Enabling poor rural people
to overcome poverty
In the previous sections, HFS was conceptualized in terms of a set of variables deemed likely to contribute to a household’s vulnerability to entitlement failure. This section uses the data to analyse each of these variables and to contrast the observed values between beneficiary and non-beneficiary households.7
Food production, food market dependency and crop diversification
Corn is the main staple in the diet of Guatemalan rural households. With the exception of Buena Vista and San Martín, where high-altitude conditions are ill suited to the crop, almost all households grow corn. Even in Suchiquer, where the harsh condition of the terrain makes corn production unrewarding, practically all households allocate the majority of their land to corn production, intercropped with sorghum (in this community, sorghum has become a forced substitute for corn in the household diet, since it appears to perform better in dry weather and poor soil conditions). That the majority of rural Guatemalan households grow corn in spite of adverse conditions and low returns appears to reflect the premium that poor households are willing to pay in order to secure a minimum level of food adequacy and to insure themselves against riskier portfolio choices.
The figures in Tables 1 and 2 in Appendix 1 highlight the differences in crop production and productivity observed across the surveyed communities. These differences reflect profound climatological and morphological variations and give rise to idiosyncratic risk of food entitlement failure. As expected, Buena Vista and San Martín have the lowest production levels, with an average production of only 80 lbs of corn per capita8 (Table 2). At the opposite end of the spectrum we find households in San Nicolas each producing an average of 790 lbs per capita. All remaining communities produce a per capita average of 260 lbs. An analysis of corn yields across communities reveals a distressing scenario for some (Table 1). Suchiquer, for example, reports an average corn yield of only 80 lbs per cuerda. Even the figures for Tituque (150 lbs/cuerda) leave ample room for improvement.
The argument that food self-sufficiency at the household level is a main determinant of HFS has been advocated by many, on the grounds of the endemic nature of the imperfections and uncertainty characterizing food markets in many LDCs. In the presence of market failures due to idiosyncratic transaction costs, rural households in LDCs tend to rely more heavily on own-produced food than can be reasonably explained by efficiency and comparative advantage criteria. Table 2 presents selected indicators measuring household dependency on food markets. Specifically, figures are reported for net corn purchases (equal to the difference between corn purchases and sales), for the corn market dependency ratio (equal to the ratio between corn purchases and total household corn consumption) and for the average duration in months of the previous corn harvest. All three indicators tell similar stories, with the figures for Buena Vista and San Martín indicating the highest dependency, and those for San Nicolas, the lowest. Among the intermediate communities, Suchiquer is characterized by a higher degree of food market dependency than either Tituque or the NTX communities in the Cuchumatanes.
However, self-sufficiency in the production of staples does not imply food security, particularly when the risk of output shocks is very high. In order to lessen this risk in either own-supplied or purchased food supplies, crop diversification emerges as a viable option in a farmer’s risk management strategy against food entitlement failure. Table 3 summarizes information on agricultural diversification, at both the household and the community level. The results are disturbing. With the exception of project beneficiaries in Tituque and the NTX communities, the other communities show little diversification at the household level, with many households growing only two crops.
This pattern also appears in aggregated community-level data. Not only do households generally grow few crops, but they also tend to grow the same crops within the same community. Suchiquer shows a particularly low level of crop diversification, which is likely to have a negative impact on both household diet and HFS. On the positive side, it seems that the projects may have been partially successful in promoting agricultural diversification among their beneficiaries, although the effect appears to have been limited to only a few households and to selected communities.
As mentioned above, food market dependency alone does not necessarily imply higher food insecurity. If we limited our analysis to food supply, it would look as if Buena Vista and San Martín were the most vulnerable communities from a food security standpoint, San Nicolas the least vulnerable and Suchiquer somewhere in between. However, this partial analysis may be grossly misleading in characterizing HFS. In the search for a more comprehensive description of food security, we continue in the next sections by evaluating some additional factors presumed to contribute to a household’s food entitlement.
Income level and composition
The documented shift in the HFS paradigm from food supply to food access has led many policy makers and development practitioners concerned with food and nutrition issues to emphasize the role of income as a determinant of food security. However, as noted above, their efforts have often focused solely on the level of household income, while frequently ignoring equally important factors such as income composition and control. Although two households can generate equal amounts of monetary income, such income can differ from a food security perspective, either because of differences in the number of sources or in their riskiness. This section considers the role of income source and income diversification.
Table 4 displays total household income by source. A household can generate income from the sale of own-produced agricultural surplus, off-farm labour, petty trading, animal sales and transfers. The data show striking differences both in the amount of total income and in its composition, across communities and between beneficiaries and non-beneficiaries.
Households in San Martín and Buena Vista earn, on average, the highest incomes (Q 8 298/year),9 with beneficiaries averaging 18% more than non-beneficiaries. The disaggregation by source reveals a heavy reliance among households in these communities on on-farm income (mainly from the sale of own-produced potatoes), especially for non-beneficiaries. About 90% of households in these communities are reported to have earned some income from on-farm labour, with non-beneficiaries earning on average Q 5 362 (representing a striking 73% of their total income) compared with less than Q 4 900 for beneficiaries (56% of total income). Many beneficiaries in Buena Vista and San Martín also derive income from off-farm labour, petty trading and animal sales.
These figures are in sharp contrast with those for Suchiquer, where household earnings average only Q 1 801 per year, and non-beneficiaries earn approximately 40% more than beneficiaries as a result of handicraft sales. For households in this community, the only two sources of income appear to be off-farm labour (mainly by men and in the form of temporary agricultural wage work in coffee plantations in nearby Honduras) and handicrafts, such as straw mats and baskets, manufactured mainly by women and sold by men (!) in the local market in Jocotan.10 Practically no households in this community earn any income from on-farm labour.
In the NTX communities, the relatively close total income figures for beneficiaries (Q 7 253/year) and non-beneficiaries (Q 6 649/year) conceal major differences in income composition. Beneficiaries in the NTX communities derive over 60% of their income from the sale of own-produced agricultural products, compared with only 30% for non-beneficiaries who rely much more heavily on off-farm income (61% of total income versus only 25% for beneficiaries).
Beneficiaries in Tituque earn on average 80% more than non-beneficiaries (Q 5 139 versus Q 2 820). The difference is a result primarily of the higher on-farm income generated by the former through coffee production. Surprisingly, in spite of the apparent dynamism shown by female beneficiaries in the manufacture of handicrafts, beneficiary households earn less than non-beneficiaries from this source (Table 5).
Non-beneficiaries in San Nicolas earn on average 20% more than beneficiaries (Q 6 090 and Q 5 000, respectively). In terms of income composition, no major differences are detected between the two groups, except for the higher income that non-beneficiaries earn from off-farm labour, which accounts for the observed difference in total income. Most of the income from petty trading in this community comes from the sale of handicrafts.
Excessive reliance on a limited number of sources may be hazardous, from a food-security standpoint, if these few options also display high risks. This may be the case of beneficiary households in NTX communities. On average, these households obtain 61% of their income from notoriously risky activities, such as the production of non-traditional crops. Past experience with NTX crops in Guatemala and in other parts of the world have shown repeatedly that these crops have proved to be deceptive for poor farmers, who lack the resources to face the increasing problems associated with NTX production. Furthermore, in the presence of limited marketing channels for the products, excessive reliance on NTX income often translates into extreme subordination by poor peasants to powerful agro-exporters. As vehemently voiced to the author by many farmers during the fieldwork, they consider the treatment they continue to receive from agro-exporters as highly unfair. This situation has persisted in spite of continual changes over the years in the makeup of the agro-export firms with whom the farmers contract. According to NTX farmers in Chochal and Torlón, agro-exporters often reject their products because of alleged low quality, or else impose excessive discounts for these perceived defects. The agro-exporters are also blamed for charging excessively for the inputs they advance to the farmers, paying late, or, in some cases, not paying at all. In these circumstances, the greater risks and exposure to unequal bargaining may put NTX households in jeopardy from a foodsecurity standpoint, despite sizeable increases in income.11
Table 3 summarizes the income composition scenario and household exposure to the risk of entitlement failure from income shocks, showing a set of indicators reflecting household income diversification strategies. One important result is the high number of households in both Suchiquer and Tituque that derive their income from only one source (33% and 34%, respectively). This finding, combined with the low income levels reported for these communities, points to the extreme vulnerability of many households to entitlement failure in both Suchiquer and Tituque.
Asset ownership
In circumstances in which financial markets are highly imperfect or non-existent, asset ownership represents a common form of savings and investment among households in rural communities. For food security purposes, asset ownership must be further characterized on the basis of liquidity. Liquid assets, such as small animals, are often used by poor rural households as a form of insurance against the risk of food entitlement failure. As has been said, liquid asset divestment, if made to cope with temporary food shortages, can be interpreted as a good indicator of moderate vulnerability. Conversely, divestment of assets that are less liquid, such as land or a house, often indicates extreme vulnerability.
Land ownership is, without doubt, the feature of a rural household that encompasses its essential nature and identity more than anything else. A household’s landholdings are a good indicator of its wealth and productive potential. Land ownership is thus important to the concept of food access. As productive capital, land provides a rural household with the ability to supply its own food and generate income with which to access food. As a low-liquidity asset, it provides the household with a last-resort hedging mechanism against food entitlement failure.
The data in Table 7 reveal large differences in patterns of land ownership, between and within communities. In all communities, however, average landholdings are small, ranging in size from .75 ha12 (for non-beneficiary households in Suchiquer) to 3 ha (for beneficiary households in Buena Vista and San Martín). In these latter communities, there is a striking difference between the average landholdings of beneficiaries and those of non-beneficiaries: non-beneficiary households own, on average, only one third of that owned by beneficiaries. Households in San Nicolas own, on average, more than 2 ha of land, which is double the value for Suchiquer. These figures do not reflect the substantial differences in land quality between the two communities, with Suchiquer being characterized by very poor and with depleted soil conditions, a situation illustrated by the yield figures reported earlier.
Table 8 provides an analysis of liquid asset ownership. This includes animals and durable goods. Fowl is by far the most widely diffused type of animal, with project beneficiaries owning, on average, more units than do non-beneficiaries in all communities. With the exception of Buena Vista and San Martín, few households own pigs and even fewer, in general, reported owning larger animals, such as cows or horses. As expected, sheep ownership is common only in the high-altitude communities of Buena Vista and San Martín.
Due to the realization that animal ownership is particularly important – as a source of food supply and food access – for a poor household’s global food security strategy, both projects have promoted animal purchases via credit assistance. Loans are primarily targeted towards the acquisition of chickens and pigs, except in Buena Vista and San Martín where the main thrust of the credit programme is geared towards the purchase of sheep.
Non-beneficiary households in Buena Vista and San Martín own, on average, a slightly higher number of sheep than do beneficiaries. However, in the past year, beneficiaries have increased the value of their livestock holdings (mainly sheep) by approximately six times that of non-beneficiaries (Table 9). This finding seems to suggest that the project’s credit programme may have been responsible for promoting a more equitable distribution of animal ownership, substantially equalizing herd sizes between the two groups. However, because of the nature of the dataset, one can only speculate on the presumed impact of the project in promoting livestock accumulation in these communities.
The ownership of durable goods may be a good indicator when locating wealthier households – although at times it may be indicative only of lumpiness in household income. The purchase of durable goods is usually made during periods of the year when there is relative abundance (for example, at harvest time, or during the months of temporary migration) sometimes even preceding the purchase of more basic necessities. While small appliances, such as radios, may not always indicate a food-secure household, the inability to purchase one of these durables (or the need to sell such an item in distress) provides some useful insights for locating food-insecure households. The data reveal an extremely low level of ownership of durable goods among households in Suchiquer. In contrast, Buena Vista and San Martín seem to be better equipped with respect to durables such as radios and bicycles.
As stated above, changes in asset ownership over time can often be a good indicator of prevailing vulnerability, particularly if stress sales can be clearly identified. Table 9 contains information on asset transactions in the year prior to the survey. In Suchiquer, for example, very few purchases of durables are observed, reflecting the inability of most households in this community to accumulate physical assets that could be used as insurance against the risk of food entitlement failure. Figures for Suchiquer also show a net decrease in the value of animal holdings, as do those for the NTX communities. Two factors may account for this pattern: (i) the value-added associated with raising animals for selling purposes; and (ii) a reduction in animal assets. The share of distress sales contained in total sales can be utilized to separate the two effects. Presumably, for households exhibiting higher shares of distress sales out of total sales, the negative value of net animal purchases are more likely to reflect a reduction in the stock of animals. Households in both communities reported having sold animals because of distress conditions in approximately 50% of the cases.
Access to credit
Credit, particularly if used for consumption smoothing, can be another instrument by which a household can avoid food shortages, at least on a temporary basis. However, IFAD projects do not extend credit for consumption directly. The credit component in these projects is directed primarily towards stimulating the adoption of improved technologies in agriculture and women’s involvement in income-generating activities (such as animal-rearing and the manufacture of handicrafts). However, by promoting access to credit for productive purposes, projects can effectively reduce the vulnerability of poor households to entitlement failure, and thereby improve their food access.
The data in Tables J and K clearly indicate that, in all communities, the projects have been instrumental in improving poor households’ access to credit facilities. Before the start of the projects, there was almost no credit available in the communities, especially for women. However, in spite of efforts to reach women and to promote their involvement in income-generating activities through the disbursement of loans targeted along gender lines, large differences are noted gender-wise in the average amounts allocated. The gap reflects the endemic and persistent problems associated with women’s lack of collateral in all the social contexts under study, as well as the social conventions impairing women’s involvement in productivity.
Claims and social network
A household can improve its food entitlement also through transfers from family members, as well as through access to extended families and social networks. Transfers from family can come in the form of one-time contributions and gifts or from periodic remittances from migrant family members. In many contexts, community networks provide households with the ultimate strategy for coping with food entitlement failures.
None of the communities studied were characterized by heavy long-term or permanent migration. The data reveal very few instances of cash transfers benefiting the selected households. Even in San Nicolas, where several households reported having received cash transfers during the previous agricultural year, the share of total income derived from this source is, on average, extremely low.
Very few households reported having access to informal sources of credit, such as loans from family members or neighbours. However, when presented with a hypothetical scenario of entitlement failure, several households responded that they would rely on family and friends as their first choice in obtaining emergency sources of cash (Table 2). Since about two thirds of the households reported having experienced an entitlement failure of some sort during the preceding year, the apparent discrepancy may be the result of confusion over the term loan.
Access to wild resources
Finally, in defining its access to food entitlement, a household can complement own-produced food supplies and earned income with harvested wild resources, such as wild fruits and plants. These can be utilized as food, as commodities for selling purposes or as productive inputs in cash-generating activities. Information was collected on whether or not household members had dedicated time during the previous year to gathering wild resources. A problem arises in the interpretation of the variable, since a household’s decision to forego the activity may reflect either lower household needs or greater scarcity of local resources. This ambiguity may obscure differentiated household needs within the community. Furthermore, there may be other factors affecting a household member’s decision to allocate time to this activity, such as a potential time clash of wild resource gathering with other activities to which household members are committed (e.g., animal husbandry). Nevertheless, the variable can still provide useful insights, particularly in locating extremely vulnerable groups from a food security perspective. For example, poor households in a resource-scarce community will tend to allocate more time to gathering than their wealthier neighbours, and cover longer distances in doing so, as a result of their greater needs and more limited range of alternatives.
As expected, the data indicate differentiated access to wild resources between communities (Table 2). Significantly fewer households in Suchiquer allocate time to gathering wild resources, compared with the other communities. This pattern probably reflects a scarcity of these resources at the community level. In this specific context, the within-community figures are indicative of idiosyncratic needs across households, with over one third of respondents dedicating some of their time to the activity, in spite of adverse local conditions.
Analysis of the single-factor indicators presented so far has shown the complexity of the food security concept. It also corroborates the call for a more exhaustive measure that would attempt to rank household vulnerability in a less ambiguous manner than single-factor indicators permit – from a food security perspective, at least. Although Section VI misses a benchmark measure against which to compare results, it does manage to combine some of the determinants of food security analysed so far, in an attempt to provide policy-makers with a starting point for the development of an easy diagnostic tool with which to locate vulnerable groups.
Women’s time Allocation and resource control
One of the goals of the IFAD projects is to foster women’s involvement in income-generating activities, and thereby improve their control over household resources. Such a programme is currently being implemented through the inclusion of gender considerations in all main project components. In this section of the present report, an analysis is made of the information collected concerning women’s time allocation and intrahousehold resource control. The data is then related to project activities, in an attempt to assess the role played by the projects themselves in fostering an improved role for women.
The data on women’s participation in agriculture (Table 12) illustrate sizeable intercommunity differences in the number of days women dedicate to agricultural activities. These differences reflect heterogeneous crop portfolios among the different communities and a well-defined division of labour by gender associated with each crop. In line with the findings of many previous studies from Latin America, women contribute only marginally to corn production, participating mainly in harvesting activities. In the NTX communities of the Cuchumatanes, women allocate the greatest number of days per year to agricultural activities, compared with the other communities. These differences are due to the considerable time women dedicate to crops such as broccoli and cauliflower.
Table 13 contains an analysis of the time allocated by women to market activities, such as manufacturing handicrafts, tending animals, selling at local markets and running family stores. The manufacture of handicrafts (mainly straw mats and baskets and rope bags) emerges as the activity to which most of the women allocate a substantial amount of time, both in Suchiquer and in San Nicolas. Ninety-seven percent of the women in Suchiquer dedicate an average of 99 days/year13 to manufacturing handicrafts. Non-beneficiary women dedicate approximately 20% more time than beneficiaries. In San Nicolas, 86% of women dedicate an average of 105 days/year to this activity. Non-beneficiaries allocate 70% more time to it. In Tituque, only 25% of the women reported dedicating time to manufacturing handicrafts, even though they spent the greatest number of days (121 days/year), on average. In the surveyed communities of the Cuchumatanes, almost no women dedicate time to the manufacture of handicrafts for selling purposes.
With the exception of San Nicolas, women in beneficiary households in all communities appear to retain more control over a larger proportion of household income from petty trading than those in non-beneficiary households. As expected, differences in the share of women’s control over income reflect household idiosyncrasies in income composition. Data indicate that women’s control is positively related to the share of the household petty trading income derived from handicrafts and from selling at local markets. In Suchiquer, women in beneficiary households fare substantially better than their counterparts in non-beneficiary households with respect to control of petty trading income, but they lag behind women in all the other surveyed communities. This finding appears to be related to the inability of the women in Suchiquer to market their products directly. Their geographical isolation makes their selling activities dependent on their husbands’ trips to town, often resulting in a partial or total loss of income control over the eventual earnings.
Animal ownership within the household, and also the type of animals owned, appears to be clearly defined across communities. This pattern is often a consequence of cultural norms and conventions. Fowl are owned mostly by women, while pig ownership varies substantially across communities. Women generally do not manage larger animals, such as cows and horses. In Buena Vista and San Martín, sheep are often considered to be owned jointly by a woman and her husband, even though women still claim a substantial share of the revenue gained.
Animal husbandry emerges as a primary activity for women in Buena Vista and San Martín, where women dedicate an average of 201 days/year to this activity. Ninety percent of the women in all households reported that they dedicated at least some time to tending livestock, while very few women appeared to sell in local markets. The only exception to this rule was in San Nicolas, where, thanks to the proximity to roads and markets, 40% of the women are involved in this activity.
In order to interpret these results concerning women’s time allocation and determine the potential repercussions on food security beyond income alone, it is important to note the degree of simultaneity each activity carries with it. Caring for animals and selling at the market necessitate women’s absence from the house or the community for extended periods during the day, or even for several days at a time. The consequences can be particularly harmful for the health and nutrition security of the children, the most vulnerable members of the household, particularly in those cases in which a suitable substitute for women’s time is not readily available.
Table 14 analyses women’s time allocation to home activities. The table shows that women dedicate a sizeable amount of time to making tortillas and to cooking. Not surprisingly, the community in which women dedicate the least time to tortilla-making is also the community where the highest percentage of households own a mill. This finding supports PROZACHI’s effort to introduce time-saving technologies geared towards freeing women from time-consuming activities such as corn grinding. In all communities other than San Nicolas, women still dedicate more than two hours a day, on average, to making tortillas.
Another task performed by almost all women, once or twice a week, is that of washing clothes. This takes from 7 to 10 hours/week, based on family size and distance to water sources. Furthermore, 80% of the women in Tituque stated that they dedicated an average of 6.6 hours/week to carrying water for household consumption. In arid communities such as Buena Vista and Suchiquer, women spend approximately 10 hours/week collecting water.
Women’s time allocation to gathering fuel wood may be an indication of the household income bracket, local availability of fuel wood and level of deforestation, or else the availability of substitutes for women’s time within the household. These factors often interact with one another. The more wealthy families in forest-poor communities will purchase their fuel wood, while poorer households will have no option but to allocate long hours to fuel wood collection.
As expected, the data on women’s allocation of time to fuel wood gathering are similar to the patterns noted above for the gathering of wild resources for food security purposes. In resource-poor communities such as Suchiquer, few women allocate any time to this activity. However, those who do, also allocate the most time, since they cover longer distances. A larger number of women from the NTX communities gather, rather than purchase, fuel wood, but they spend less time on the activity, since the wood is more readily available in these areas.
Table 15 analyses variables related to women’s decision-making within the household. In Buena Vista and San Martín, and also Tituque, a significant number of women claim to be in charge of handling household savings (31% and 34%, respectively). Project participation does not appear to be a differentiating factor in this regard. As for which of the family members decide on the purchase of basic necessities, this decision is generally taken jointly by the male and female heads of household in all communities except in Buena Vista and San Martín, where it appears to be more of a unilateral decision taken by either household head. Women’s involvement in decision-making with respect to household consumption purchases also appears to be greater in beneficiary households than in non-beneficiary households.
Finally, the data on credit revealed that although the projects have played a catalytic role in introducing credit to many women in the surveyed communities, loans continue to be granted along well-separated gender lines. In fact, women receive substantially less money than men, as they generally cannot meet the project’s collateral requirements for larger loans. Also, the small number of loans granted to women for agricultural inputs, both in NTX communities and in Tituque, supports the view that cash crops are controlled by men, in spite of women’s sizeable labour contribution. In addition, evidence suggests that because of the insufficient income generated from the productive investments for which the loans were taken, and also due to the possible loss of control over the borrowed cash, such loans often become a liability for women who are left with the responsibility for repaying them.
In conclusion, information on the intrahousehold sharing of labour and resource control is extremely important for targeting purposes. The data illustrate that activities such as handicraft manufacturing and petty trading in local markets are linked to an improvement in women’s control of the household income. Furthermore, the data confirmed that small animals, such as chickens, are generally under the control of women. Finally, in spite of an increased involvement in market activities, women continue to dedicate long hours to home production. This pattern raises some concern regarding an excessive increase in women’s workloads.
Agricultural commercialization and HFS
Both of the IFAD projects surveyed include a component geared towards the promotion of cash crops. The type of crop promoted varies according to local conditions. For example, coffee is vigorously promoted by PROZACHI in Tituque. In the Cuchumatanes, the project’s commercialization component has focused on supporting potato production in the highland communities of Buena Vista, and non-traditional export crops, such as broccoli and cauliflower, in the mountainside communities of Chochal, Torlón and Chichalum. Examining the relationship between the latter type of crops and HFS, the relevance of NTX adoption to the HFS debate can be seen from the following features of NTXs: (i) NTX are highly labour intensive, thus inducing large labour reallocation within the household; (ii) the link among NTX adoption and nutritional performance and food security has proven weak in other studies; (iii) past experience with NTX adoption among Guatemalan smallholders, and the potential agronomic and marketing problems that generally follow, raise some concerns about the use of NTXs as an instrument for rural development; (iv) the temporary success of previous NTX experiences appears to have been conditioned by strong financial support from outside; and (v) NTXs are not used for consumption purposes.
The objective of this section is to assess the impact of NTX adoption on a set of variables presumed to have a direct or indirect link with food security – namely, household food dependency, income, food production and productivity and some intrahousehold factors such as women’s time allocation and resource control.
In the three communities of the NTX sub-sample (Chochal, Torlón and Chichalum) 34 households out of 54 reported growing broccoli. Based on this classification, the results of the preliminary analysis are reported in Table 16.
As shown by previous studies, NTXs appear to generate significant improvements in household income, with NTX households displaying average income levels of four times that of non-NTX households. The difference is a result of higher on-farm income, mainly from NTX crops. NTX farmers derive over 70% of their total income from on-farm labour alone – a figure that suggests inadequate income diversification and excessive reliance on risky income-generating activities. Although many NTX households are characterized by very low diversification both in income and in crops, they still score significantly better than non-NTX households.
On average, NTX farmers own and cultivate more land than non-NTX households. In view of the fact that no significant differences are noted in average food crop yields, the lower dependency on food markets observed for NTX households may be a reflection of the differences in land ownership and use.
As expected, women from NTX households allocate twice the amount of time to on-farm labour than that allocated by their non-NTX counterparts. However, in spite of this greater involvement in farming activities, the former do not appear to have made any significant reduction in the time they allocate to home production, although they claim control over a larger share of both household petty trading income and small animal stocks.
7/ A more detailed analysis of the dataset is given in Appendix 1.
8/ For approximately four months a year, the household diets in these communities also include potatoes (as a complement to corn), of which the communities are major producers.
9/ 6 quetzales = USD 1 in July 1996.
10/ Some women in beneficiary households also sell their handicrafts through a store in Guatemala City set up with the support of the project. Thanks to this new retail avenue, they get better prices for their products and are able to retain control of the resulting income, since payment is made by the project directly to them. However, because of limited demand, only a few women in Suchiquer and Tituque have been able to benefit from this initiative.
11/ The income figures from on-farm sources do not always reflect net revenues, due to input costs. The discrepancy may be especially large when cash crops are involved, because of the higher intensity in purchased input use in these types of production. Consequently, the substantially higher income levels reported for the communities growing cash crops can be the result in part of differences in input costs.
13/ The computation of the number of days is based on eight-hour working days.
