditions.
VIII Concluding Remarks
As is evident from parents’ plans and actual land transfer, equal division among all children is the most favored practice in the villages surveyed, despite the intensifying scarcity of land and increased labor migration in recent years. This implies that there will be further farm fragmentation in the villages surveyed. At the same time, however, parents with very small land endowment are unable to divide land equally among all their children, and some of them choose not to give land to children who can make a living without land. Though this situation leads us to expect an increase in the proportion of children unable to receive land from their parents in the near future, the data also suggest that the increase would not be very dramatic.
The recent increase in migration opportunities helps parents to maintain equal division to a certain extent, by enabling children to make their living with only a small piece of land. Partly for that reason, expanding migration opportunity has not caused fundamental changes in land transfer practices, but it has indeed affected intergenerational land transfer. Especially in villages S and K, where youth labor migration has been widespread, children with premarital migration experience are less likely to receive land from their parents even after controlling for parents’ land endowment. Premarital migration experience of children is negatively associated with land transfer, especially when children settle in the migration destination or when they marry a person from another province whom they met at the migration destination, then move to their partner’s place of origin.
On the other hand, children who migrated before marriage and then married a person from another province—mostly with partners encountered at the migration destination—are less likely to receive land from their parents, even if they live in their home province after marriage. This indicates that children who had little expectation of receiving land from their parents try to find their partner at the migration destination.
Given this perspective, the argument that expansion of labor migration opportunities increases the number of children not receiving land is not very accurate. A more realistic description of the current situation of the villages surveyed is that, taking advantage of labor migration experience, children with little prospect of receiving land from their land-poor parents, choose to leave their home province and find means to make a living without land. To put it from a different perspective, as long as parents have enough land, their children are inclined to settle in their home province even if they migrated when they were single; such inclination on the part of the children also leads their land-rich parents to give land to all their children.
The contrast between land-poor families and land-rich families—in the emigration of children of land-poor families and the division of land among children of land-rich families—is very similar to the situation in a rural village in Tokugawa Japan outlined by Hayami (1983).26) Hayami argued that different responses to changes in economic conditions between the classes (defined by the size of landholding) led to inter-class mobility. Similar changes can happen in the long run in the Cambodian villages surveyed. That is, the above-mentioned differences between classes in land transfer practice and children’s behavior imply that increase in migration opportunities can eventually reduce landholding inequality among households within a village by the following mechanism. Further fragmentation of smaller farms would be avoided as parents choose to give land only to a limited number of their children, made possible by the expansion of migration opportunities. In addition, because children who do not receive land tend to move out of the village, increase in landless households in the village would be contained.27) On the other hand, large-scale landholders will experience farm fragmentation as they continue to divide land among all their children, and the landholding of their offspring will get smaller and smaller.
The decrease in inequality does not necessarily imply improvement in livelihood for the lower economic stratum. As presented in Section VII, most of the children who did not receive land from their parents are in difficult economic conditions, as they do not receive other assets and their spouses are also less likely to receive land from their parents. This situation also implies positive assortative matching exists in marital partner selection in the villages surveyed, which may increase economic inequality among households.
Though increasing migration opportunities can have a large impact on land distribution or economic inequality among households in a village, as discussed above, whether, how, and to what extent that actually happens or will happen are beyond the scope of this paper and require future research. To tackle these questions, well-planned research with a long-term perspective is needed to collect detailed longitudinal data at the household as well as individual level in specific villages.
Another issue that this paper does not address is family conflict that may arise due to unequal division of land among children. With unequal division of land, children who receive neither land nor financial compensation from their parents may feel dissatisfied. Such dissatisfaction may cause a rift within the family, which can have a negative socio-economic impact on rural households and society. Our data is insufficient to examine this issue and further research is required.
Accepted: September 24, 2012
This study was supported by KAKENHI [Grant-in-Aid for Young Scientists (B) (20780168)]. I am indebted to Mr. Dork Vuthy from the Royal University of Phnom Penh and the research assistants for preparing and conducting the field survey. I also extend my heartfelt gratitude to people of the surveyed villages for cooperating with our survey.
Amakawa Naoko 天川直子. 2001. Nochi Shoyu no Seido to Kozo: Poru Poto Seiken Hokaigo no Saikochiku Katei 農地所有の制度と構造―ポルポト政権崩壊後の再構築過程 [System and structure of farmland ownership: Restructuring process after the Pol Pot regime]. In Kambojia no Fukko-Kaihatsu カンボジアの復興・開発 [Reconstruction and development in Cambodia], edited by Naoko Amakawa 天川直子, pp. 151–211. Chiba: Institute of Developing Economies.
André, Catherine; and Platteau, Jean-Philippe. 1998. Land Relations under Unbearable Stress: Rwanda Caught in the Malthusian Trap. Journal of Economic Behavior & Organization 34(1): 1–47. doi: 10.1016/s0167-2681(97)00045-0.
Cambodia, National Institute of Statistics, Ministry of Planning. 2010a. General Population Census of Cambodia 2008. Analysis of the Census Results Report 1, Fertility and Mortality in Cambodia. National Institute of Statistics, Ministry of Planning, Cambodia.
―. 2010b. General Population Census of Cambodia 2008. Analysis of the Census Results Report 6, Migration. National Institute of Statistics, Ministry of Planning, Cambodia.
Ebihara, May Mayko. 1971. Svay: A Khmer Village in Cambodia. PhD dissertation, Columbia University, Ann Arbor, Mi, Umi Dissertation Services.
Emigh, Rebecca Jean. 1997. Land Tenure, Household Structure, and Age at Marriage in Fifteenth-century Tuscany. The Journal of Interdisciplinary History 27(4): 613–635.
Goldschmidt, Walter; and Kunkel, Evalyn Jacobson. 1971. The Structure of the Peasant Family. American Anthropologist 73(5): 1058–1076. doi: 10.1525/aa.1971.73.5.02a00060.
Habakkuk, H. J. 1955. Family Structure and Economic Change in Nineteenth-Century Europe. The Journal of Economic History 15(1): 1–12. doi: 10.1017/S0022050700066092.
Hayami, Akira. 1983. The Myth of Primogeniture and Impartible Inheritance in Tokugawa Japan. Journal of Family History 8(1): 3–29. doi: 10.1177/036319908300800101.
Khera, Sigrid. 1973. Social Stratification and Land Inheritance among Austrian Peasants. American Anthropologist 75(3): 814–823. doi: 10.1525/aa.1973.75.3.02a00120.
―. 1972. Kin Ties and Social Interaction in an Austrian Peasant Village with Divided Land Inheritance. Cross-Cultural Research 7(4): 349–365. doi: 10.1177/106939717200700404.
Kobayashi Satoru 小林知. 2007. Poru Poto Jidai Igo no Kambojia ni okeru Nochi Shoyu no Hensei Katei: Tonre-Sappu Ko Togan Chiiki Noson no Jirei ポルポト時代以後のカンボジアにおける農地所有の編成過程―トンレサップ湖東岸地域農村の事例 [Agricultural land right resolution process in post-Pol Pot rural Cambodia: A case study from the eastern Tonle Sap region]. Ajia, Afurika Chiiki Kenkyu アジア・アフリカ地域研究 [Asian and African area studies] 6(2): 540–558.
Ledgerwood, Judy L. 1995. Khmer Kinship: The Matriliny/matriarchy Myth. Journal of Anthropological Research 51(3): 247–261.
Platteau, Jean-Philippe; and Baland, Jean-Marie. 2001. Impartible Inheritance versus Equal Division: A Comparative Perspective Centered on Europe and Sub-Saharan Africa. In Access to Land, Rural Poverty, and Public Action, edited by Alain de Janvry, Gustavo Gordillo, Elisabeth Sadoulet, and Jean-Philippe Platteau, pp. 24–67. Oxford: Oxford University Press. doi: 10.1093/acprof:oso/9780199242177.001.0001.
Ram, K. Anantha; Tsunekawa, A.; Saha, D. K.; and Miyazaki, T. 1999. Subdivision and Fragmentation of Land Holdings and Their Implication in Desertification in the Thar Desert, India. Journal of Arid Environments 41(4): 463–477. doi: 10.1006/jare.1999.0495.
Salamon, Sonya. 1980. Ethnic Differences in Farm Family Land Transfers. Rural Sociology 45(2): 290–308.
Takahashi Miwa 高橋美和. 2001. Kambojia Inasaku Noson ni okeru Kazoku Shinzoku no Kozo to Saiken: Takeo Shu no Jirei カンボジア稲作農村における家族・親族の構造と再建―タケオ州の事例 [Family-Kinship structure and its reconstruction in rural Cambodia: A case study of a rice-farming village in Takeo province]. In Kambojia no Fukko-Kaihatsu カンボジアの復興・開発 [Reconstruction and development in Cambodia], edited by Naoko Amakawa 天川直子, pp. 213–274. Chiba: Institute of Developing Economies.
Takeuchi Takao 竹内隆夫. 2000. Kazoku, Shinzoku, Jinko no Henyo 家族・親族・人口の変容 [Transformation of family, kinship, and population]. In Zoku, Tai Noson no Kozo to Hendo: 15 Nen no Kiseki 続・タイ農村の構造と変動―15年の軌跡 [The Structure and transformation of Thai farm villages, second series], edited by Osamu Akagi 赤木攻, Atsushi Kitahara 北原淳, and Takao Takeuchi 竹内隆夫, pp. 103–132. Tokyo: Keiso Shobo.
Wegge, Simone A. 1999. To Part or Not to Part: Emigration and Inheritance Institutions in Nineteenth-Century Hesse-Cassel. Explorations in Economic History 36: 30–55. doi: 10.1006/exeh.1998.0703.
Wiegandt, Ellen. 1977. Inheritance and Demography in the Swiss Alps. Ethnohistory 24(2): 133–148.
Yagura Kenjiro 矢倉研二郎. 2015. Effect of Intergenerational Asset Transfers on Land Distribution in Rural Cambodia: Case Studies of Three Rice-Growing Villages. Agricultural Economics 46: 173–186. doi: 10.1111/agec.12149.
―. 2012. Does Labor Migration Offer Opportunities for Meeting Prospective Spouses? The Case of Migrant Workers in Cambodia. Population, Space and Place 18(3): 277–294. doi: 10.1002/psp.683.
―. 2008. Kambojia Noson no Hinkon to Kakusa Kakudai カンボジア農村の貧困と格差拡大 [Poverty and widening economic disparity in rural Cambodia]. Kyoto: Syowa-do.
―. 2005. Imperfect Markets and Emerging Landholding Inequality in Cambodia. Japanese Journal of Rural Economics 7: 30–48.
1) In this paper, the term “transfer” is used instead of “inheritance” to refer to the giving of land by parents to their children, because it is generally conducted while parents are still alive.
2) For example, in the eastern part of the Treang district, Takeo province, a vast area of dry season rice fields was reclaimed in the 1980s, but the reclamation has stopped as unreclaimed land has almost run out (Yagura 2008).
3) Yagura (2005) examines factors determining land distribution in two Takeo villages, but changes in land distribution in the long run are not fully captured because data used is basically cross-sectional. In addition, the effect of changes in land transfer practice is not addressed.
4) See Yagura (2015), which, using the same household data as the present paper, is focused on the effect of intergenerational asset transfers on land distribution in the three villages surveyed. (This note is added after the acceptance of the present paper.)
5) For example, under a manorial system where the lord of a manor imposes tax and labor on tenant farms and sets the rules for transfer of the rights to land between tenants (Platteau and Baland 2001).
6) In rural villages in northeastern Thailand, the practice of partible land inheritance has largely been maintained in the 1980s and 1990s even though farm size has shrunk. Takeuchi (2000) attributes this phenomenon to productivity improvement in rice farming.
7) The reverse can also bring about discord. André and Platteau (1998) show that in a Rwandan village, a change in land inheritance practice from primogeniture (all land given to the eldest son) to division among sons brought about intra-family dispute, provoked by the eldest son who was deprived of his privilege.
8) According to the population census conducted in 2008, population outflow due to inter-provincial migration for five years up to 2008 was 70,357 for Prey Veng and 12,508 for Pousat (Cambodia, National Institute of Statistics 2010b). The ratio of the outflow to total population as of 2008 is 7.4 percent for Prey Veng and 3.1 percent for Pousat, which respectively rank 1st and 7th among 24 provinces including Phnom Penh.
9) The krŏm samokki is a group composed of about 10 households organized to practice collective farming in rural Cambodia under the People’s Republic of Kampuchea regime. As is the case of the three villages in this paper, krŏm samokki were mostly dissolved by the mid-1980s (Amakawa 2001).
10) In this paper, land that the informant considered as his/her own is regarded as being owned by the informant (or his/her household), whether the plot is officially registered or not.
11) Among 582 married children aged below 40, 72 persons received residential land from their parents and all but 4 of them also received farmland from their parents.
12) Among married children under 40, the proportion of those who live with their parents or their spouse’s parents is 54 percent for the youngest children and 30 percent for the other children.
13) The age group 15–19 of village K is exceptional in that it includes only four persons. Thus no overall trend can be drawn from it.
14) Though the size of land plot can have an effect on parents’ land transfer decision, we are unable to explore this factor because of the lack of plot-level data. Nevertheless, the effect of plot size in the surveyed villages is probably negligible because parents can divide a plot up when distributing land to children.
15) “Cow” here indicates either male or female and castrated or uncastrated cattle.
16) The amount of money given to each of the three children is 4 million riels, 3 million riels, and 9.5 million riels, respectively. These amounts are comparable with one to two hectares of wet season rice field, as one hectare of wet season rice field in K is priced at 3–5 million riels as of 2009.
17) Of 39 households (couples) with more than one married child aged under 40 and who gave farmland to only some of these children, just 4 gave financial assets exclusively to those children who did not receive any land.
18) Kampong Leav district, which is very close to village K (in Prey Ro district), is such a case. It would have been preferable to measure the remoteness of residency by calculating the distance from the home village, but the necessary information was not available.
19) For example, many villagers from Y paired up with people from Battambang province through such interactions.
20) Incidentally, the prevalence of the residency in another province tends to increase with OLE for the cases of provincial exogamy. This is because a large portion of children marrying under provincial exogamy are from land-rich village Y.
21) This involves collecting waste such as cans, bottles, and paper by cart around the city to sell to junk dealers.
22) These categories include only a few cases each.
23) According to the data collected through the survey for this study, parents tend to help their children in farming if the child lives in the home village, irrespective of the gender of the child. From the perspective of a married couple, this means they can count on the help of both sets of parents (the husband’s and the wife’s) if their parents live in the same village.
24) In the Cambodian context, parents give land to their own child, not to the couple. But in this paper, I use statements such as “children received land from their spouse’s parents” for simplicity.
25) Existence of the positive assortative matching in the surveyed villages is more rigorously examined by Yagura (2015). (This note is added after the acceptance of the present paper.)
26) According to Hayami (1983), contrary to the stereotype of the Japanese family system, diverse patterns of land inheritance practice, including partible inheritance, existed in the Tokugawa era-village that he studied.
27) This prediction does not necessarily contradict with the existence of landless households in the surveyed villages. In fact, in accordance with this prediction, the proportion of landless households is lower for younger generations than for older generations (4.4 percent for households whose heads are in their twenties or thirties and 11.7 percent for households whose heads are in their forties, fifties, or sixties).