FINDINGS FROM VILLAGE AND VILLAGE/TOWN SAM MULTIPLIER MODELS

The village and town SAMs provided the basic data input for the six multiplier models. In the case of the Erongaricuaro region, two SAMs, one for the town of Erongaricuaro and one for the three villages, were stacked into a single SAM matrix with four rest-of-the-world accounts: a local endogenous trade account within the microregion, two other rest-of-Mexico accounts, one for non-competitive intermediate inputs and the other for commodities, and the rest of the world outside Mexico, essentially the United States. Lewis and Thorbecke (1992) combined village and town production activities into a single set of accounts in their SAM multiplier study of a Kenyan village/town economy. Here they are kept separate in order to highlight differential impacts of income changes on production in the two economies. The village and town multiplier models were all programmed using the General Algebraic Modelling System (GAMS) software.

Three experiments were carried out on each of the villages and on the Erongaricuaro region. They were designed to illustrate household income linkages and the implications of agricultural supply constraints for rural income multipliers. The first set of experiments explores the impacts of a US$100 change in exogenous household incomes upon production, factor and household incomes, and trade linkages. This is the least constrained type of SAM multiplier, because it assumes that the output of goods is perfectly elastic; that is, the model is constrained only on the demand side.

Where input supplies are tight or there are short-term capacity constraints, the assumption of perfectly elastic supply response may not be tenable; the model may be constrained on the supply rather than the demand side. In a second set of experiments, agricultural production was constrained to be perfectly inelastic in each of the villages and the village-town economy, as in semi-input-output and economic-base models (see Haggblade, Hammer and Hazell, 1994), using the procedure suggested by Lewis and Thorbecke (1992). We then re-estimated the household income multipliers. In both cases the experiment corresponds to the PROCAMPO programme, in which farmers receive direct income payments per hectare cultivated with basic grains. The first two experiments thus examine the likely impacts of US$100 of PROCAMPO payments. The third set of experiments explores the local economy-wide effects of loosening the agricultural supply constraint by US$100. These experiments might be viewed as simulating the multiplier effects of supply-augmenting technological change in agriculture. The results of all three experiments for the five villages are presented first, followed by the findings for the Erongaricuaro village-town microregion.

Village income multipliers

SAM multipliers without supply constraints

Multiplier effects including farm/non-farm linkages from changes in rural incomes were estimated by increasing total exogenous household income by US$100 and distributing this income across household groups in proportion to their initial shares in total household income in the village. The resulting village SAM multipliers unfold as follows. In the first instance, incomes of the directly affected households increase by the amount of the change in exogenous income. This increases household demands for farm and non-farm goods by the amount of the exogenous income gain times the household-specific marginal expenditure propensities. Because both farm and non-farm production sectors are unconstrained in this model, that is their supplies are perfectly elastic, output increases to meet these increased household consumption and investment demands. This sets in motion a round of production linkages as sectors purchase inputs from one another to support their production expansion. As production increases, new value-added is generated; this value-added is channelled back into households as income, which in turn unleashes a new round of household expenditure increases. Along the way, demands for consumption goods, investment goods and production inputs from outside the village create income leakages, dampening the income multipliers at each round of the process. Although these village imports are a leakage from the point of view of the village, they are mostly non-farm goods and thus represent an important growth linkage between villages and regional or national economies. The SAM multipliers reported in Table 6 represent the end result of multiple rounds of increases in household incomes and production that result from the initial increase in exogenous household income.

TABLE 6
Village SAM multipliers resulting from a US$100 increase in rural household incomes^a (unconstrained model)^b

^a Distributed across household in proportion to initial income shares.
^b All supplies perfectly elastic.

The impacts of the US$100 exogenous income change vary strikingly across all villages and across production activities within the villages. Total household income increases by US$113 in Concordia, US$121 in El Chante, US$161 in Napizaro, US$124 in Naupan and US$146 in Reyesoghpan. Much of the impressive increase in income in Napizaro comes from livestock, on which households spend a large marginal share of consumption and investment expenditures. Because livestock production uses little labour, the gains from output expansion in this activity accrue mostly to capital and land. As for Napizaro, most of the increase in Reyesoghpan GVP comes from livestock. Livestock production in the Sierra consists of small animals, however, using family labour. For this reason, the gains from output expansion in Reyesoghpan accrue mostly to family labour.

Apart from livestock, farm/non-farm linkages within the three villages are relatively modest: non-agricultural production rises by just over US$2 in Concordia, by US$12 in Napizaro, by US$7 in Naupan and by US$0.3 in Reyesoghpan; there is no non-farm sector in El Chante. Farm/non-farm linkages outside the villages are striking, however. In all four cases, total demand for goods produced outside the village, mostly manufactured, rises by nearly as much as the amount of the initial exogenous income injection; in one case it is more than this. In Concordia, Napizaro and Naupan most of this external linkage is created by consumer demand for local commercial-sector goods, whereas in El Chante and Reyesoghpan most of it is with the regional products markets. In El Chante, where saving and investment propensities are very high, much of the external linkage is a result of investment demand for outside goods.

The distribution of income gains across household groups is unequal in Napizaro, where most of the indirect income gains go to owners of capital. Income gains are more evenly distributed in Concordia and especially Naupan, Reyesoghpan and El Chante, where most value-added accrues to family labour. In the present experiment an initially unequal income distribution, as in Napizaro, tends to promote an unequal distributional outcome, because the US$100 exogenous income gain is allocated across households according to their initial shares in total income. In all cases, the higher second-round effects of the income change contribute significantly to household incomes. In addition to the initial income injection, for example, the increase in village production stimulated by higher rural household expenditures adds another 62 percent to household income in Napizaro. In general, the households that are most engaged in the production of goods and services for village markets benefit most from the higher second-round effects of the income change.

Village SAM multipliers in the presence of agricultural supply constraints

The size of SAM multipliers resulting from an increase in rural household income depends critically on the supply response of village or town production sectors. In the extreme case of perfectly inelastic supply response in all production sectors, there can be no income multiplier. In real life, supply elasticities vary across sectors. In agriculture, supply is generally inelastic in the short term. Technological and environmental constraints may result in sharply decreasing marginal returns to factor inputs. If new technologies are not available or accessible to small farmers, the supply response may be inelastic in the long term as well. A number of factors affect the accessibility of new technologies and farmers’ willingness and ability to adopt them, including information, risk and the absence of formal or informal insurance markets, lack of access to credit, environmental constraints and various other types of input and output market failures (Feder, Just and Zilberman, 1985; Bellon and Taylor, 1993; Smale, Just and Leathers, 1994; Meng, 1997).

In the second set of experiments, the supply response is constrained in all agricultural sectors to be perfectly inelastic. The exogenous increases in household incomes in this experiment are identical to those in the first experiment. The resulting increase in household demand for village-produced goods, however, now stimulates an increase in output only in the unconstrained non-agricultural sectors. In the agricultural sectors - crop and livestock production - a rise in demand by households or firms can be satisfied only through a decrease in village exports, or marketed surplus. An inelastic supply of agricultural goods constrains the expansion of production sectors that demand inputs from agriculture.

The findings from the constrained-SAM multiplier experiments appear in Table 7. The increases in village income resulting from exogenous rural household income changes are much smaller than in the unconstrained experiments. On the production side, output changes in all agricultural sectors are assumed to be zero. The effect of inelastic agricultural supply on the non-agricultural sectors varies, depending critically upon reliance by the production sectors on the constrained agricultural sectors for inputs. The extreme case is Reyesoghpan, whose economy is the most dependent on agricultural production.

Although increases in household incomes are less than in the previous experiment, the change varies strikingly across villages and household groups. All household groups are hardest hit by agricultural supply constraints in Napizaro and Reyesoghpan, where agricultural production accounts for an important share of household income. In all villages, the household groups most closely linked to the constrained agricultural sectors suffer the most from the supply constraints.

The smaller total income multipliers result in weaker demand linkages between the villages and outside markets. The adverse effects of agricultural supply constraints are larger for village trade than for village incomes. In short, the capacity of changes in farm incomes to stimulate demand for non-farm goods turns critically on the agricultural supply response.

TABLE 7
Village SAM multipliers resulting from a US$100 increase in incomes^a (constrained model)^b

^a Distributed across household in proportion to initial income shares.
^b Agricultural production sector supplies perfectly inelastic.

Village multipliers from loosening agricultural supply constraints

Loosening agricultural supply constraints can have a multiplicative effect on village incomes and stimulate farm/non-farm linkages. The third set of experiments was designed to estimate agricultural supply multipliers in the five villages. To do this, the constrained SAM of the previous experiment was used as the base model; agricultural output was then exogenously increased by US$100, with this increase distributed across agricultural production sectors in proportion to their initial shares in total agricultural output. This experiment might represent a loosening of supply constraints through technological change, credit market development or public investments in marketing, extension or transportation infrastructures. Alternatively, by changing the sign of the supply shift from positive to negative, it might represent the negative multiplier effect of environmental degradation diminishing production.

The increase in agricultural supply has several impacts on a village economy. It increases intermediate input demands by the constrained agricultural sectors, stimulating production and imports of these inputs and setting in motion a round of production linkages in the village. It also increases payments of value-added income to village households, which results in a rise in household expenditures on goods and services supplied inside and outside the village. The household demands stimulate a new round of production increases in the village, increased demands for intermediate inputs and value-added payments to households. The SAM multipliers resulting from the increase in agricultural-sector supply are presented in Table 8. The exogenous increases in agricultural-sector supply appear in the first three rows of the table.

This experiment produces by far the largest increases in village value-added, reflecting the pivotal role of agricultural supply constraints in shaping rural income linkages. In this experiment, all village income effects are indirect, through production. The increased agricultural production generates a large increase in labour value-added in the four most labour-intensive villages; in Napizaro, where most of the output gain is in the capital-intensive livestock activity, physical capital and land value-added represent most of the increase in value-added.

It is not surprising that agricultural households benefit disproportionately from the supply shift. The fact that non-agricultural households also benefit, although relatively modestly (see Concordia and El Chante columns), reflects the existence of some farm/non-farm commodity and factor-market linkages within these villages. As in the previous experiments, the distributional effect on household incomes is much more equitable in Concordia, El Chante, Naupan and Reyesoghpan than in Napizaro, again reflecting the capital intensity of Napizaro livestock production and an unequal distribution of capital in that village.

TABLE 8
Village SAM multipliers resulting from a US$100 increase in agricultural supply^a (constrained model)^b

^a Distributed across agricultural production sectors in proportion to total initial output levels.
^b Agricultural production sector supplies perfectly inelastic.

The increase in agricultural supply has modest positive effects on non-farm production in the five villages, but strong effects on village demands for manufactures. Again, the increase in local commerce is much lower in the villages that are well connected with the regional product markets, El Chante and Reyesoghpan.

Income multipliers in the village/town microregion

The findings reported above reveal modest farm/non-farm income linkages within Mexican villages. The Erongaricuaro microregion survey affords the opportunity to estimate farm/non-farm linkages between village economies and the town economy to which they are most directly linked. For each variant of the SAM multiplier model presented for the villages, two sets of experiments were performed for the village/town microregion. The first involves an exogenous US$100 increase in village household incomes for the unconstrained and constrained cases, or an exogenous US$100 increase in constrained-sector supply for the constrained case. These experiments were repeated for the town economy. The multiplier effects of these two experiments on village and town incomes are summarized in Table 9, Table 10 and Table 11. Columns A and B in these tables present multiplier impacts on the village and town economies that result from a US$100 exogenous increase in village-household incomes (Table 9) or in constrained-sector supply (Tables 10 and 11). Columns C and D present impacts on village and town economies associated with the same exogenous increases in town-household incomes or supply.

Unconstrained village/town SAM multiplier

The multiplier effects of the exogenous income increase are substantially larger in the villages than in the town; for example, total income increases by US$231 as a result of the US$100 exogenous increase in village household incomes (column A). By contrast, town income rises by only US$133 as a result of the US$100 exogenous increase in town household incomes (column D). The much smaller value-added multiplier for the town can be explained by the town’s high degree of openness to markets outside the village/town region. These outside market linkages transfer the multiplier out of the village/town microregion. The villages, on the other hand, are less integrated with outside markets; a large proportion of changes in demand is satisfied through changes in village production. In the villages (column A), the exogenous income gain is associated with production increases that dwarf the production increases in the town associated with a US$100 exogenous income change there (column D). In both town and village economies the largest increases are for commerce, which mainly represents a leakage from the local economy.

The cross effects of the village income increase on the town economy appear in column B; the town income increase on the village economy is shown in column C. In all cases, cross effects are small relative to the own effects. Because the town’s commercial sector plays an important role in satisfying village consumption demands, it is the town’s main beneficiary from increases in village incomes. The cross effect of village income increases on town commerce of US$38 actually exceeds the US$17 effect of a US$100 increase in the town’s income on its own commerce.

TABLE 9
Village/town SAM multipliers resulting from a US$100 increase in rural household incomes^a (unconstrained model)^b

^a Distributed across household in proportion to initial income shares.
^b All supplies perfectly elastic.

TABLE 10
Village/town SAM multipliers resulting from a US$100 increase in incomesa (constrained model)b

^a Distributed across household in proportion to initial income shares.
^b Agricultural production sectors supplies perfectly inelastic.

In the village/town SAM, there are three household groups. Commercial households are household-farms or household-firms engaged in commercial production of some sort. Subsistence households are households that produce for their own subsistence only. Net-buyer households include households that satisfy little or none of their consumption demand from own production. Commercial households enjoy the largest income gains; they benefit directly from the exogenous income gain and indirectly through increased demand for the goods they produce. Taking into account the multiplier effects of the exogenous income change, the income of all household groups more than doubles. The large divergence between total and direct effects on household incomes reflects the involvement of households in production activities stimulated by income linkages in the village/town microregion.

In the town, net-buyer households have the highest incomes to begin with and thus initially benefit most from the income transfer. As in the villages, the indirect effects of the income change are largest for commercial households, though overall indirect effects are more muted. They are smallest for subsistence households.

Although the income change stimulates a village demand for town commerce, external linkages with the rest of Mexico are relatively small. The US$100 exogenous income increase in the villages increases village demand for goods from the rest of Mexico by just over US$22. By contrast, the US$100 increase in town incomes increases town demand for goods from the rest of Mexico by US$73. Village households spend a significant part of their income gains on goods brought home by migrants from the United States; the demand linkage with the United States is smaller for the town.

Village/town constrained multipliers

Multiplier effects of the exogenous income change plummet when the output of agricultural and livestock activities is constrained. The effect of the US$100 exogenous income change in the village on village income falls from US$231 in the unconstrained case to US$143; the effect in the town on town household incomes falls from US$133 to US$116. The cross-multipliers decrease as well, although in the case of the village-to-town effect the decrease is much smaller in relative terms than the decrease in village and town own multipliers.

By assumption, there is no change in production in the output-constrained agricultural and livestock sectors in the village. To compensate for these constraints, the village economy shifts its demands away from village supply-constrained goods in favour of unconstrained-sector goods, including those supplied by the town. As a result, the output of unconstrained sectors decreases by less in relative terms than the decrease in total village value-added; so does the multiplier effect on gross town product. In short, supply constraints in the village substantially dampen the village income multiplier. The negative effects of village production constraints on the town economy are smaller, however, as village production activities and households attempt to compensate for the supply constraints by shifting their intermediate and consumption demands in favour of town-produced goods.

TABLE 11
Village/town SAM multipliers resulting from a US$100 increase in agricultural supplya (constrained model)b

^a Distributed across agricultural production sectors in proportion to total initial output levels.
^b Agricultural production sector supplies perfectly inelastic.

The decrease in village income and production multipliers is reflected in household incomes, which increase by about 40 percent less than in the unconstrained case. Agricultural supply constraints substantially reduce the indirect effects of the income change, especially for commercial households, which are heavily involved in the constrained-sectors production. As in the first experiment, commercial households enjoy the largest gains from the exogenous income change in the village. The ratio of indirect to direct income changes drops from 149 percent to 46 percent for commercial households, however; in the supply-constrained case it is similar to the 44 percent of village net-buyer households. In the town, the ratio of indirect to direct income effects for commercial households falls from 40 percent to 12 percent.

Village multipliers from loosening agricultural supply constraints

Loosening agricultural supply constraints unleashes powerful income multipliers in the village/town economy, as in the village economies examined earlier. In contrast to the effects of the exogenous income increases in the first two experiments, increases in the exogenous supply of agricultural and livestock activities has a similar impact on the village and town economies. A US$100 increase in constrained village agricultural and livestock supply stimulates a US$101 increase in total income in the village. The same exogenous increase in town supply results in an US$87 increase in town income, a much larger indirect effect than in the other two experiments. The constrained-production multiplier is less than 20 percent larger in the villages than in the town; by contrast, the income multipliers reported above were of the order of five times larger in the village. This finding highlights the strong local linkages emanating from agricultural and livestock activities within the village/town economy. By targeting these sectors, the US$100 increase in supply exploits these linkages directly, thus creating larger income multipliers than those generated by the US$100 increase in household incomes, which affect the agricultural and livestock sectors indirectly through household expenditures.

Because households do not benefit directly from the exogenous increase in supplies from agricultural and livestock activities, all the household-income multipliers are indirect in this experiment and the impacts on households’ total incomes are smaller than in the other two experiments. The households most involved in agricultural and livestock activities benefit disproportionately from the supply change. This is the only experiment in which town commercial households have an advantage over net-buyer households because of their involvement in agricultural and livestock activities.