# VIII. Analytical Results and Discussion

This chapter summarizes and compares estimates of profit per unit output of broiler and hog production between small and large-scale farms and between independent and contract farms, first without incorporating the cost of family labor. It then proceeds to incorporate the effects of policy distortions, explicitly accounting for the cost of family labor, the capture of environmental externalities, and the impact of transaction costs on farm technical/allocative efficiency. The effect of prices, fixed factors, and technology on profit per unit are also discussed prior to the treatment of transaction costs.

The results for broiler farms and hog farms are reported separately beginning with the discussion on broiler farms.

## 8.1 Broiler Farms

8.1.1 Comparative Profit Performance

Profits per kilogram of output by scale, with and without contracts

Figure 8.1 shows farm profits per kilogram output across farm types. Per kilogram profits from production and sale of main output (live broilers) were highest for contract growers. Small contract growers received marginally higher profits (4.1 pesos per kg) than their larger counterparts (4.0 pesos per kg).

Large independent producers realized the lowest profits per kilogram (1.1 pesos). Small independents earned significantly higher profits per kilogram (1.6 pesos per kg) than the large independents.

Regarding smallholder contract farms, distinguishing between price and fee contracts reveals the advantage of the fee contact system over the price contract and the rest.

Figure 8.1 Profit per kilogram by scale, with and without contracts, Broiler Farms, Philippines, 2002

Total farm profit by scale, with and without contracts

Total farm profit refers to profits over the latest cycle. Over 93 percent of the sample practiced the 'all-in, all-out' scheme, with an average of four to five cycles per year, each production cycle lasting 41-42 days. Figure 8.2 shows a comparison of total farm profits per cycle.

The contrast in total farm profits is stark. Scale matters. Although profits per kilogram were also relatively high among smallholder independents, total farm profits per cycle were only five percent of those obtained by large fee contract producers, and less than a tenth of the profits earned by large independent producers. Although independent smallholder profits were close to half the profits earned by small price contract producers, they were only about one-sixth those earned by smallholders with fee contracts.

Figure 8.2 Total farm profits per cycle, by scale, with and without contract, Broiler farms, Philippines, 2002.

Small fee contract growers realized more than twice the earnings of their price contract counterparts. Large fee contract producers, on the other hand, earned more than three times more than the fee contract smallholders.

Between the two types of large producers, the fee contract growers had 60 percent higher profits than the large independents, notwithstanding a smaller average flock size and lower investments per bird. Among the large farms, larger scale did not translate to higher profit overall.

Profit per unit output and scale of operations, by size quintile

Figure 8.3 shows profit per kilogram by quintile. Profit per kilogram of output, using pooled data, revealed that households belonging to the lowest quintile, producing an average of 695 kg of output per cycle, generated the highest profit at 5.3 pesos per kg. Profits per kilogram fall progressively as scale of output increases, but ticks up in the last quintile. On a per unit output basis therefore, smallholders are not inferior in generating earning.

Figure 8.3 Profit per kilogram, by quintile, Broiler farms, Philippines, 2002

8.1.2 Accounting for Market distortions

Market distortions considered are three: the non-costing of family labor; the non-accounting of social costs generated by negative environmental externalities from by-products in the form of livestock waste (manure) and dead animals; and trade and social subsidies to large integrators and to cooperatives.

Adjustments for cost of family labor

Costing unpriced family labor is expected to reduce the economic profits of the group that uses this resource most intensively: smallholders. Figure 8.4 displays comparative profits per unit of output among broiler farms after costing family labor. Using the sample wage rates to value the family labor involved, independent smallholders' average profit per unit of output is slightly reduced, from 1.6 to 1.3 pesos per kg of output. The adjustment has negligible impact on the three other groups. While this was to be expected for the large producers, the marginal effect of costing family labor on the smallholder contract group is revealing. This arises from the fact that the small contract producers had flock sizes on average five times larger than those of the independent smallholders. Profit per unit output, however, remains lowest among large independent producers.

Figure 8.4 Profit per kilogram after costing family labor, Broiler farms, Philippines, 2002

Adjustments for cost of environmental externalities

Comparative expenditures on mitigation of environmental externalities. Expenditures on mitigation of environmental externalities consisted of the following: revenue from sales of animal manure; the value of manure spread on own fields; costs of disposing of livestock waste and dead animals; fees, environmental fines, and environmental taxes paid; and compensation paid to neighbors for pollution. Where none of the related activities to mitigate environmental externalities is done, the value of expenditures on mitigation is zero. That is to say, none of the environmental costs imposed is visibly captured by the producer in this private economic activity.

Expenditures on environmental mitigation by scale, with and without contracts. Table 8.1 presents estimates of expenditures per 100 kg of output on environmental mitigation. Dividing the sample into independents plus price contract growers on the one hand and fee contract producers on the other hand, and distinguishing them further by scale of operations, yields two significant findings in terms the amount spent by broiler producers on environmental mitigation (measured in pesos per 100 kg of live-weight output). First, contract growers spent more per unit than independents, regardless of scale. Second, small independents spent about the same amount per unit as large contractors.

The question is, if mitigation expenditures differ, what determines differential rates of environmental mitigation among broiler producers?

Table 8.1 Estimates of expenditures on environmental mitigation per 100 Kilograms of output, Broiler farms, Philippines, 2002

 ENTRY SMALLHOLDER (<10,000) COMMERCIAL (>10,000) Independent Contract Independent Contract Expenditure per 100 kg output 48.3 23.5 14.9 22.2

Source: UPLB-IFPRI LI Field Survey, 2002-03.

Determinants of expenditure on mitigation of externalities. Tobit regressions were run on environmental mitigation expenditures by broiler raisers on household, farm, and production characteristics that were theoretically conceived to affect mitigation expenditure behavior. Table 8.2 presents the results of the pooled and regrouped regressions.

For the pooled sample the salient results were four:

• In the regression for the pooled sample, all statistically significant explanatory variables had the expected directional relationship.

• The age of the household head, the ability to sell manure in the last two months, and the existence of own or operated croplands increased the value of the latent dependent variable (the average producer expenditure on mitigating environmental externalities per kg of output).

• The farm's being situated in an area officially classified as non-agricultural and being located closer to residential communities tended to increase producer expenditures on environmental mitigation.

• The lower the physiological efficiency of the production process (i.e., in converting feed to meat (higher FCR value)), the higher the average producer expenditures on environmental mitigation.

To distinguish the differential behavior of expenditures on mitigation of externalities by scale of operations, the sample was divided into two groups: 'all smallholders' and 'all large farms'. Table 8.2 also presents these regression results. The salient results follow, first for smallholders, then for large farms.

For smallholders:

• For smallholders, all the statistically significant explanatory variables in the pooled sample held their ground and retained their directional relationship with the latent value of mean expenditure on environmental mitigation.

• The strength of the relationship between these significant explanatory variables and the value of mitigation expenditure were each, respectively, stronger among smallholders themselves than in the pooled sample.

• For smallholders, the level of education of the farm decision-maker (household head) emerged as statistically significant but carried the wrong directional sign.

Table 8.2. Determinants of farm expenditure on mitigation of environmental externalities from broiler production, Philippines, 2002.

 EXPLANATORY VARIABLES Pooled Sample N=116 All Small- holders n=70 All Large Farms n=46 All Contract Farms n=62 All Independents n=54 Small Contracts n=31 Small Independents n=39 Coefficient Coefficient Coefficient Coefficient Coefficient Coefficient Coefficient Education of HH head (yrs) ns -0.214 ns ns -0.246 -0.092 -0.314 Age of HH head (yrs) 0.041 0.058 ns ns 0.078 ns 0.066 Able to sell manure in last 2 mons. (Dummy) 2.861 4.119 0.410 0.794 4.495 1.283 5.432 Land class is Agricultural(Dummy) -0.717 -1.202 ns ns -1.379 ns -3.347 Has crop land (Dummy) 1.120 2.214 ns ns 1.702 1.043 4.402 Within LLDA jurisdiction(Dummy) ns ns ns ns ns ns ns Wage rate (PhP per hr) ns ns ns ns ns ns ns No. of mortalities in last cycle(birds) ns ns ns ns -0.008 0.004 ns Distance to nearest Residential community (Kms) -0.154 -0.213 0.127 ns ns -0.119 ns No. of DOCs loaded previous batch (birds) ns ns -0.00001 ns ns ns ns FCR 2.214 2.956 ns ns 4.140 ns 4.396 Constant -6.487 -9.075 ns ns -11.540 ns -12.240

Source: UPLB-IFPRI LI Field Survey, 2002-03.

For large farms:

• For large farms, only the ability to sell manure retained statistical significance and the directional relationship as was hypothesized.

• Being situated within the jurisdiction of the Laguna Lake Development Authority (LLDA), a government body with juridical authority to monitor and impose sanctions on entities that pollute Laguna Lake and its tributaries, emerged as significant only among large producers and carried the expected directional relationship with mitigation expenditure behavior.

• The size of day-old chicks loading in the previous cropping, a proxy for scale within the sample, emerged as significant, with a negative sign, although with an extremely small coefficient size (close to zero).

• In general, among large farms, even for the explanatory variables that are significant, the sizes of the coefficients are relatively small (i.e., the magnitude of change in expenditure on mitigation per kilogram of output is not expected to be large).

With regard to production arrangement effects, the following results were found for independents and contract growers.

For independents:

• For all independents, regardless of scale, the age of the household head, the ability to sell manure, the classification of farm lands, existence of croplands operated, and FCR were retained as significant, all with coefficient sizes larger than those in the original pooled sample.

• Distance to the nearest residential community loses significance; education of household head and number of animal deaths over the last cycle emerge as significant but with the opposite directional relationship than expected. The coefficient size of the number of animal deaths is small (close to zero), and appears to be capturing the reverse 'scale effect' exhibited previously by large producers on mitigation expenditures per unit of output.

• For small independents, all variables explaining the behavior of smallholders, except distance to nearest residential community, retain their statistical significance, and all have coefficient sizes larger than in the pooled smallholder sample. In addition, the wage rate comes into play with the expected negative directional sign.

• As a group, the small independent producers exhibited the strongest response of environmental mitigation expenditure on these hypothesized variables.

For contract growers:

• For the group of small contract growers, only the ability to sell manure, the existence of croplands operated, and distance to the nearest residential community are retained as significant, relative to the original pooled sample. The coefficient sizes of these variables are all smaller than for the independent subgroup.

• The new explanatory variable that emerges as significant for small contract growers is the number of animal deaths in the last cycle, with the expected positive sign but a very small coefficient size (close to zero).

In sum, regarding environmental mitigation behavior we can deduce the following:

• The ability to sell manure is crucial for all farms in positively contributing to environmental mitigation behavior. All producers respond to the market for manure, where it exists. Smallholders as a group, and independent smallholders in particular, respond the strongest to the market for manure.

• The existence of croplands operated by the household is a key factor that positively contributes to environmental mitigation behavior, except among large farms as a group. The effect of this factor was strongest among independent smallholders and significant to all smallholders in general. This means that where croplands are available, smallholders tend to exert greater effort to mitigate environmental externalities. This is important, particularly in the absence of a market for manure for independent smallholders, when their volume of manure by-product per cycle is relatively, or absolutely, too small to merit pick-up by regular manure traders.

• Only smallholders as a group are sensitive to being close to residential communities. Although this variable loses significance among small independents as a sub-group, this is compensated by an even stronger response of being situated in an area classified as non-agricultural (i.e., residential, industrial/commercial). Smallholder farms, in general, are closer to residential areas than large farms. Smaller farms appear to respond to residential community pressure to control externalities from their operations.

• The technology variable, FCR, emerges strongly as a determinant of mitigation expenditures for all farms, except for large farms and small contract farms. The non-significance of FCR among large farms (which include fee contract farmers) and on small contract farms may simply indicate the narrower variation in FCR among contract farms, since this measure of performance is contractually stipulated as a determinant of compensation to these producers. An implication of the significance of FCR is that access by smallholders, particularly independent smallholders, to better bred day-old chicks and better feed quality, all things equal, could reduce their expenditures on environmental mitigation (i.e., reduce per unit cost of production in general) and increase per unit and total farm profits, as well as net societal benefits.

• Whether a farm is located within the jurisdiction of the LLDA is significant only among large farms. The fact that this variable is second in coefficient size only to the ability to sell manure implies that aside from the market for manure, large farms additionally respond to threats of sanctions by an agency that has environmental enforcement authority. It may also imply that LLDA trains its attention and penalty exaction mainly on easily visible commercial farmers (those with capacity to pay environmental pollution taxes and penalties and other fees).

Trade and social subsidies to integrators and cooperatives

Differential access to credit. Access to inputs credit is one feature that differentiates the production opportunities of contract growers from those of independent producers. Integrators automatically provide contract growers their intermediate inputs of day-old chicks, feeds, and vaccines and pharmaceuticals. For independent producers, that is not the case.

As Table 8.3 shows, only 13 percent of independent smallholders could obtain feeds credit from their suppliers, compared to 22 percent of large independents. For day-old chicks, only 10 percent of small independents obtained credit, and for vaccines and pharmaceuticals the figure was only three percent. In contrast, 17 percent of large independents obtained day-old chicks and pharmaceuticals on credit.

Table 8.3 Access to credit for inputs by broiler farms, by scale and production arrangement, Broiler Production, Philippines, 2002

 ENTRY SMALLHOLDER (<10,000) COMMERCIAL (>10,000) Independent Contract Independent Contract Day-old chicks 9.7 100.0 17.4 100.0 Feeds 12.9 100.0 21.7 100.0 Vaccines/Pharmaceuticals 3.2 100.0 17.4 100.0

Source: UPLB-IFPRI LI Field Survey, 2002-03.

Access to formal credit is likewise difficult. As shown in Table 8.4, only six percent of all farmers actually borrowed from a bank for production purposes, with the interest paid averaging 15.4 percent per annum. If the farmers really needed credit, fewer independents thought they could borrow from a bank (24-28%) as compared to contract farmers (33-41%).

Table 8.4 Access to bank credit and lowest interest rates at which farmers could borrow, Broiler farms, Philippines, 2002

 ENTRY SMALLHOLDER (<10,000) COMMERCIAL (>10,000) Average Independent Contract Independent Contract Percent of farms that actually borrowed (%) 6.5 6.5 8.7 3.2 6.0 Average interest rate paid (% p.a.) 27.6 19.8 11.0 14.0 18.4 Percent of farms who think they could borrow from the bank (%) 27.6 41.4 25.0 33.3 32.4 Perceived bank interest rate (% p.a.) 9.7 14.4 13.8 14.8 13.2 Farmer's perception of bank interest for a general loan (% p.a.) 12.1 13.7 11.0 11.6 12.2 Percent of farms who believe they can get loan at 12% p.a., or less (%) 54.8 64.5 73.9 51.6 60.3

Source: UPLB-IFPRILI Field Survey, 2002-03.

The interest rate that the producers thought a bank would charge varied by group, with the only consistent pattern being that contract growers expected higher interest rates than their independent counterparts. On average, the expected rates were between 12 percent and 13 percent per annum. The table also shows the proportion of producers who believed they could get a loan at less than 12 percent per annum. No distinct pattern arises in this distribution, except the surprising result that the lowest proportion of farmers that believed they could obtain a loan at less than 12 percent per annum was unexpectedly in the large independents group. It may be worth noting that broiler producers believed they could borrow elsewhere-from relatives and friends-at much lower rates of interest. In a certain sense, smallholders may have greater access than commercial producers to informal sources of socialized credit for production purposes.

The extremely low incidence of farmers obtaining formal credit emphasizes the importance of suppliers' credit for inputs. The willingness of suppliers to provide inputs on credit depends on the cost of supplying them, which, among others, includes the cost of actually producing the inputs (e.g., feeds).

Subsidy on feeds and day-old chicks to the contract growers. It is often asserted that the large integrators offering fee contracts obtain main feed ingredients at lower prices than the smaller feed-milling integrators or other commercial feed mills. Likewise, their costs in producing day-old chicks are said to be much lower than those incurred by other producers.

On major feed ingredients, the sheer size of the integrators allows them to gain access to corn imported under the minimum access volume (MAV) commitments of the Philippines to the WTO for the period 1995-2004. With its strong protectionist policy on corn, government MAV imports are obtained at a 35 percent in-quota tariff rate, while imports outside the quota are slapped with a 60 percent rate. Feed millers who procure corn from domestic sources face domestic prices that are at par with at least 60 percent implicit tariff protection.

The large integrators are also able to import wheat in bulk, at 10 percent tariff on feed wheat and three percent tariff on food wheat. There is no enforceable mechanism, however, preventing an integrator from transforming wheat imported as food wheat (by import declaration) into feed ingredients. Thus, large integrators have access to facilities that effectively lower the cost of producing their own mixed feeds.

As a tax incentive for locating industrial operations outside Metro-Manila, in line with the government program of developing identified regional industrial centers (RICs), firms under the Board of Investments (BOI) incentives program are entitled to enjoy the privileges of duty-free import of breeding stock and capital equipment.

The large integrators are said to pass on part of their benefits from these differential tariff privileges-on grandparent and parent stock of day-old chicks and on imported feed ingredients-to their fee contract growers, to lure them away from competing integrators. This would be reflected in higher base fees per unit output than would be the case in the absence of such privileges.

On the other side of the fence, it is asserted that the large independent commercial broiler farms, together with the smaller sized local feed-milling firms that engage in price contracts with smallholder producers, do not individually have sufficient scale to effectively take advantage of formal access to MAV imports of corn or duty-free imports of grandparent and parent stock for day-old chick production. Although large independent producers may obtain wholesale price discounts from the large integrators for day-old chicks and high-quality commercial feeds, they cannot obtain them at par with the prices internal to the integrators in their conduct of fee contract growing operations.

Feed and credit subsidy to price contract growers of feed-milling and multipurpose cooperatives. Cooperatives, by legislation, are entitled to exemption from the corporate income tax. Cooperatives are perceived as generating social good in being of assistance to small and asset-poor producers. The corporate tax is levied at a uniform 32 percent on corporate net income or before-tax profit.

Most of the price contracts with smallholders were engaged in by local feed millers with no grandparent/parent stock breeding operations. Inputs are provided on credit to contract growers at market prices, plus some mark-up. The contract growers rely on the prevailing market prices of output at harvest. The integrators deduct from total sales the value of all intermediate inputs advanced on credit. Other integrators, however, charge the grower interest on the inputs advanced at prevailing bank rates for short-term loans. The integrator and contract grower then split net income 50-50. The scheme adds trust and reputation to transactions within a market system that is otherwise similar to that of the independents.

8.1.3 Determinants of profit inefficiency in broiler production

Profit frontiers were estimated where farm profit per kilogram of output was regressed on output and input prices, fixed factors (resources), interaction of fixed factors and prices, and on two variables capturing technology (the feed conversion ratio (FCR) and the animal mortality rate). With the frontier determined, the efficiency level of each farm was established, and the effects of the variables hypothesized to affect farm efficiency were simultaneously estimated.

Effects of prices, fixed factors, and technology on profit per unit output

Pooled sample. Table 8.5 shows the results of the econometric estimation of the profit frontier and the magnitude of effects of hypothesized determinants of inefficiency for the pooled sample. The salient results are below:

• In the estimation of the profit frontier, the price of output (broilers) and all input prices were significant in determining the level of profit per kilogram of output. All exhibit the expected directional sign.

• Only family labor on a per unit output basis is a significant fixed factor in explaining the variations in profit per unit, and it carries the expected positive sign.

• The feed conversion ratio (FCR), the parameter for neutral technical change, has the expected negative sign.

• The interaction of family labor per unit output and the wage rate is significant, and it carries the expected negative sign when the opportunity cost of labor is considered, even if uncosted family labor is used.

• The interaction between feed price and interest rate is positive.

• All things being equal, average profit increases significantly when one has a fee contract in broiler production.

Table 8.5 Profit frontier estimates in broiler production, with and without contracts, Philippines, 2002

 EXPLANATORY VARIABLES Pooled Independent and price contracts Contract fee Small- holders Large/Commercial Constant -4.535 21.896 4.790 -5.125 -31.514 Broiler price 5.551 4.849 - 4.57 8.616 DOC price ns ns - ns -3.139 Feed price ns -5.120 - 0.60 -3.167 Wage rate -3.527 -5.502 -0.620 -2.06 3.403 Lowest interest rate ns -6.265 ns 0.10 ns Buildings and eqpt/Q ns -4.700 ns ns ns Family labor/Q 1.741 2.057 0.334 1.075 -1.702 Land/Q ns -0.122 ns 0.012 ns FCR ns ns -0.593 -0.793 -0.397 Family labor x wage/Q -0.775 -0.924 -0.162 -0.457 0.705 Buildings and eqmt/Q x wage rate ns 1.658 ns 0.156 ns Buildings and eqmt/Q x inter rate ns 0.954 ns -0.34 ns Feed price x inter rate ns 4.232 - -0.18 ns Fee contract Dummy 23.007 ns - 17.64 17.618 Mean efficiency 0.578 0.506 0.839 0.595 0.827 log likelihood function -69.06 -55.079 18.060 -36.85 21.370

Source: UPLB-IFPRI LI Field Survey, 2002-03.

Production arrangement effects: independent versus contract growers. Table 8.1.5 also shows the comparison of frontier estimation results between independents and contract growers. The following are the salient results:

• Relative to the pooled results, while the number of significant explanatory variables increased in the group of independent producers, the number drastically fell for fee contract growers.

• While the response of independent producers to output price, feed price, and wage rate became stronger relative to the pooled sample results, only the wage rate remained a significant price variable for the fee contract growers, and with a far smaller absolute coefficient size than that of independents.

• Family labor and its interaction with the wage rate is significant to both independent and contract growers, but the response is again stronger among independents.

• While the FCR and animal mortality rate are significant among independents, they are surprisingly not significant for fee contractors, for whom these performance measures are in fact explicit incentive and penalty instruments applied by the large integrators.

• For independents, certain coefficients have perverse coefficient signs (interest rate, value of buildings and equipment per unit of output).

• The main difference between the two groups is the strong price-responsiveness of independents in the determination of profits, in contrast to the fee contract growers' almost exclusive focus on the allocation of family labor and hired labor in their profit-maximizing decisions.

Scale effects: smallholders versus commercial producers. Table 8.5 also presents the differential response of profit per kilogram to prices and fixed factors by smallholders and commercial producers. The following are the salient results:

• In general, for commercial producers, profit was far more responsive to changes in prices and fixed factors than for smallholders. For smallholders, the number of insignificant variables was greater than among commercials.

• Smallholder profits, however, were sensitive to the output price and the price of day-old chicks. Yet they were less responsive than those of the commercial farms.

• For commercial producers, the FCR and animal mortality rates were significant in explaining variations in profit, with their respective expected signs. The same measures of farm performance were not significant among smallholders.

• All things equal, regardless of size, engagement in fee contracts increased average profits, with commercial fee contracts gaining higher profit per unit as opposed to independents.

• Taking into account the scale of operations, profit per kilogram output was predicted to be higher on farms that had a fee contract with a large integrator compared to those with a fee contract with a small feed-milling firm.

In summary, smallholder profit per kilogram of output was determined by the price of broilers and day-old chicks and by investments in facilities and equipment, apart from whether or not one gets a fee contract. In contrast, commercial broiler producers' profits are most sensitive to changes in output and input prices.

Farm inefficiency and its determinants

The hypothesized determinants of farm technical/allocative inefficiency pertained to differential attributes of the decision-makers of the business. These embody differential human capital; differential access to markets, services, and information; and differential behavior in the mitigation of environmental externalities.

The pooled sample. Table 8.6 shows for the pooled sample the effect of the factors hypothesized as influencing farm inefficiency. The salient results are as follows:

• None of the decision-maker human capital attributes were significant in explaining profit inefficiency.

• Better access to feeds credit and to information/communications infrastructure was significant in reducing farm-level inefficiency.

• Environmental externality mitigating behavior was not significant in changing the level of inefficiency of the farm.

Table 8.6 Determinants of inefficiency in broiler production, Philippines, 2002

 EXPLANATORY VARIABLES Pooled Independent and price contracts Contract fee Small- holders Large/Commercial Constant ns ns ns ns 2.445 Age ns ns ns ns -0.842 Educ ns ns ns 2.512 -2.058 Experience in poultry ns ns ns ns 0.660 Formal/informal training ns ns ns ns -1.453 Picked up services by buyer ns ns ns ns 0.245 Has phone services (Info) -1.535 ns ns -5.38 0.245 Environmental capture ns ns ns 0.234 -0.237 Feeds credit dummy -3.782 -3.132 - -6.39 -2.340

Source: UPLB-IFPRI LI Field Survey, 2002-03.

Production arrangement effects: independents versus contract producers. Comparing the results on determinants of profit inefficiency (Table 8.6) between independents and fee contract growers, the following are observable:

• Among the fee contract producers, only the environmental capture variable was statistically significant in affecting farm efficiency. For fee contractors, the attempt to capture externalities improves efficiency.

• For the contract producers, access to feeds credit is a non-issue since access comes automatically with the engagement in a contract growing arrangement. In contrast, access to such a service is all-important to independents.

Scale effects: smallholders versus commercial producers. Table 8.6 also shows the results on the determinants of technical/allocative efficiency between smallholders and commercial producers. The salient results are below:

• Within the group of commercial broiler producers, human capital attributes of education and training, as well as access to feeds credit, were statistically significant as determinants of farm inefficiency.

• Access to feeds credit and information/communications infrastructure were strongly significant in reducing profit inefficiency.

• An unexpected coefficient sign was exhibited in the age variable.

• The education variable has an unexpected positive sign. The perverse effect of education on profit efficiency for smallholder farms may reflect the intervention of opportunity cost of educated human capital, particularly when the scale of operations of the broiler business is relatively small. In such a case, the attention of more educated human capital may be diverted from alternative primary or secondary economic undertakings.

Farm efficiencies of classes of broiler production, when family labor is not costedMean efficiency for the pooled sample was 0.58. Relative to this mean, efficiency levels varied among and within the producer classes. The following are the major results:

• Fee contract producers, both smallholders and large contract growers, were on average most efficient, with mean efficiency of 0.84. This class of producers also generated the highest profit per kilogram of output.

• Smallholders were the second most efficient group, with mean efficiency about 20 percent off that of the best producers (mean efficiency = 0.67). The smallholder sample, however, included 23 fee contract growers who belonged to the most efficient group.

• Commercial producers came in third with a mean efficiency of 0.61, roughly the average of the pooled sample and close to 10 percent off the levels achieved by the smallholders as a group.

• Least efficient were the classes of independents and price contractors. These groups' efficiency (mean efficiency = 0.51) was some 40 percent lower than that of the most efficient group. These classes of producers, including both large and small, faced the largest price risks on both the input and output sides. Their efficiency levels were thus strongly affected by their ability or inability to get feeds on credit.

## 8.2 Hog Farms

8.2.1 Comparative profit performance

Profits by scale and production arrangement

Table 8.7 summarizes profits per kilogram of output and average farm profits by scale of production, with and without contracts. Observations include the following:

• Profit per kilogram of output is highest among independent smallholders, but this group realized the second lowest overall average farm profits.

• Profit per kilogram of output is least among contract growers, with small contract growers earning the lowest farm profits. Large contract growers, however, generated the second largest farm profits. Small margins per unit output were compensated by relatively large scale of operations.

• Profit per kilogram of output was about even between medium- and large-scale independents. The difference in total farm profits arises from the difference in scale.

Table 8.7 Profit per kilogram of output and total farm profits, Hog Production Philippines, 2002

 ENTRY SMALLHOLDER (<100 heads) COMMERCIAL (100 heads and above) Independent Contract Medium Independent Large Independent Contract (PhP/kg) 26.69 2.08 19.43 20.37 2.33 Farm profits (PhP) 15,942 6,732 114,332 497,923 143,044

Source: UPLB-IFPRI LI Field Survey, 2002-03

Profit per kilogram and scale: by quintile

Figure 8.5 shows the behavior of profit per unit output by size quintile.

Figure 8.5 Profit per kilogram of output by size quintile, Hog Production, Philippines, 2002

The quintile distribution does not distinguish between independent producers (which generally earned higher profit per unit) and contract growers (which had uniformly lower profit margins). The following results are of interest:

• Profit per unit was highest (38.8 pesos) among the smallest quintile, with its mean output of 261 kilograms. This group would involve mainly the farrow-to-weaning (Type 1) independent smallholders.

• Profit per unit falls abruptly in the second quintile (mean profit = 14.5 pesos per kg), then continues to gradually fall through to the fourth quintile.

• The second quintile still involves independent smallholders, mainly those engaged in farrow-to-finish (Type 2) and combined output (Type 4) operations.

• The quintiles with the lowest profit per unit, the third and the fourth, involve the smallholder contract growers and the independent medium-scale commercial farms.

• The last quintile includes mainly the independent large-scale commercial producers and the large-scale contract growers. Profit per unit output is second highest in this group, at 18.5 pesos.

• While the smallest quintile (the smallest scale) of farms registered the highest profit per kilogram of output of all size categories, it cannot be said that beyond the first quintile, smallholder profits per unit are still higher than those of large farms. What can be stated is that, in general, smallholder farm profit per unit is higher than that of medium-sized commercial farms (up to the fourth quintile).

• The second statement that could be made with respect to profit per unit and scale is that, except for the smallest size quintile, large farms' profit per unit output is higher than that of smallholders and of medium-sized commercial farms. The distinct position of the very large farms (fifth quintile) points to market and technological advantages of operating on a sufficiently large scale, whether one is operating as an independent firm or through a fee contract with a large integrator.

8.2.2 Accounting for market distortions

Adjustments for cost of family labor

Figure 8.6 shows the behavior of profit per kilogram of output by size quintile, after the costing of family labor. The salient results are the following:

• As expected, adjusting for cost of family labor affects smallholder profit per unit output most strongly. The absolute and relative change (reduction) in smallholder profit per unit is not large.

• For the first size quintile, profit per kilogram falls from 38.8 pesos to 38.3 pesos. For the second quintile, profit per kilogram falls by even less, from 14.5 pesos to 14.4 pesos.

• As expected, per unit profit of the medium- and large-scale commercial farms are hardly affected by adjustments for the cost of family labor.

• The adjustment for the cost of family labor does not change the relative positions of farms according to scale.

• Unexpected is the relatively small impact of costing of family labor on smallholder farms. This reveals the small relative share of labor costs in pig production, even when labor is hired (e.g., labor costs amount to 5% among medium-sized commercial farms). It also emphasizes the dominant position in the cost structure of expenditures for feeds and growing stock.

Figure 8.6 Average profit per kilogram after costing family labor, by quintile, Hog Production, Philippines 2002

Adjustments for the cost of environmental externalities

As in broiler production, adjustments for the cost of internalizing environmental externalities from animal wastes and mortalities were measured in terms of the value of manure sold or used on own farm, evaluated at market prices, and the cost of disposing of non-marketed or unused waste and animal carcasses. These were aggregated to comprise a measure of environmental abatement or mitigation expenditures per unit of output. The assumptions used for the validity of this measure are identical to those discussed in the broiler section.

Comparative expenditures on mitigation of environmental externalities. Table 8.8 presents the expenditures on environmental capture of externalities from waste and dead animals by smallholders and commercial producers of hogs, with and without contracts. The following are observed:

• Independent producers spent more on environmental mitigation per unit of output than contract growers, regardless of scale.

• Among independents, smallholders spent two to three times more than the large producers.

• Small independents spent a factor of 14 more than small contract growers.

• Medium and large independents spent more than large contract growers by a factor of four and seven, respectively.

Table 8.8 Expenditure on mitigation of environmental externalities in hog production, by scale, with and without contracts, Philippines 2002

 ENTRY SMALLHOLDER (<100 heads) COMMERCIAL (100 heads and above) Independent Contract Medium Independent Large Independent Contract Expenditure per 100 kg output 57.3 4.1 17.1 27 4.1

Source: UPLB-IFPRI LI Field Survey, 2002-03

It does appear that independent producers tend to capture larger amounts of externalities from waste and dead animals than contract producers. It also appears that among independent producers, there are economies of scale in dealing with environmental capture.

Determinants of farm expenditure on mitigation of environmental externalities. Tobit regression was run to determine the differential effects of household, farm, and locational characteristics on mitigation expenditures by farm scale and production arrangement. The dependent variable is the total mitigation cost (pesos per kilogram of output), which is the sum of the following elements: the value of manure sold; the value of manure used on the farm; expenditures for disposal of dead animals; compensation to neighbors; transport costs associated with waste disposal; costs for licenses, taxes, and permits related to pollution; and depreciation and maintenance costs of waste treatment and control facilities. The explanatory variables include age and education of the household head or decision-maker (in years), the wage rate (pesos per hour), number of animal mortalities in the last cycle, distance to nearest residential community (in kilometers), size of inventory, feed conversion ratio (FCR), connection to piped-in water supply, and dummy variables for land classification, existence of croplands, location within Laguna Lake Development Authority (LLDA) jurisdiction, ability to sell manure in the last two months, and region (Bukidnon = 1).

Table 8.9 shows the results of the Tobit runs for the pooled sample. The important determinants of mitigation expenditure turned out to be three: existence of croplands, location within LLDA jurisdiction, and distance to the nearest residential community.

Table 8.9 Determinants of farm expenditure on mitigation of environmental externalities from hog production, Philippines, 2002.

 EXPLANATORY VARIABLES Pooled Sample N=203 All Indep. Farms n=152 All Contract Farms n=51 All Large Farms n=95 Large Indep. n=66 Small Indep n=86 All Small Farms n=108 Coefficient Coefficient Coefficient Coefficient Coefficient Coefficient Coefficient Education of HH head(yrs) ns ns ns 0.017 0.028 ns ns Age of HH head (yrs) ns ns ns ns ns ns ns Able to sell manure in last2 mons. (Dummy) ns ns 0.104 ns ns 3.678 3.695 Land class is Agricultural(Dummy) ns ns 0.060 ns ns ns ns Has crop land (Dummy) 0.501 0.578 ns ns ns ns 0.917 Within LLDA jurisdiction(Dummy) 1.342 1.364 0.074 0.283 0.431 ns ns Wage rate (PhP per hr) ns ns 0.006 ns ns ns ns No. of mortalities in last cycle (heads) ns ns ns ns ns ns ns Distance to nearest Residential comm. (Kms) 0.010 ns 0.001 ns -0.038 ns ns Connection to piped-in water supply ns ns 0.078 0.118 0.170 ns ns FCR-FE ns ns ns 0.126 0.109 ns ns Size of inventory (no. of heads) ns ns ns ns ns ns ns Bukidnon (dummy) ns ns ns 0.343 0.382 ns ns Constant ns ns -0.143 -0.944 -1.070 ns ns Log pseudo-likelihood -345.96 -279.36 33.01 -34.80 -26.50 -160.35 -191.23

Source: UPLB-IFPRI LI Field Survey, 2002-03.

The coefficient of the croplands dummy and the LLDA jurisdiction dummy carried the expected positive sign (as shifters of the intercept). Distance to the nearest residential community exhibited the sign opposite to that expected; the expectation was that farm proximity to population centers would prompt greater efforts to mitigate adverse environmental effects.

Regarding the effects of production arrangements, Table 8.9 also shows the Tobit results for the environmental capture regressions for all independents and all contract farms. The salient results are three:

• Only operation of croplands and location within LLDA jurisdiction mattered for independent producers, while six explanatory variables emerged as significant among contract farms.

• For contract farms, the ability to sell manure, land classification as agricultural, and location within LLDA jurisdiction carried the expected positive coefficient sign.

• The coefficient signs for wage rate, distance to nearest residential community, and connection to piped-in water supply carried the opposite signs than expected.

Regarding scale effects, the Tobit results for large and small farms also yielded three main observations:

• The significant determinants of mitigation expenditures differ entirely between scales of operations.

• For large farms, location within LLDA jurisdiction, feed conversion ratio (FCR) performance, and location in Bukidnon were significant. The coefficients carried the expected directorial relation.

• For small farms, the ability to sell manure and the operation of croplands were significant with the coefficients carrying the expected positive sign.

Looking at scale effects among independent farms, on average, independent hog farms were found to spend more per unit of output on environmental mitigation than contract farms. As a group, however, the cropland dummy was the only significant factor that appeared to explain the independent farms' mitigation expenditure behavior. However, small and large independent farms may respond differently to the factors hypothesized to explain environmental capture behavior.

Table 8.9 shows the Tobit results for small and large independents. The main results are the following:

• For small independent farms only the ability to sell manure remains as significant.

• For large independents, six explanatory variables emerge as significant.

• Education (years of schooling) carries the expected positive sign of coefficient.

• Location within LLDA jurisdiction and location in Bukidnon carry the expected positive coefficient sign.

• Distance to the nearest residential community has the expected negative sign of the coefficient.

• FCR performance carries a positive coefficient sign, as expected.

• Connection to piped-in water supply emerges as significant but with a positive sign of the coefficient.

Determinants of environmental mitigation behavior that turned out to be non-significant. Some of the hypothesized determinants of environmental mitigation behavior in hog production turned out to be non-significant. These fell into two groups. The first group is made up of the age of the farm decision-maker, the scale of operations (inventory size), and the number of animal mortalities. Although appearing only among medium and large independent farms, the level of education is more important than age, as such, in influencing environmentally relevant behavior. Scale had no systematic continuous effect on environmental mitigation, although small independent farms spent more than large producers on environmental mitigation per unit of output. Neither did the absolute number of animal mortalities on the farm appear to influence the level of environmental mitigation. This may simply imply that farms do not deal with the disposal of carcasses as an activity distinct and separable (as an environmental mitigation activity) from their normal production operations. Thus, only very slight variation that could be captured in explicit expenditures related to disposal of dead animals, regardless of scale.

The second group of non-significant factors is comprised by the wage rate and the official land classification (e.g., agricultural, residential, industrial). These determinants were insignificant to all except contract growers. Even for the contract group, the directional relation was opposite that hypothesized. It appears that the official classification of lands where hog farms are situated is irrelevant to hog farms inasfar as environmental mitigation behavior is concerned. The fact that a farm is situated on 'residential' or 'industrial' (non-agricultural) lands is not an automatic guarantee that the farmer will engage more intensively in environmentally friendly behavior. The mere classification of land as non-agricultural, without accompanying institutional pressures to define and enforce what pollutive activities are permitted, thus does not generate environmental mitigation activities on the part of farms.

Similarly, the wage rate appears to be non-significant in decisions on environmental mitigation expenditures, despite the hypothesis that the disposal and management of hog waste and dead animals is a labor-intensive activity. Perhaps this implies that factors other than the wage rate influence environmentally friendly behavior more.

Perverse responses in hypothesized determinants. Some perverse responses were observed in the hypothesized determinants of environmental mitigation behavior. First in this category are distance to residential areas and access to piped-in water supply. Distance to residential communities came out as significant only in the pooled sample and on contract farms and was insignificant for the others. Thus, the distance factor appears to work only with contract growers, but in a perverse way. Among contract farms, large-scale operations are, on average, farther away from residential centers than their smaller counterparts. But their mean expenditures on mitigation per kilogram of output are the same. Large fee contract farms are in fact governed by regulations on location requiring some minimum distance from other farms to control incidences of diseases. Moreover, incentives are provided (e.g., cost sharing) for spending on waste disposal facilities. Since the distance factor was not significant among small farms (including the small contract farms), the distance factor must be working on large contract farms. But the large contract farms are already, on average, farther away from residential communities. This distance variable may in fact be capturing something else, with an opposite effect on the environmental mitigation behavior than was hypothesized for pure distance.

The perverse effect of access to piped-in water supply applies to large farms and contract farms, but was non-significant among small farms. Access to piped-in water was hypothesized as allowing for easier, less costly disposal of hog waste from the pens to the nearest body of water through crude canals and gutters. The results, however, do not support the hypothesis. For large and contract producers, it appears that when access to piped-in water is present, they tend to make greater expenditures in controlling waste. Possibly, access to piped-in water enables them to build waste treatment facilities more easily, thus raising expenditures on environmental mitigation, all things being equal. Where there was no access to piped-in water, large farms and contract farms may try to minimize the flow of waste by non-capital-intensive means, for which costs are not explicit and thus were not captured by the survey instrument.

Conclusions on determinants of environmental capture on hog farms. From this discussion, a number of conclusions can be drawn regarding determinants of environmental capture on hog farms:

• While independent farms on average spend more per unit output on environmental mitigation than contract farms, independent farms cannot be lumped together in extracting the determinants of environmental capture particular to them.

• While contract farms in general spend far less than independents on environmental mitigation, contract farms respond significantly positively to market opportunities for manure sales when they exist and to using manure on crops when they operate croplands. They also respond to the existence of an anti-pollution authority that imposes fines, penalties, and other sanctions.

• Contract growers, however, strangely behave in a perverse way to proximity of operations to residential communities. Those located farther out spend more than those that are closer, although the coefficient size is small (0.001).

• Contract growers also behave in the opposite directions to what was hypothesized with respect to connection to piped-in water supply. Casual observation showed that connection to piped-in water facilitates disposal of hog waste from the pens to the nearest body of water through canals. The absence of piped-in water was expected to make disposal of hog waste a more labor-intensive activity. Possibly, however, connection to a supply of piped-in water allowed for means of dealing with and treating waste other than simply flushing it out of the pens and off the farms, particularly among larger producers.

• The segregation in the analysis of large independents from small independents was quite fruitful in that it enabled the capture of distinct behavioral patterns. On the one hand, small independents, on average, spent more on environmental mitigation per unit of output. The rationale for this behavior cannot be explained by the various factors that were thought to influence such behavior. Only a single factor-the opportunity to market manure-was significant, and the coefficient was relatively large (3.7 pesos per kg of output when sales were realized).

• Also notable is that when all smallholder farms are considered, the operation of croplands was also significant and the coefficient was relatively large (i.e., an additional 0.9 pesos per kg where croplands exist). With the removal of the contract growers from the smallholders group, the cropland dummy became insignificant. This may reflect independent smallholders' lack of access to croplands.

• The segregation of large independent farms from the pooled sample and from the all-large-farms sample highlighted the distinct behaviors of independent commercial farms (medium and large). As readily identified and easily targeted entities, they appear to respond more sensitively to legal authorities and to actual and potential community (residential) pressure. The emergence of the education of the household head or farm decision-maker as significant appears to strengthen that greater responsiveness and sensitivity.

• Among large independent farms, technology, as embodied in the FCR, appears to be an important vehicle for influencing the amount of environmental capture, all things being equal. The more feed-efficient the entire activity is, the less the resulting waste should be in pure physical terms. This should hold for all farms, regardless of size and production arrangement, even if statistical significance in the sample applies only to large farms as a whole, and large independent farms in particular.

• The significance of the Bukidnon locational dummy for large farms may simply capture the effect proper of the cropland dummy. In Bukidnon relatively extensive use of pig manure on orchards was observed among the independent commercial farms.

Policy implications. A few policy implications follow from the analyses described above. First, both market-based instruments and institutional pressures work in inducing environmentally relevant behavior. These two instruments, however, work differentially on smallholders and large producers. Smaller farms tend to respond to opportunities to sell manure or make use of it as fertilizer on their own farms and crops. Larger farms respond to the existence of an institution that makes and effectively enforces rules against environmental pollution.

Such formal institutional instruments work on large or commercial entities possibly because they are more easily targeted for the imposition of sanctions (as permit sources) and they have the size and capacity to invest in facilities and structures to minimize the potential magnitude of externalities on others. Moreover, the education of decision-makers appears to function positively for large producers. Likewise, the feed efficiency factor comes into play. This points to the potential of a technological-institutional approach to minimizing environmental pollution from commercial hog farms. If the technologies are developed, and are commercially supplied, when decision-makers are well informed and the rules are made clear in matters of environmental pollution, commercial producers are likely to respond by spending more per kilogram of output on environmental abatement.

Smallholders, on the other hand, will experience an institutional solution, such as enforcement of pollution standards, as an additional burden that increases their production costs. Nonetheless, they will respond if there are opportunities for them to benefit from the disposition of livestock waste, whether through the sale of manure or by its use on croplands. Since no functioning market for hog manure exists of the stature of the domestic chicken manure market, the policy direction suggested is, at the very least, the development of a feasible market for processed hog manure alongside increased access by smallholders to croplands. As with the broiler sample, hog-raising smallholders have the least access to manure markets even where they exist, and smallholders also tend to have very little assets in terms of agricultural cropland holdings.

Scale and environmental mitigation

Figure 8.7 shows the distribution of mitigation expenditures by scale of output. Farm expenditures per kilogram of output are higher among smallholders than among large producers. Those in the first quintile spend the most per kilogram of output, about 40 percent more than those in the second quintile and about eight times more than the third, fourth, and fifth quintiles.

Figure 8.7 Environmental mitigation cost by quintiles, Hog Production, Philippines, 2002

There appear to be economies of scale in environmental mitigation, as exhibited in the behavior of average cost of environmental mitigation, although average cost tends to remain constant after the third quintile. This can be understood in the context of the focus of expenditures on environmental mitigation by large commercial farms: in waste treatment facilities that are characterized as lumpy investments. The resort to such facilities, however, is viable only with sufficient scale of output.

Smallholders, particularly those with very small inventories and output per cycle must deal with the disposal of waste and dead animals using more labor-intensive methods. Where labor is not scarce, the smallholder would still end up spending more on environmental mitigation per unit of output by virtue of their smallness in scale.

Conclusions on scale and environmental mitigation. The analyses on scale and environmental mitigation show that small producers spend more per unit of output on environmental mitigation than large producers. But does this finding lead to the conclusion that smallholders pollute less than commercial (large-scale) producers?

Assuming that a unit of output (e.g., 100 kg) generates, on average, the same amount of physical waste and environmental cost, smallholders pollute less than large producers per unit of output under the condition that the value of environmental damage (cost) reduced per unit expenditure on environmental mitigation is the same. Under conditions where cost-effectiveness enters into the equation, and its direction with respect to scale is biased in favor of large-scale producers (third, fourth, and fifth quintiles), then smallholders still end up mitigating larger amounts of environmental cost per unit of output if the technologies of the large producers are only, at best, three times more cost-effective than the technologies or waste management and disposal practices that smallholders currently use. The already relatively high mitigation costs per unit of output internally borne by smallholders should warn against a strict and purely institutionalized approach to reducing environmental externalities from wastes in hog production.

The imposition of waste-treatment facilities (e.g. biogas digesters) on small-scale hog producers would further jack up their per unit expenditures on environmental mitigation, with direct adverse impacts on profits per unit output. This arises mainly from the lumpiness of such investments. A pure and strict institutional solution to environmental problems from hog waste would hit hardest operations where scale is small and constraints to expansion exist. The large producers at least have the scale to cushion the impact of large investments, other things being equal.

Trade and social subsidies to integrators and cooperatives

Differential access to credit. Table 8.10 shows the credit behavior of hog farmers. Differential access to formal credit is revealed in the lower incidence of borrowing for production purposes by smallholders (4-10%) compared to the group of large-scale farms (24-27%), as discussed in Chapter 5. This is reinforced by the relatively high percentage of smallholders who do not attempt to borrow from a bank because they perceive their chances of being granted a loan as being low (31%). Fewer commercial producers (19%) thought they would be ineligible for a bank loan.

The interest rates paid for loans averaged 17.6 percent per annum. There is no systematic pattern in the behavior of interest rates paid, except that among all groups the large independent producers obtained the lowest interest rate, at 15.6 percent per annum, while independent smallholders paid the highest average rate, at 18.2 percent.

Table 8.10 Borrowing experiences and perceptions on credit by hog farmers, Hog Production, Philippines, 2002.

 ENTRY SMALLHOLDER (<100 heads) COMMERCIAL (100 heads and above) Contract Average Independent Contract Medium Independent Large Independent Percent of farms that actually borrowed (%) 11.5 4.3 8.0 23.5 26.7 13.0 Average interest rate paid (% p.a.) 18.2 16.0 17.5 15.6 18.0 17.6 Percent of farms who believe they can get loan at 12% p.a., or less (%) 48.3 8.7 50.0 29.4 30.0 42.0

Source: UPLB-IFPRI LI Field Survey, 2002-03

There is no systematic pattern in which group has greater access to subsidized credit (12% rate or lower). Around half of the independent smallholders and medium-scale commercials believe they could access loans at that rate. Yet only about 30 percent of large commercial farms (independent and contract) perceived they could get those preferential rates.

Except for the smallholder contract growers, whose access to subsidized rates was also restricted, it appears that large-scale farms have less access to subsidized credit than small- and medium-scale farms. However, this distribution may be depicting the differential access to credit from sources other than banks, since credit sources such as friends and family might charge interest rates lower than 12 percent. Quite possibly, smallholders have greater access to informal sources of loans at lower than bank rates.

Subsidy on feeds and breeding stock to contract growers. The assertions made in the broiler section on the advantages conferred to large integrators due to their large stature apply to hog production as well. These advantages include access to minimum access volume (MAV) imports of corn at lower than regular tariff rates (35% vs. 60%); access to imported food wheat at tariff rates lower than feed wheat (3% vs. 10%); and access to duty-free imported breeding stock as a Board of Investment (BOI) incentive for locating operations in regions outside Metro-Manila.

The large integrators are said to pass on part of the benefits of such privileges to their commercial fee contract growers. These take the form of a higher fixed fee base per kilogram of output than could be the case had the advantages not been enjoyed.

Feed and credit subsidy to smallholder price contract growers of feed-milling and multipurpose cooperatives. The assertions in the broiler section on the advantages of feed-milling and multipurpose cooperatives engaged in livestock contract growing generally applies to hog production as well. This advantage arises from the cooperatives' exemption from the corporate income (profits) tax, levied on corporations and business entities at 32 percent of net income.

An example of a multipurpose feed-milling cooperative engaged in hog contract growing that enjoys such a privilege is the Soro-Soro Ibaba Development Cooperative (SIDC) in Batangas City in Southern Luzon. As of 2002, it had 322 contract growers, 287 of which were grow-to-finish hog raisers, 265 broiler raisers, and 9 farrow-to-weaning hog farmers. The SIDC incurred a net loss of 9.6 million pesos in 2002 due to very high production costs. Therefore, to illustrate the potential magnitude of the cooperative's implicit subsidy, SIDC's net income in 2000 is used (Table 8.11).

The cooperative earned a total net income of 33.9 million pesos in 2000, about 75 percent of which was generated by its main activity, feed milling. Its contract growing operation was only fourth in the order of net income contribution, accounting for only about three percent of cooperative net income. If the regular corporate tax rate of 32 percent were imposed on the total net income of 33.9 million pesos, then the tax bill would have been 10.9 million pesos in 2000. This is the equivalent implicit subsidy (tax exemption) which corporate feed-milling firms of similar size engaging (or not engaging) in contract growing do not enjoy. The magnitude of this implicit subsidy (10.9 million pesos) is even larger than the absolute contribution of the contract growing operations of the cooperative to overall net income (1.0 million pesos).

Table 8.11 Net income and implicit subsidy on SIDC feedmilling, contract growing, and other operations, 2000

 Activity/Source Net Income (PhP Million) Percent Share of Net Income (%) Implicit Subsidy (PhP Million) Feed milling 25.3 74.6 8.10 Mini-mart 7.7 22.7 2.46 Expanded credit line 1.1 3.2 0.35 Contract growing 1.0 2.9 0.32 TOTAL NET INCOME 33.9 10.85

Source: Annual Report, SIDC, 2001.

The SIDC considers its contract growing operations to be a social program to build goodwill among the nearby communities. The apparent 'insignificance' of the contribution of the contract growing to total net income gives the cooperative leeway to 'subsidize' this activity from its larger and more important profit centers.

8.2.3 Determinants of technical/allocative efficiency

As with broiler farms, profit frontiers were estimated for hog production. With the frontier determined, the efficiency level of the farm classes was established and the effects of the variables hypothesized to influence farm efficiency estimated.

Effects of prices, fixed factors, and technology

Table 8.12 presents the estimated profit frontier and determinants of technical/allocative efficiency among hog farms. For the pooled sample, the following are the salient points:

• Prices of outputs-slaughter hogs and piglets-were both positively signed. Considering all activities, profit per unit of output responded positively to output prices.

• Profit per unit output responded negatively to prices of major intermediate inputs-feeds in all activities and weanlings for activities that require the purchase of growing stock.

• Neither the wage rate nor the interest rate was significant in influencing the direction of profit per unit output.

• None of the fixed factors, or their interactions with prices, was significant in influencing the direction of profits per unit output.

• Neither the feed conversion ratio (FCR) nor mortality rate was significant in affecting the direction of profit per unit.

• Of the activity types that deviated from the standard farrow-to-finish (Type 2) activity, two were significant: engagement in grow-to-finish (Type 3) and engagement in combined output of piglets and slaughter hogs (Type 4). All things being equal, engagement in Type 3 and Type 4 activities tended to reduce mean profit per unit output. Engagement in grow-to-finish is embodied in the contract dummy.

Table 8.12 Determinants of profit and inefficiency, Hog production, Philippines, 2002

 EXPLANATORY VARIABLES Pooled (N=207) All Independents (N=154) All Contracts (N=53) All Small-holders (N=110) All Large-scale (N=97) Constant 3.564 ns 2.412 ns ns Price of hogs 0.685 1.233 1.08 0.789 1.506 Price of piglets 0.727 0.738 - 1.110 -0.495 Price of weanlings -0.119 -0.119 -0.266 -0.116 - Price of feed -0.687 -0.718 -0.814 -0.696 -0.544 Contract dummy -0.405 - - -0.373 3.227 Wage rate ns ns ns ns ns Free board and lodging dummy ns ns ns -0.106 ns Lowest interest rate ns ns ns ns ns Value of inventory/Q ns ns ns 0.018 ns Buildings and eqpt /Q ns ns ns ns ns Family labor /Q ns ns ns -0.379 ns Farmsize (in ha)/Q ns ns ns ns ns FCR ns ns ns ns -0.115 mortality rate ns ns -0.0178 ns ns Family labor x wage rate ns ns ns ns ns Buildings and eqpt /Q x lowest I.r. ns ns ns ns ns farrow-wean dummy ns 2.315 - ns 8.695 combined farrow-wean and farrow-finish -2.986 -3.012 - -4.599 2.555 Very large dummy - - - - ns Constant -6.594 -1.978 ns ns 2.197 Age ns ns ns ns ns Experience in hog business ns -0.426 0.102 ns ns Education ns ns ns ns ns Formal/informal training 2.052 2.833 ns 1.799 ns Feeds credit dummy ns 1.086 0.271 2.220 ns Capital credit dummy -1.448 -0.986 -0.161 -1.945 ns Information -4.972 -3.078 0.702 -3.665 ns Buyer is viajero 4.12 2.08 - 2.857 ns Environmental cost ns -0.77 -0.106 -1.217 ns Fee contract dummy -1.391 - -0.747 - -1.838 Very large dummy - - - - Region dummy (Bukidnon=1) ns -1.41 ns ns -0.703 sigma squared 2.085 1.357 0.0127 1.583 0.203 gamma 0.994 0.992 0.7 0.997 0.984 Mean efficiency 0.802 0.781 0.746 0.802 0.812

Source: UPLB-IFPRI LI Field Survey, 2002-03

Effects of production arrangements: independent versus contract production.

• Profit per unit of output of both independents and contract growers responded positively to their respective output prices, with contract growers only producing finished slaughter hogs. Independent producers, however, were relatively more sensitive to changes in slaughter hog prices than contract growers as a whole. This stems from the guaranteed fee arrangements enjoyed by most of the large contract growers.

• Profit per unit in both groups responded negatively to prices of weanlings and feeds.

For independent producers, the two relevant activity type dummies were significant-farrow-to-weaning (Type 1) and dual output (Type 4). Other things being equal, average profit per unit was higher on farms engaging in farrow-to-weaning activity; but profit per unit was lower on farms engaging in Type 4 activity (combined operations).

Scale effects: smallholders versus large-scale producers.

• Profit per unit output of smallholders and large-scale producers responded positively to prices of slaughter hogs, but the response of large producers was twice as strong as that of smallholders. This may stem from commercial producers' treatment of the hog-raising enterprise as strictly business.

• While smallholders' profit per unit responded positively to prices of piglet output, the reverse was manifested for commercial producers. It was noted that piglet revenue is not as significant to total revenue among large producers as it is for smallholders. Piglets are the sole output of Type 1 smallholders. Very few commercial raisers are engaged in purely farrow-to-weaning operations. Particularly for large-scale producers, piglet production appears to be only an appendage to the main activity of farrow-to-finish, from which maximum farm profits can be derived by fattening the hogs to the optimal age or weight. If the sale of piglets is influenced by motives other than profit maximization, the perverse reaction of profit per unit to piglet prices can occur.

• Both groups responded negatively to the prices of intermediate inputs relevant to their respective activity types, with weanlings and feeds relevant to smallholders. But weanling prices were not relevant to large producers, whether engaged in independent production or in contract operations. Large contract growers were mainly engaged in fee contracts, with growing stock unpriced.

• For smallholders, the free board and lodging dummy for hired workers became relevant, and signed negatively as hypothesized. This may be capturing the effect that the wage rate was unable to capture. For smallholders, who rarely employ hired labor, the provision of free board and lodging to just one worker may already significantly impinge on profits due to the smallness of operational scale. For large-scale producers, free board and lodging privileges may not significantly and systematically affect profit per unit due to size of operations.

• For the smallholders group, two fixed factors (evaluated on a per unit output basis) became significant-value of inventory and family labor. While value of inventory is positively signed as hypothesized, family labor has the opposite sign (negative). For smallholders, it may not necessarily be the number of household members involved in the livestock activity (with various degrees of involvement), but rather the intensity and quality of involvement of household members other than the operator or decision-maker. The number of other family members involved in the livestock activity is a potential resource, but not a sufficient condition for increasing profit per unit output. At an extremely small scale of operations, additional family labor may be redundant, and to the extreme, counterproductive.

• For commercial farms, the parameter for neutral technical change, the FCR, was significant and signed negatively as hypothesized. The FCR is a parameter that is explicitly monitored on commercial farms. The significance of the parameter's effect on profit per unit may reflect the relatively more serious business approach to the enterprise among commercial-scale producers, whether independents or under fee contract.

• For the activity type dummies, two activities came out as significant for smallholders-Type 3 (contract dummy; grow-to-finish) and Type 4 (combination). Both have negative coefficients as in the pooled sample.

• For commercial farms, all three activity dummies were significant, and all are signed positively. Two results that for the commercial farms are different from the other groups are the positive coefficient signs of Type 3 activity (contract dummy; grow-to-finish) and Type 4 (output combination). While involvement in these two activities, other things being equal, resulted in lower mean profits per unit relative to the standard farrow-to-finish operations, the reverse happens among commercial farms.

• The contract dummy in the commercial farm sample refers virtually exclusively to fee contracts. The positive coefficient sign signifies that if Type 3 activity is done under fee contracts and undertaken in sufficiently large scale, mean profits per unit output are higher. It is to be noted the commercial fee contracts involve quite technologically advanced processes (superior genetic quality of stock, high feed quality, close veterinary supervision, performance-tied incentives, and penalty provisions).

• The positive coefficient sign of Type 4 activity for commercial-scale operations is a surprising result. It implies that under certain conditions, partial engagement in sale of piglets yields higher profit per unit output than strictly farrow-to-finish operations. This is quite plausible, as it gives rationale to the engagement in this activity by the commercial independents in our sample. If the commercial farm is into breeding stock of superior genetic characteristics, its weanlings will also command a premium price on the commercial market. Although intended for internal utilization, market conditions for weanlings may provide temporary windows of advantage for selling highly priced weanlings.

Determinants of technical/allocative efficiency

The pooled sample.

• Of the individual hog raiser's human capital characteristics, only training in the business in the last two years emerged as significant, but had the opposite sign to that hypothesized.

• Access to credit for capital and communications infrastructure came out as significant in reducing inefficiency.

• Engagement in a fee contract reduced inefficiency.

Production arrangement impacts: independent producers versus contract growers.

• More years of experience in the hog business reduced inefficiency among independent producers, although it was not significant for contract growers. Additional years of experience may imply mastery of the business among independents, who are left to themselves. However, for contract growers, mastery in the more recent technological advances may be more crucial in influencing efficiency within the group.

• For independents, recent training in hog production or in business, as in the pooled sample, had the opposite sign to that hypothesized. For contract growers, recent training in livestock or business reduced inefficiency. Quite possibly, there are qualitative differences in the type and intensity of training that the two groups undergo.

• Access to feeds credit had a perverse effect on efficiency among independent producers. This result is the reverse of that expected.

• For both independents and contract growers, access to capital credit improves efficiency, but the effect of access to capital on efficiency is stronger among independents than among contract growers.

• As in the pooled sample, for independents, access to information and communications infrastructure significantly reduces inefficiency.

• Consistent with the pooled sample, among independent producers, access to viajeros as regular buyers had the unexpected sign, with the effect of increasing inefficiency.

• Environmental mitigation costs are significant in both the independent and contract grower samples, and respective coefficients are signed negatively. The directional sign of the relationship between the value of environmental capture by the farm per unit of output and inefficiency suggests that being environmentally friendly promotes farm efficiency.

• Among contract growers, those with fee contracts are, on average, more efficient than those under price contracts. This may arise from the more rigid terms of fee contracts, where net compensation is closely tied to the passing of base physical productivity standards.

• For both independents and contract growers, those located in Bukidnon, Northern Mindanao, were more efficient, on average, than those located in Luzon.

Scale effects: smallholders versus large-scale producers.

• Only two significant determinants influenced efficiency among large-scale producers: being located in Bukidnon and being a fee contract producer significantly improved mean efficiency.

• For smallholders, there were six significant determinants of inefficiency, but two were wrongly signed, as in the all-independents sample. These two were recent training in livestock or business and access to feeds credit.

• Among smallholders, mirroring the all-independents sample, access to credit for capital and access to information and communications infrastructure significantly reduced inefficiency. The parameter magnitudes, however, were larger among smallholders as a group than among all independents, with the impact of capital credit almost twice as strong.

• Expenditures on environmental cost mitigation worked to improve efficiency. The magnitude of its impact was strongest among smallholders.

Comparative efficiency among hog farms

Figure 8.8 shows the efficiency levels achieved by size quintile.

Figure 8.8 Mean efficiency among hog farms by size quintile, Hog Production, Philippines, 2002

Scale effects: comparative efficiency by size quintile.

• Mean efficiency levels by size quintile are relatively high, ranging from 0.78 to 0.84.

• The variation of mean efficiency levels among quintiles is relatively narrow, suggesting that within the groupings considered (quintiles) there is not much difference in efficiency performance.

• The most efficient are those belonging to the last quintile, that is, where scale is largest (mean efficiency = 0.84).

• Considering the first four size quintiles, smallholders belonging to the smallest quintile are relatively more efficient than the rest (mean efficiency = 0.81).

• The middle three quintiles are relatively even in efficiency performance, within the 0.78-0.79 range.

• In general, smallholders are less efficient than the large-scale hog farms. Smallholders, however, are as efficient as the medium-sized commercial farms.