Descriptive Statistics of Food Expenditure among Phililippine Household, FIES Data

Associated Press photo by Aaron Favila

Credit: AARON FAVILA

The average food expenditure by income group and geographical location in 2006 presented in Table 1 below (Food Income and Expenditure Statistics, 2006) . Among the food groups, rice grain accounts for the highest budget share at 30% of total households’ food expenditure. Rice remains the main staple food in the Philippines, contributing 35% of the total food calorie intake but as much as 60-65% of the households in the lowest income quintile (CPBD, 2008). Low income families spend a larger part of their income for rice consumption. Rice accounts for about 25% of total household’s expenditure in the low-income group. Families belonging to the high income bracket, on the other hand, allocate lower budget from their income for rice consumption at 17.8%.

Between urban and rural households, the latter allocates more budget for rice consumption at 27.5% of its share relative to urban households with 17.3%. Unlike rice, corn is a staple commodity for a very small portion of the Filipino household particularly in Central Vizayas and in some parts of Mindanao. This explains the small percentage of consumption for corn. The leafy vegetables, however, posted the lowest budget allocation relative to other food item in all population groups. Meat accounts for nearly 12% share of all total households with the budget allocation of urban households higher (15.3%) than those in their rural counterpart (11%). The budget allocation for meat increases with increase in income. On the other hand, the budget allocation for fish decreases with increase in income.

Table 1. Average Expenditure By Income Group and Geographical Locations
	Entire	Location		Income Group
Food Group	Sample	Urban	Rural	Low	Middle	High
Rice	0.2990	0.1773	0.2750	0.2430	0.2090	0.1748
	(0.1303)	(0.1012)	(0.1360)	(0.1334)	(0.1214)	(0.1226)
Corn	0.0314	0.0120	0.0471	0.0349	0.0256	0.0183
	(0.0817)	(0.0407	(0.1010)	(0.0864)	(0.0729)	(0.0735)
Fresh Fruits	0.0320	0.0332	0.0311	0.0318	0.0324	0.0320
	(0.0201)	(0.0205)	(0.0198)	(0.0204)	(0.0196)	(0.0200)
Leafy Vegetables	0.0181	0.1010	0.0205	0.0186	0.0172	0.0178
	(0.0129)	(0.0096)	(0.0146)	(0.0134)	(0.0118)	(0.0012)
Fruit Vegetables	0.0220	0.0199	0.0237	0.0225	0.0213	0.0216
	(0.0135)	(0.0116)	(0.0147)	(0.0137)	(0.0133)	(0.1222)
Meat	0.1290	0.1530	0.1097	0.1229	0.1383	0.1606
	(0.0748)	(0.0172)	(0.0713)	(0.0744)	(0.0741)	(0.0720)
Eggs	0.0216	0.0226	0.0208	0.0215	0.0218	0.0200
	(0.0131)	(0.0130)	(0.0130)	(0.0132)	(0.0216)	(0.0134)
Fish	0.1300	0.1166	0.1409	0.1335	0.1251	0.1197
	(0.0608)	(0.0530)	(0.0643)	(0.0620)	(0.0581)	(0.0592)

The data consists of 18,411 observations (Philippines); 8,243 observations (Urban Area); 10,168 observations (Rural Area); 11,490 observations (Low-income Group); 5,482 observations (Middle-income Group); 1, 436 observations (High-income Group); collected from individual Filipino households, 2004-2006.

* Standard deviations  are in parenthesis

Brief Review of Philippine Food Demand Studies

image source: farmlandgrab.org

There have been few food demand studies done using the Philippine household demand data employing AIDS model. Recently, Balisacan (1994) used LA-AIDS model to explained the consumer behavior of urban and rural areas towards the changes on price and income using average expenditure shares for each region and by area ( urban and rural). The study was focused on the aggregated commodities (i.e. cereal, meat beverages, fuel, house and clothing) and not  mainly on food commodities in specific, as this paper is aiming to estimate. The expenditure elasticities, however, suggested that in the commodity group such as cereal, meat, beverages, clothing and miscellaneous groups were inelastic with the expenditure elasticities of 0.382, 0.805, 0.990, 0.967 and 0.25 respectively. Further, the own price elasticities of identified food commodities were not found significant. On the other hand, the uncompensated price elasticities showed a strong evidence between the food groups and non food groups. The price of cereal, for example, has a significant positive effect on the demand for fuel and house with the estimated coefficients of 0.161 and 0.431 respectively.

The result also showed the substantial difference in the demand response  by various population groups to changes in income. In case of cereal, the expenditure elasticity was considerably lower for urban areas than for rural areas, especially in 5^th  income quantiles with estimated expenditure elasticity of -0.463 compared to 0.179 for rural area. Further, there were little variation in price elasticities across population groups. This result, however, was not unexpected since the data set used in the study did not contain information found among households in different circumstances (Balisacan, 1994).

Llanto et.al. (1996), however, made used of QUAIDS model specification by Blundell et.al. (1993) using a cross-sectional household data of Food Expenditure Survey in year 1991. The analysis was focused mainly on the food groups such as cereal, fruits and vegetables, meat and fish products, dairy products and other food items across different population groups. The demand elasticities for cereal was inelastic to its own price for various household classification of urban households, farm households and deficit households with the calculated elasticities of 0.87, 0.89 and 0.90 respectively. Cereal products however, was found elastic, among rural households, at varying household sizes, agricultural households and surpluses households. On the other hand, the result further showed that fruits and vegetables , meat and fish products and dairy products, irrespective to various  household groups, cereal is positive. This implied that cereals was served as substitute good for those  who cannot afford those food groups mentioned at the prevailing prices. Further, the studies showed that the estimated income elasticity were all positive in all population groups, this means that all food groups were all normal goods.

A more recent study applying Philippine household data was conducted by Orbeta et.al. (1998) in simulating the impact of macroeconomic changes on the nutrition status of Filipino households using the LA-AIDS. Using the Tariff Reform Program implemented between 1988-1992 as policy change, the study showed a more progressive impact on nutrition compared to the impact of income. The result of LA-AIDS demand parameters were used to simulate the effect of policy change across income distribution.

Source of basic data: FIES-NSO, various years

image source: ats-ses.agr.gc.ca