PART II CASE-STUDY: GUINEA-BISSAU
7. Village-level case-study I: Caiomete
7.7 Migration and its effects
Migration is a very widespread strategy amongst the population of Caiomete. This is not surprising, given that the village was selected largely on the basis of this a priori criterion.
Some telling indicators of the current significance of the phenomenon are presented in Table 7.11: a full 76% of the interviewed households reported the existence of a former member who is currently living elsewhere (abroad in 53% of the cases and in Europe in 29%); and in as many as 72% of the households, at least one of the current members of the households lived outside of the region for more than twelve months in the past (abroad in 53% of the cases and in Europe in 6%).
218 Frequency
(Yes)
Valid percent Households with current
migrants away
55 76.4%
Households with current migrants abroad
38 52.8%
Households with current migrants in Europe
21 29.2%
Households with economically active (or student-worker) migrants currently way
39 59.1%
Households comprising return migrants
52 72.2%
Households comprising return migrants from abroad
38 52.8%
Households comprising return migrants from Europe
4 5.6%
Table 7.11 Caiomete: Household participation in migration - basic indicators
Because this process dates back to many decades ago, including in its intercontinental variant, it is hard to find precise information with regard to its early characteristics. However, my qualitative sources concurred that early international migrants from Caiomete followed the Guinea-Bissau–Senegal–France route and that most of those napat ubabu151 migrated by ship. In the present, migration has a diverse character:
intra-regional migration to Canchungo for the purpose of pursuing education or work, inter- regional migration to Bissau to take advantage of the opportunities available there, cross- border migration to Senegal (mostly by women, to work in cleaning/laundering) and inter- continental migration to Portugal and France (mostly by men, to work in construction) constitute the main layers in the constellation of contemporary migration flows from Caiomete.
All of the above is reflected in Figures 7.13 and 7.14: the 167 migrants reported by the 72 respondents are divided almost evenly between those who are living elsewhere in Guinea-Bissau and those who currently reside outside the country. Amongst the former, Bissau stands as the main destination by far, with almost all the others living somewhere in
151 This is the general name given to the pioneers of intercontinental migration. According to Mr.
Mendes, the expression literally means “those who row to the white man’s land”.
219 the Cacheu region (typically in Canchungo). Among the latter, Senegal constitutes the single most common international destination (22%), followed by Portugal (18%) and France (8%).
Figure 7.13 Caiomete: Breakdown of migrants by country of current residence
Figure 7.14 Caiomete: Breakdown of current migrants residing elsewhere in Guinea-Bissau by region
With regard to their demographic characteristics, we find that the migrants from Caiomete are quite heavily concentrated in the age group that typically corresponds to economic activity, although this information must be taken with caution due to the extremely high percentage of migrants whose ages the respondents were unable to report (almost 50%). Amongst those whose ages were reported, however, none are over the age of 59 and less than 10% are under 15. The mean age is 27, and 50% fall in the [20-35] age- group (Figure 7.15). This is typical of migration contexts in which the flows continue to rejuvenate themselves, as indeed is also suggested by the distribution of the migrants’ year of departure from the village: 75% of the migrants left eight or less years ago (Figure 7.16).
220 It may also be considered indicative of a migration context characterised by a high level of return migration, in so far as part of the reason for the lack of older migrants in the sample lies in their subsequently having returned to the village. Another partial explanation may lie in the design of the survey, given that migrants were identified in relation to existing households and that the households of origin of some of the older migrants are more likely to no longer exist (hence those migrants will fail to appear in the sample). However, taking into account the high share of return migrants among the current resident population of Caiomete (Table 7.11), the previous interpretations of the age characteristics of the migrant population seem quite plausible.
Figure 7.15 Caiomete: Current age of the migrants as reported by the respondents, boxplot
221 Figure 7.16 Caiomete: Current migrants’ year of departure from the village, boxplot
In terms of its gender composition, the migrant population from Caiomete is almost perfectly balanced: 84 female and 83 male migrants were reported by the 72 respondents.
However, this overall balance conceals significant gender differences in terms of migration patterns, as already suggested above. Thus, while male migrants are significantly over- represented amongst those who migrated to Europe (Portugal and France), women migrants are considerably more numerous amongst migrants to Senegal (Table 7.12).
Arguably, this is accounted for by both the greater ability on the part of men to mobilise the resources required to undertake transcontinental migration and by the characteristics of the migrants’ labour market incorporation in these two types of destinations – most (male) migrants who go to Europe work in the construction sector, whereas most (female) ones who go to Senegal work in cleaning, laundering, weaving and embroidery.
´
Current place of residence
Total Guinea-Bissau Senegal Portugal France Spain
Gender Male 45 10 17 10 1 83
Female 41 26 13 3 1 84
Total 86 36 30 13 2 167
Table 7.12 Caiomete: Cross-tabulation of the migrants’ gender with their current place of residence, absolute frequencies occurring in the sample
222 It is important to note that a significant share of out-migration from Caiomete is not work-related. In particular, we find that as many as 34% of the migrants moved elsewhere with a view to pursuing their studies. These, too, follow typical geographical patterns, being overwhelmingly found in either Canchungo or Bissau. It is amongst the economically active, however, that the geographical-occupational patterns are most remarkable (Table 7.13).
Out of the 70 migrants reported as being either economically active or student-workers, those living in Senegal mostly work as cleaners/launderers, those living in Europe overwhelmingly work in the construction sector (or are currently without a job) and those who live elsewhere in Guinea-Bissau are much more evenly spread across the various occupational categories.
Current Place of Residence
Total Guinea-
Bissau Senegal Portugal France
Occupation Unemployed 2 1 4 0 7
Construction worker 9 3 8 3 23
Cleaning/laundering 6 12 1 0 19
Commerce 4 2 0 0 6
Weaving/Embroidery 1 3 0 0 4
Primary sector worker 2 1 0 0 3
Other (skilled) 6 0 0 0 6
Other (unskilled) 1 0 1 0 2
Total 31 22 14 3 70
Table 7.13 Caiomete: Cross-tabulation of the economically active (and student-worker) migrants’
current place of residence with their occupations, absolute frequencies occurring in the sample
So far, evidence has been provided as to the significant magnitude of past and present migration from Caiomete. We have found that the ratio of reported current migrants to the resident population is around 47%, which can be grossly interpreted as a full one third of the village being away. Information has also been provided on the characteristics of the migrants and the logics that drive the migration processes. But what have the socioeconomic effects of migration been in Caiomete?
With regard to this, it is useful to begin by putting forth some of the qualitative evidence. Most respondents to the survey, as well as the participants in the focus group
223 and semi-structured interviews, concur that migration has constituted an important livelihood strategy for the migrants themselves, complemented the income of their households of origin and families through remittances, and helped the village as a whole in several ways. In particular, the generalised substitution of zinc for thatch roofs in individual housing units is widely associated with remittances and the migrants’ return with accumulated savings. At the level of the village as a whole, the migrants from Caiomete, through their hometown associations in France and Portugal, have collectively played a key role in financing three main initiatives: the building of a new school building, which currently complements the older one (which could no longer accommodate all the students); maintenance work on the Caió-Caiomete road; and the building of a local basic healthcare centre. In each of these cases, the Caiomete migrants’ HTAs paid for the construction materials, the tools, and food and wine for the workers, whereas Caiomete residents provided the labour.
However, arguments were in some cases put forth in support of the view that migration has had negative effects, too. One of those arguments regards the fact that the reduction in the number of young men has made it more difficult to organise large-scale collective initiatives like the building and repairing of the main dykes that separate the rice paddies from the rising sea water. Given the high share of respondents who reported having lost rice paddies due to salinisation, as well as the fact that those initiatives used to mobilise the community as a whole and have not been undertaken for at least ten years, there is certainly plausibility to this argument. Another argument concerns the effect of migration upon community cohesion: some people view migrants as privileged vehicles for the introduction of new consumption patterns and cultural habits, which they feel have been gradually undermining tradition, equality and cohesion (see Gable 1995, 2000 and 2006 for an especially pessimistic account of the effect of emigration in another nearby Manjaco village).
In spite of the above, 60% of the respondents to the survey consider that, overall, migration has had a positive impact upon the village as a whole, compared to 18% who feel that the impact has been a mostly negative one (13% regard it as neither positive nor negative and a further 9% see it as both). Interestingly, when we turn to the respondents’
assessment of the impact of migration upon their own household, we find a less optimistic outlook: the share of those who are neutral about migration increases (as might be expected, given that some households never participated directly in migration), but the ratio of positive to negative assessments decreases from 3.3/1 to 2.3/1. Qualitative
224 information provided by the respondents on a few occasions helps to interpret this difference: seemingly, there is a significant amount of resentment on the part of some migrants’ relatives as to what they perceive as “neglect” on the part of the migrants, who not only are no longer there to contribute to the household’s agricultural production but also fail to send (enough) money.
Indeed, the quantitative information with respect to the migrants’ actual practices seems to corroborate this somewhat less optimistic view of the consequences of migration for the households who stayed behind. The last visit to the village by most migrants took place in either 2010 or 2009 (Figure 7.17), which suggests that, at least in this sense, they maintain strong linkages with their household and community of origin152. However, the extent to which they contribute to the domestic economy of their households of origin is surprisingly low.
Figure 7.17 Caiomete: Year of last visit to the village by current migrants as of May 2010, boxplot
Indeed, we find that only 17 out of the 56 households that currently participate in migration report having been sent any money remittances during the course of the previous twelve months153. Additionally, only as few as 21 out of 155 migrants over the age
152 Naturally, those living in Senegal or elsewhere in Guinea-Bissau typically visit the village more frequently than those who have moved to Europe.
153 In two other cases, the respondents either did not know or did not answer.
225 of 15 were reported to have sent any money in the previous year154. In relative terms, migrants in Portugal and France were the ones most likely to remit, whereas that practice was reported as most infrequent among migrants currently living in Senegal and elsewhere in Guinea-Bissau (Table 7.14). Additionally, remittance-sending migrants were four times more likely to be men than women155.
Sent money last year Y/N
Total Yes No
Not Applicable
(Under 15) DNA/DNK Current
place of residence
Guinea-Bissau 7 70 6 3 83
Senegal 2 31 3 0 36
Portugal 7 23 0 0 30
France 4 9 0 0 13
Spain 1 1 0 0 2
Total 21 134 9 3 164
Table 7.14 Caiomete: Cross-tabulation of the migrants’ current place of residence with their remittance behaviour in the previous 12 months, absolute frequencies occurring in the sample
Now, as mentioned in chapter 5 (above), there are reasons to believe that the respondents may have under-reported this practice to some extent. However, the fact that the failure to send money back to the households of origin exhibits such a strong association with gender (women) and locations (Guinea-Bissau and Senegal) indicates that there is likely to exist relevant socioeconomic explanations for the low prevalence of remittance practices beyond any possible under-reporting156. Clearly, female migrants (and those in Guinea-Bissau and Senegal) are much less able to send money than male ones (and those in Europe) by virtue of their socioprofessional status (and there may also be less of an expectation that they do so).
154 DNK/DNA = 3.
155 An additional piece of information of some interest consists of the fact that the 21 migrants reported to have remitted money in the previous 12 months were divided evenly in terms of the channel that they used to transfer the money: 1/3 brought the money in-hand upon a visit to the village, 1/3 used a money-transfer operator (Western Union or Money Gram) and 1/3 sent the money through friends or relatives.
156 What the migrants from Caiomete did in larger numbers than remit money was to send or bring other material help – clothes, food, cell phones, etc. –, a practice reported for the previous 12 months in the case of 21 out of 70 migrants currently living in Guinea-Bissau and 14 out of 33 living in Senegal.
226 At any rate, it is clear that the prevalence of remittance-sending practices is very low in the context of this village, and it therefore comes as no surprise that migration is not statistically associated with greater long-run wealth as reflected in the asset index score. If we refer back to Table 7.9, we find that current participation in migration is associated with a lower asset index score, whereas past participation in migration does not have any significant differentiating effect. This appears to indicate either that migration has negative self-selection characteristics (i.e. migrants hail from relatively poorer households) or that it causes a relative deterioration in the economic situation of the households of origin, presumably through the reduction in those households’ pool of labour. The answers to the survey do not make it possible to disentangle these two effects, but qualitative information collected from various sources seems to point to the latter effect having greater prominence. It remains true that those migrant households that report receiving remittances do seem somewhat better-off than the average in terms of long-run wealth (Table 7.9); however, as we have seen, there are few of these.
Moreover, when we look at the reported use of remittances on the part of those households that do receive them (an information that must be taken with caution given the fungibility of remittances vis-à-vis other sources of monetary income), we find that purchasing foodstuffs is the most frequent answer (12 out of 21), followed by religious ceremonies (11/21) and health-related expenses (7/21). Only one respondent reported hiring someone using remittance money (to do repair works in the house) and never was the money reportedly used to buy means of production, or otherwise valorise commodities or intensify production. This shows that even for those households that receive them, remittances may play an important role as a complementary source of money income, but are also typically “unproductive”, insofar as they are used for consumption, rather than invested or converted into capital. Of course, it remains possible that the additional demand brought on by remittances, to the extent that it does not ‘escape’ from the village, contributes to accumulation by households other than the immediate recipients of remittances. However, what this particular data seems to preclude is the likelihood of endogenous upward class differentiation directly based on participation in migration and the reception of remittances.
Indeed, when we examine the likelihood of a given household having hired in agricultural labour in the previous twelve months, we find that it is nearly identical whether or not that household was sent any remittance money in the previous year (Table 7.15). As
227 regards the hiring-in of non-agricultural labour, remittance-recipient households are in fact even less likely to have done so in the previous twelve months.
Paid Agricultural Hiring-in
Total
No Yes
Remittance-recipient household Y/N
No 17 36 53
Yes 5 12 17
DNA/DNK 0 2 2
Total 22 50 72
Table 7.15 Caiomete: Cross-tabulation of the households’ remittance-recipient status with their recourse to paid agricultural labour in the previous twelve months, absolute frequencies occurring
in the sample
So much for upward class differentiation driven by remittances. Conversely, however, could remittances somehow act to alleviate the pressure upon local residents to engage in wage labour, such that they might constitute a force (however small) constraining the local labour supply and, as a consequence, the generalisation of capitalist social-productive relations and capitalist accumulation? Here, the evidence is mixed. In order to compare the recourse to the hiring-out of wage labour by remittance-recipient households with that by the remaining households, two new variables were constructed by computing the share of adults (over 15) in each household who engaged in the hiring-out of labour in the previous twelve months, respectively within the village and to employers outside (Table 7.16). When we compare the means of these two variables, we find that the likelihood of an adult individual who belongs to a remittance-recipient household engaging in the hiring-out of wage labour within the village (0.41) is almost identical to that of an adult individual in a non-recipient household (0.42) (Table 7.16). Only as regards the hiring- out of labour outside the village is there a significant difference (0.07 compared to 0.20).
This may be interpreted as suggesting that migration and remittances may indeed have an impact of the sort described above – reducing the labour supply by acting as a ‘substitute’
for the local hiring-out of labour –, but also that that effect occurs at the regional (rather than village) level and on a limited scale.
228 Remittance-recipient
HH Y/N
Hire-out village HH
share
Hire-out outside HH
share
No Mean 0.427 0.204
N 52 52
Yes Mean 0.417 0.068
N 17 17
p-value of significance of differences (T-test, 2-tailed)
0.929 0.01*
Table 7.16 Caiomete: Comparison of the mean recourse to the hiring-out of wage labour in the previous twelve months by remittance-recipient and non-recipient households
Still, comparing the recourse to the hiring-in and hiring-out of wage labour by remittance-recipient and non-recipient households only provides a partial picture of the effect of migration upon the labour market. Not only is it possible that some households may be under-reporting current remittances, it may also be the case that these or other households could have benefited from remittances (or return migration with accumulated savings) in the past and used them to engender a dynamic of upward class differentiation, or prevent one of downward differentiation. It is therefore useful to examine whether households with current and return migrants, regardless of their current remittance- recipient status, exhibit significant differences with respect to their participation in the labour market when compared to the other households.
Table 7.17 provides information in this respect. With regard to the hiring-in of labour, we find that current participation in migration reduces the likelihood of a household hiring both agricultural and non-agricultural labourers – the only exception being the hiring- in of agricultural labour by households with migrants in Europe (in which case the likelihood is more or less identical to that of non-migrant households). Although the absolute figures in some of the cells of these cross-tabulations are small (hence the risk of stochastic error influencing the conclusions is greater), the fact that participation in migration consistently fails to be associated with greater recourse to the hiring-in of wage labour enables us to conclude that the endogenous emergence in this context of a proto-class of ‘yeoman farmers’ driven by migration seems unlikely to say the least. Indeed, this corroborates the conclusions previously reached based on the remittance data.