Table 36 - Selected demographic and socioeconomic characteristics of households: Bangladesh, Malaysia and Philippines
|Fish. com.||Farm. com.||DHS 96/97||Fish. com.||Farm. com.||Censusa||Fish. com.||Farm. com.||DHS 1993|
|Type of households in study area:|
|% fisherfolk households||96||2||97||10||100||1|
|Fisherfolk/farming households contacted||(100)||(100)||(100)||(101)||(100)||(99)|
|Age distribution (%):|
|50 and above||6||7||12||5||7||13||10||23||14|
|(Female household members)||(311)||(299)||(385)||(337)||(252)||(244)|
|50 and above||8||9||13||13||14||12||12||24||12|
|(Male household members)||(338)||(367)||(392)||(381)||(300)||(278)|
|Average household size:||7.3||7.2||5.3||8.5||8.0||4.8||5.5||5.2||5.3|
|Percent female-headed households:||0||0||9.0||6.8||11.6||…||6.0||5.0||12.2|
|Rating of financial situation of household by heads of households (% of households):|
|More than sufficient||4||1||4||0||1||2|
|Percent of households who:|
Notes: …=not reported/not applicable/not asked.
a = 1991, rural areas only.
Table 37 - Selected demographic characteristics of eligible household members: Bangladesh, Malaysia and Philippines
|Fish. com.||Farm. com.||DHS 96/97||Fish. com.||Farm. com.||DHS 96/97||Fish. com.||Farm. com.||DHS 92/93|
|Average age of respondents:|
|Median age at first marriage:|
|Married men, 20–60||23||25||…||24||24||265||25||25||…|
|Married women, 20–49||16||15||14||19||19||22||23||22|
|Appropriate age at first marriage (median):|
|Married men: For men||25||27||25||25||25||25|
|Married women: For men||28||30||25||25||28||30|
|Average number of children ever born:|
|Married women 15–24 **||2.2||1.7||1.4||1.1||1.0||1.5||1.1||…|
|Married women 40–49 **||7.8||7.6||6.2||5.0||4.4||5.1||3.9||3.8|
|Married women 15–49||4.3||4.9||3.9||3.6||4.4||3.9|
|Perceived ideal number of children (average):|
|Sson preference - ideal sex ratio:||1.3|
|Use of any modern contraceptive method: (married women, 15–49)|
|% current users||28.7||27.8||30.8||26.2||13.2||27.6||11.2||16.7|
|Desirability of last birth by married women (15–49)|
|Percentage who wanted then||48||30||90||82||67||68|
|Percentage stating ‘Better if born later’||15||22||3||6||16||16|
|Percentage that did not want child||37||48||7||12||17||16|
|Percentage of married women (15–49) with more live births than the ideal number:|
|Excess is 1 child||8||12||3||8||19||12|
|Excess is 2 children||7||16||2||4||5||7|
|Excess is 3 children or more||26||26||2||1||15||12|
|Percentage of married women (15–49) in a polygamous union:|
Notes: … = not reported/not available/not asked.
1 = However, 39 percent of fisherfolk women report to be able to read “with difficulty”; in farming communities the percentage is 30.
2 = In practically all countries averages for this age group were based on less than ten responses.
3 = Chittagong Division.
4 = Instead of illiteracy rates, percent with no education was reported.
5 = Both sexes.
In conclusion, Bangladeshi fisherfolk were more disadvantaged in terms of living conditions and social welfare, than farming households. Part of this may be attributed to the fact that they experienced discrimination due to their low social status as Sudra caste Hindus in a predominantly Muslim environment. Illiteracy was particularly high among women in fishing communities. Fisherfolk women had somewhat higher fertility then farming women. Roughly half of the women in both types of community had ever used a modern contraceptive method, but only one in four was a current user, so discontinuation rates were high. The survey data show that a very large proportion of women did not successfully control their wanted fertility and this problem is perceived to be more acute in farming communities than in fishing communities.
Fisherfolk households were somewhat larger (8.5 persons) than farming households (8.0 persons), probably due to the higher prevalence of joint families in fisherfolk households. As table 36 shows, women less frequently headed a fisherfolk household (6.8 per cent) than a farming one (11.6 per cent).
Table 37 shows that few indicators exhibited differences between the respondents from the two types of household. As mentioned before, educational attainment in the villages was fairly high. Therefore, illiteracy was only observed in a small segment of the population. Female respondents, who were on average aged 31, had married fairly young (at 19) in both types of community. Respondents in fishing communities believed that the appropriate age of marriage for women should be 20 years, about two years less than the age perceived by married women in farming communities. Focus group members mentioned that men should marry much later than women, so they would be economically and psychologically better equipped to support a family. In the past, judging from the lifetime fertility experiences of older respondents, fertility must have been higher in fishing communities than in farming communities.31 Overall, respondents in fisherfolk households desired a higher number of children compared to those in farming households, particularly among married male respondents. This was also confirmed in the accounts from focus group discussions. Fisherfolk wanted more children, especially sons, to help take over fishing activities, to ensure financial support in old age and to carry on the lineage.
As can be seen in table 37, more than two thirds of the married women in fishing communities had ever used modern and effective contraceptives; this is almost 25 percent higher than the women in farming communities. Despite the high rates of discontinuation in both types of community, the current use rates of married women in fisherfolk households were still twice as high than in farming households. Discontinuation of modern contraceptives is partly explained by a switch to traditional methods, such as rhythm, periodic abstinence and withdrawal. Men in both types of community mentioned a preference for such natural family planning methods, whereas women preferred modern methods, particularly the pill. Despite the high discontinuation rates and the low current use rates of modern contraceptives, the majority of the women, particularly fisherfolk women, seemed to have control over their wanted fertility. Relatively few women mentioned that they did not want their last child born or that the last birth should have come later. Also, in both types of community, and particularly in fisherfolk households, couples succeeded in having the actual number of children they desired. However, a rationalization of unwanted births by respondents may have taken place. Focus group discussions indicated that having more children than initially desired was not considered problematic, as it encouraged parents to work harder and seek more sources of income to meet the cost of raising them. Having a large number of children was financially attractive once the children were grown. It meant more income for the household and provided old age security for the parents.
31 A difference also seems to exist between the younger generations of women of the two types of community, but the difference is not statistically significant.
About one in twelve marriages were polygamous. These unions appeared somewhat more frequently in farming than in fishing communities. Women accepted such marriages, provided the husband could afford it and was fair to his wives regarding sleeping arrangements. Some men in focus groups had a negative view and mentioned that men involved in polygamous marriages may be untrustworthy and that these marriages were not free from fights. There did not seem to be any significant differences between fishers and farmers regarding opinions on polygamous marriages.
In conclusion, although in the past the number of children born was higher in fishing than in farming communities, this difference was not (yet) observed in the fertility of the younger generation of women. Overall, however, men in fishing communities reported higher numbers of desired children and preference for sons than those in farming households. This could be related to the fact that fisheries more than farming are perceived as labour intensive, but it might also reflect higher perceived risks of infant/childhood mortality among fisherfolk. Overall, it seemed that fisherfolk women had slightly better control over their fertility than women in farming households.
Table 37 shows that respondents considered the appropriate age of marriage for women to be around 25. Focus group discussions resulted in lower estimates, but, in general, especially in the younger generation, there was no strongly preferred age of marriage. The main criterion was that the person, male or female was “ready” for the role of spouse, father or mother. Past fertility levels were higher in fishing than in farming communities, and this is partly due to the higher desired number of children among respondents in their mid-thirties from fisherfolk households. It was recognized across generations that the cost of children had increased, and that the ideal number of children had decreased. Focus group members mentioned that because of the high cost of raising them, it is better to have fewer children. On the other hand, it is deemed that in the long run families with a larger number of children will be better off economically. In both types of household men wanted more children than women, mainly for the continuation of the family name and the expectation of adding income-earning labourers to the household.
Focus group discussions revealed that in both types of community, modern family planning methods (pill, IUD, condom) were primarily used by women below 30. Older women preferred traditional methods (withdrawal, abstinence). In addition, some methods (tubal ligation, IUD) were said to be avoided because women - and even more men - believed that these methods might weaken a woman and lessen her work capacity. Household survey data revealed that ever-use and current-use rates of modern family planning methods in fishing communities were more than twice as high as in agricultural communities. However, fisherfolk women reported to have more children than desired to a greater extent (40 per cent) than women in farming households (31 per cent). When asked whether the last child born was wanted, there appeared to be no difference between the women from the two types of household: two out of three married women mentioned that the last birth was actually wanted at that time. A possible explanation for this may be that contraceptive use-effectiveness among fisherfolk women was lower than among farming women.32
32 Alternatively, the degree of rationalization of unwanted pregnancies is higher among fisherfolk women, perhaps as a consequence of low use-effectiveness.
In conclusion, the traditional livelihoods in the Philippine fishing and agricultural communities do not in themselves provide sustainable and sufficient income sources to support adequate living conditions for the inhabitants. Although fisherfolk were still better off than farming households, the perceived prospects were grim. In both types of community, the younger generation was seeking alternative and/or secondary livelihoods, and some people preferred to migrate out of the communities to urban areas. The conditions favouring high fertility in the past have changed. Younger people favoured marriage at a later age, wanted fewer children, and used modern contraceptives more often than the older people, particularly in fishing communities. However, fisherfolk women had more children than the number desired by farming women, which may indicate that use-effectiveness of modern contraceptives was lower among fisherfolk.
Fertility and childhood mortality: a comparative perspective
The populations surveyed are located in countries that differ greatly in economic and sociocultural terms Therefore, if, in spite of small sample sizes and great variations in context, the survey data revealed consistent differences in specific socioeconomic and demographic variables, it must be concluded that these types of community differ economically and demographically.
The data from Tanzania and Bangladesh posed conceptual and analytical problems In Tanzania, it was not possible to identify “genuine fishing communities” and “genuine farming communities”. According to the focus group reports and the survey data analysis, most of the fisherfolk in the selected households combined farming with part-time fisheries and animal husbandry. Therefore, the hypothesized effects of livelihood characteristics on reproductive perceptions and behaviour are very hard to disentangle. As for Bangladesh, a predominantly Muslim country, coastal fisheries are the domain of lower-caste Hindus, who are a culturally distinct and economically disadvantaged sub-group. The research data revealed that all the farming community households were Muslim, and all the fishing community households were Hindu. If the lower-caste Hindus had different fertility aspirations because of religious norms and values than the Muslim families, the impact of livelihood on fertility is difficult to assess. This is especially true if there is little or no variation in the religious denomination of households in the two community types. Deceptive associations between livelihood and fertility variables in the two types of community will result.33 These limitations should be borne in mind in perusing the analyses below in respect of these two countries.
Despite the important role of fisherfolk women in the processing and marketing of fish, fish capture at sea is exclusively done by men. Not surprisingly, focus group discussions in villages and the household survey data (see tables 35 and 37) almost invariably showed that fisherfolk, both women and men, had a strong son preference, particularly in Senegal. Tanzanian households in fishing communities were similar to farming households, in son preference as well. In fishing communities many fisherfolk households depend for their sustenance on a mixture of farming, animal husbandry and operation of fishing boats.
33 A deceptive or spurious association occurs in this case because fertility (the dependent variable) is affected not so much by the hypothesized independent variable (N1=type of livelihood, i.e. fishing or farming) as by another independent variable (N2=religion) that is strongly associated with N1.
In the past, the association of son preference with higher fertility levels has been the subject of research. However, the findings are not conclusive (Chowdhury & Bairagi, 1990; McClelland, 1979; Repetto, 1972; Mason & Taj, 1987). Moreover, there is controversy about whether men or women show a greater demand for children, particularly for male children, and there are conflicting theories regarding men's and women's reproductive goals (Cain, Khanam & Nahar, 1979; Caldwell, 1982). The reasons for son preference are most frequently the low status of women (Cain, 1984), patriarchal structures (Cain, Khanam & Nahar, 1979), differences between fathers and mothers regarding claims to the major share of children's love, loyalty and labour (Caldwell, 1982) and prevention of the financial risk of divorce, widowhood and old age dependency (Cain, Khanam & Nahar, 1979). However, research and debate rarely indicate the effect of the nature of the livelihood on gender preference of children. Although in most rural households a clear division of labour exists between genders, few farming activities exclude involvement of women to the extent that women are excluded from capture fisheries at sea. This may explain why, on average, preference for sons among coastal marine fisherfolk is somewhat higher than in farming households.
Assessment of focus group discussions indicates that son preference is more pronounced among older people (40–49), particularly among men. Focus groups also mentioned other reasons for son preference: sons ensure continuation of the family and lineage name, and sons are important for the financial support of elderly parents, particularly if widowed. More than in farming households, fisherfolk try to achieve a desired number of sons. However, they are faced with the biological fact that the probability of having a son is about equal to having a daughter. This may explain why, across countries, fisherfolk births are somewhat more numerous, on average, than farming births. The differences are most pronounced in comparing the lifetime fertility experience of the older women in the two communities; fertility differentials are less pronounced in the younger women (15–24).
In all the communities, the younger participants in the focus group discussions mentioned desiring a smaller number of children than their parents. Moreover, most of them preferred other types of work to fishing and farming, a view supported by parents in countries where fishing is no longer perceived to be a profitable undertaking. The traditional gender division of labour in fisheries may change by the adoption of new fishing methods and new forms of organization. As a result, livelihood-related fertility differentials and child gender preferences may decrease.
Fertility and child mortality differences between fishing and farming households are addressed in Table 38. Since the age distributions of female and male respondents differed by type of community (see Tables 35 and 37), a direct comparison of numbers of children ever born (CEB) and children surviving (CS) would be biased by the age factor - the “older” community would tend to have had more children, regardless of its economic set-up. In order to compare households in the two types of community, therefore, it was necessary to standardize the average values of CEB and CS by using as covariate the age distribution of all married female respondents. The results indicate a consistent pattern of higher lifetime fertility levels in fishing communities. In all countries except Bangladesh, fisherfolk families had more births than farming families.
Table 38 - Standardized average numbers of children ever born and children surviving and average percentage of deceased children among currently married women (aged 15–49) in fisherfolk and farming households
|Own children ever born alive (CEB)||3.9||3.8||3.8||3.6||3.3*||2.6*||4.5||4.7||2.7||2.3||4.3||3.9|
|Own children surviving (CS)||3.7||3.6||3.2||3.2||3.1*||2.5*||3.5||3.7||2.5||2.2||4.0||3.7|
|Percent deceased children||5.1||5.3||15.8||11.1||6.1||3.8||22.2||21.3||7.4||4.3||7.0||5.1|
|Number of women||(259)||(286)||(49)||(47)||(104)||(63)||(94)||(90)||(92)||(87)||(57)||(42)|
* The difference between results for fishing and farming households is significant at the 5 percent level.
In addition to observed fertility levels, it is interesting to note current fertility perceptions in the two types of community. As can be seen in tables 35 and 37, in all countries except Bangladesh, fisherfolk, and particularly fishermen, consistently reported higher ideal numbers of children than farming communities. The ideal number of children affects, among others, whether and how contraceptive methods are used. However, the relationship between livelihood characteristics and contraceptive use is more complex because of the existence of various intervening and underlying factors, such as education and income levels, livelihood related demands for (more) children, infant and childhood mortality. Contrary to what one may expect, across countries observed higher fertility levels in fishing communities were associated with higher ever- and current use of modern contraceptives. A tentative explanation for this anomaly could be that use-effectiveness is lower and rates of discontinuation higher in fishing communities, but there is need for further research. If true, this would call for new or improved reproductive health services in fishing communities.
Overall, ever- and current use of contraception was low. The high illiteracy observed in the communities in four of the six countries could be one significant explanatory factor (Johnson-Acsadi and Weinberger, 1980; Phillips, 1980). Generally speaking, contraceptive use (and effectiveness) rates will be low when the number of surviving children is less than the stated ideal number. This was found in three of the six countries: Senegal, Tanzania and Malaysia. The high illiteracy levels shown in tables 35 and 37 are associated with high numbers of unwanted and mistimed children in these countries. In the other countries, India, Bangladesh and the Philippines, married women had more surviving children than their ideal number, and, not surprisingly, stated more often that the last child born was unwanted or mistimed.
The reported child mortality levels, expressed in terms of the average percentage of dead children among children ever born, was equal or slightly higher in fishing communities than in farming communities. This finding conflicts with assertions made elsewhere that living and health conditions in fishing communities were generally worse than in farming communities. If this were indeed true, we should have found more pronounced differences in child mortality levels in the two types of community.
In spite of somewhat higher rates of child mortality in fishing communities, table 38 shows that in all the countries except Tanzania and Bangladesh, the average number of surviving children in fishing families was still higher than in farming communities.
Adult mortality: a comparative perspective
The survey collected limited information on adult mortality in the two communities. Respondents were asked whether their father and mother were still alive and, if not, their age at death. The results are presented in Table 39. Care is required in interpreting the results because of the following: (a) To determine and compare adult mortality levels, age and birth-cohort effects must be taken into account; the required information, e.g. date of death, to do this has not been collected. (b) The debatable assumption must be made that, on average, the age at death of adult men and women who were not parents was the same as the age at death of adults who were parents. (c) The lower the age of the respondent reporting that a parent had died, the lower the expected age at death of the parent. Consequently, if the age distribution of respondents in the two communities differed, average ages at death of parent(s) could not reflect genuine differences in adult mortality levels.
In Table 39, reported ages at death of parents have been standardized, using as covariate the ages of respondents that report that one or both parents have died. Standardized ages take out the effect of differences in ages of respondents in the two community types. These indicators were used to compare countries. In all countries, except the Philippines, adult men in fishing communities died on average at a younger age than men in farming communities. This could be related to the higher occupational health risks involved in coastal fisheries compared to coastal farming. In all countries, except the Philippines and India, adult women seemed to die on average at a younger age than women in farming communities.
Table 39 - Observed and standardized average age at death of parent(s) as reported by respondents mentioning that one or both parents had died
|Country||Type of community||Father's age at death (FAD)||Standardized FAD||Mother's age at death (MAD)||Standardized MAD||Number of respondents (N)|
* The difference between fishing and farming households is significant at the 5 percent level.
It is of interest to note that across countries and communities, adult women died at a younger age than adult men, sometimes substantially so. This is contrary to what one would generally find in most countries, because national and sub-national mortality level indicators such as the life expectancy at birth of women state higher values for women than for men. Although the concept of average age at death is different than the concept of life expectancy at birth, the two indicators are often highly correlated. The hypothesis, yet to be further investigated, could be that the socioeconomic and health conditions, particularly for women, are much worse in the studied communities than the average conditions prevailing in the countries and rural areas as a whole.
Migration and isolation: a comparative perspective
At the beginning of chapter 1, attention has been drawn to the fact that in the literature coastal fisherfolk have often been described as living in settlements which are more isolated than those of other rural communities. Given the mostly “open-access” nature of coastal fisheries, fisherfolk may have more reason than farmers not to welcome potential competitors in the catching, handling and trading of fish. In four of the six countries, eligible respondents were asked whether they had been born and raised in the village. This enables identifying and counting the lifetime immigrants and therefore provides some indication of the degree of openness in the communities.
Table 40 shows that the fishing communities did not seem more closed than the nearby farming communities. In three of the four countries, the percentages of lifetime male immigrants in fishing communities were slightly higher than in farming communities. In Tanzania and the Philippines, about one fourth of the male respondents in fisherfolk households were lifetime immigrants. Much smaller percentages of immigrants were observed in the Indian and Bangladesh fishing communities, perhaps because coastal artisanal fisheries there are almost exclusively the domain of the lower caste Hindus. In Bangladesh, because the non-fishing communities are mostly Muslim, fishing communities may not attract migrants. Overall, the female percentages of lifetime immigrants were much higher than those of the male, probably because women traditionally move to their husband's home after marriage. Finally, in all four countries, the differences between the degrees of isolation, or openness to migration, experienced by the two types of community, were smaller than generally thought.
Table 40 - Percentage of lifetime immigrants in fishing and farming communities among male (aged 15–64) and female (aged 15–49) respondents at the time of the survey
In the fishing communities studied, a number of households had different household income sources from non-fisheries industries, such as farming, livestock and service sector work. Across countries, variation is observed in socioeconomic and demographic conditions of households in fishing communities. This reflects expected variations in levels of development and demographic conditions. In all the countries, except Senegal and the Philippines, fisherfolk households were worse off regarding perceived living conditions and ownership of production and household assets than farming households. Moreover, excluding Senegal, fisherfolk households in all countries were somewhat larger than farming households. Reports from focus group discussions suggest that fisherfolk more often lived in an extended or joint family structure because preference, for artisanal fisheries-related work, was given to kin-based groups.
The surveys showed that fertility and mortality in fishing communities are not very different from those in farming communities. The indicators are higher in the former type of communities, but the differences are small. In addition, higher fertility is compensated by higher childhood and adult mortality, so that the resulting natural population growth rates are quite similar. In the context of accepted theories of the economic determinants of fertility, this indicates that the economic value of children and family labour in general is similar in both types of household. In other words, there does not seem to be a specificity of fishing households with regard to the intensity of need for family labour, be it for fishing as such or for income diversification and security.
The patterns of family labour utilization observed in both the fisherfolk and farming households studied involved some members taking on other activities than fishing (be it in main or secondary fashion). The typical reasons for this are that other income sources are more rewarding, or more stable, and that diversifying income sources is an insurance against income fluctuations and/or failure. The surveys provide some supporting evidence for this: fishermen and their sons often mentioned that they had sought work outside the fisheries and/or farming, particularly in areas where the fisheries is no longer perceived to be a profitable undertaking (Groenewold, 1998).
Desk studies have shown that the rate of population growth among fisherfolk generally was high, and fluctuated widely, in the areas covered by the Project and during the period concerned.34 When one examines the factors of growth through the surveys, however, it emerges that the differences, and changes over time, in fertility, mortality and natural growth (i.e. the excess of births over deaths) are not consistent with the observed differences and changes in the overall pace of fisherfolk growth. In fact, population growth rates were higher, or accelerated, where and when labour transfers to the fisheries sub-sector occurred because fisheries was perceived as a profitable undertaking and accessible to newcomers. In other words, changes in livelihoods, motivated by the need to alleviate poverty, were responsible for rapid growth rates of fisherfolk. Conversely, fisherfolk populations tended to stagnate in areas where and when the fisheries was not considered profitable, especially if additional or alternative occupations were not available locally.
34 Given the age distribution of women surveyed, the fertility and childhood mortality characteristics reported here should be taken to reflect mostly the experience of the period from 1980 to 1996.
Looking ahead, such observations suggest that numbers of fisherfolk should be expected to increase or decline primarily depending on whether fisheries are profitable or not and fisherfolk have access or not to alternative livelihoods for their sustenance.35 As for the natural growth factor (fertility and mortality), what could be the effect of fertility control and reproductive health programmes on population growth in fisherfolk populations? One should be aware that the said effect might be limited where there exist strong economic incentives to move into the fishing sector, at least where migration and/or intersectoral mobility are easy. Obviously, this detracts nothing from the fact that those programmes are entirely justified on the basis of individual and family health and human rights. In this respect the poorer communities require attention in order to catch up with the rest.
From the demographic viewpoint, migrant flows induced by changes in livelihoods certainly accelerate population growth in the areas of destination. The survey was not geared to measure migration flows and therefore does not allow an in-depth analysis of their role in the rapid growth of the fisherfolk populations studied. But that role is widely acknowledged locally in most cases. In addition, a change of residence is not a necessary condition for becoming involved in coastal fisheries: rural dwellers living in coastal communities may find work in artisanal fisheries to supplement their income without changing their residence.
Rural households, including fisherfolk, must not be perceived as stagnant entities, but as dynamic decision-making units. Some of the households studied demonstrated the capacity to successfully implement livelihood diversification strategies to cope with income fluctuations and poverty. However, it is understood that there may not exist local alternatives to fishing. In such situations, fisherfolk are forced to continue to work in fisheries, or to migrate to urban areas. The lack of access to alternative income sources for fisherfolk is of major concern to policy makers. It contributes to the exploitation of marine natural resources above the level that would occur if alternative livelihoods and income sources were available to fisherfolk. Moreover, the fact that some fisherfolk remain in the sector against their will negatively affects the incomes of all fishers.
Overall, at subsequent points in time, change in rural livelihood affected the number of full/part-time fishermen and their households. Therefore, in addition to demographic factors, the observed national and sub-national growth rates will be affected by the changes in the main and secondary livelihoods of rural households that responded to perceived profitability of existing and new livelihood opportunities, including fisheries. Households previously characterized as farming households may become fisherfolk households and vice versa.
35 In India, alternative sources of household income were present in the form of wage labour in the industry and services of nearby urban areas. The numbers of fisherfolk have stagnated and are declining. In the Philippines, such alternatives are absent near the communities that were investigated, and the number of fishers will continue to grow in spite of declining catches per unit of effort and marginal costs that exceed marginal financial benefits.