Научная статья на тему 'Animal food demand in Jakarta, Indonesia: using quadratic almost ideal demand system'

Animal food demand in Jakarta, Indonesia: using quadratic almost ideal demand system Текст научной статьи по специальности «Сельское хозяйство, лесное хозяйство, рыбное хозяйство»

CC BY
136
37
i Надоели баннеры? Вы всегда можете отключить рекламу.
Ключевые слова
Animal food / protein / demand system / urban household

Аннотация научной статьи по сельскому хозяйству, лесному хозяйству, рыбному хозяйству, автор научной работы — Khoiriyah Nikmatul, Anindita Ratya, Hanani Nuhfil, Muhaimin Abdul Wahib

All of the households in Jakarta are urban households, but when viewed from the income elasticity of animal food, all animal food is still a luxury item except eggs. This study aims to analyze the influence of socio-demographic variables, price, and income on animal food demand in Jakarta. The estimation of demand system using Quadratic Almost Ideal Demand System model with the application of Iterated Non-Linear Seemingly Unrelated Regression. Research data using Susenas 2016 is 4,298 households. The results showed that a 1% increase in income would increase demand for eggs, chicken, beef, fish and milk by 0.38%, 1.07%, 2.19%, 1.44%, and 1.84%. Eggs are normal goods while chicken, beef, fish, and milk are luxury items. Beef is most sensitive to income changes. Beef is a substitute for eggs and chicken. The increase in household members of 1 person decreased the consumption of beef by 0.07%. Households in Jakarta are very sensitive to changes in the price of chicken, beef, and fish. To meet protein consumption according to national standards, the stability of beef prices needs to be maintained. In Jakarta, pricing policies are more effective than income policies.

i Надоели баннеры? Вы всегда можете отключить рекламу.
iНе можете найти то, что вам нужно? Попробуйте сервис подбора литературы.
i Надоели баннеры? Вы всегда можете отключить рекламу.

Текст научной работы на тему «Animal food demand in Jakarta, Indonesia: using quadratic almost ideal demand system»

DOI 10.18551/rjoas.2019-04.29

ANIMAL FOOD DEMAND IN JAKARTA, INDONESIA: USING QUADRATIC ALMOST IDEAL DEMAND SYSTEM

Khoiriyah Nikmatul*

Doctoral Program of Agricultural Science, University of Brawijaya & Department of Agribusiness, University of Islam Malang, Malang, Indonesia

Anindita Ratya, Hanani Nuhfil, Muhaimin Abdul Wahib

Faculty of Agriculture, University of Brawijaya, Malang, Indonesia

*E-mail: nikmatul@unisma.ac.id

ABSTRACT

All of the households in Jakarta are urban households, but when viewed from the income elasticity of animal food, all animal food is still a luxury item except eggs. This study aims to analyze the influence of socio-demographic variables, price, and income on animal food demand in Jakarta. The estimation of demand system using Quadratic Almost Ideal Demand System model with the application of Iterated Non-Linear Seemingly Unrelated Regression. Research data using Susenas 2016 is 4,298 households. The results showed that a 1% increase in income would increase demand for eggs, chicken, beef, fish and milk by 0.38%, 1.07%, 2.19%, 1.44%, and 1.84%. Eggs are normal goods while chicken, beef, fish, and milk are luxury items. Beef is most sensitive to income changes. Beef is a substitute for eggs and chicken. The increase in household members of 1 person decreased the consumption of beef by 0.07%. Households in Jakarta are very sensitive to changes in the price of chicken, beef, and fish. To meet protein consumption according to national standards, the stability of beef prices needs to be maintained. In Jakarta, pricing policies are more effective than income policies.

KEY WORDS

Animal food, protein, demand system, urban household.

Three provinces with the lowest share of food expenditure is Yogjakarta (43.00%), Bali (42.73%) and Jakarta (39.94%) (BPS, 2016). Monthly expenditure per Capita in Jakarta by Rp. 1,997,446,-. Percentage of Monthly average expenditure per capita in the food and nonfood by Jakarta by 39.94 and 60.06%. The monthly share of food expenditure per capita in Jakarta in March 2017 by 39.94%, East Java and Bali is 50.79 and 42.73% (Suhariyanto, 2017). Monthly average expenditure per capita of food items in Jakarta for fresh fish and shrimp by 1.48 kg (Rp. 45,638), beef by 0.12 kg (Rp. 12,317), broiler and local chicken meat by 0.76 kg (Rp. 23,158), chicken eggs by 10.57 unit (Rp. 13,511), duck eggs by 0.01 unit (Rp. 23), infant formula by 0.1 kg (Rp. 8,805). Along with increasing income and public awareness of nutrition and food quality, there has been a change in consumption patterns including increased consumption of animal foods (Bharumshah & Mohamed, 1993). Furthermore, Fabiosa (2005) said that income growth would shift the consumption of high-carbohydrate staple foods into more expensive foods such as meat and milk.

The increase in income will increase the demand for animal food (Bharumshah & Mohamed, 1993, Wood, Nelson, & Nogueira, 2012). Increasing demand for Indonesian animal food in the future requires adequate, quality and safe supply readiness. It is consistent with the goal of self-sufficiency, self-reliance, sovereignty, and resilience in food development. Indonesia was targeting self-sufficiency for animal food in 2010, but until now domestic animal food availability has not been sufficient, so imports are still being carried out, except for fish whose needs are met by domestic production. Weber (2015) also explained that if only relying on domestic production, it would be difficult for Indonesia to be self-sufficient in meat. Meat imports in 2010 amounted to 28% and in 2015 imports were still

quite high at 37%. During this time the highest imports occurred in 2014, amounting to 246,509 tons. Domestic supply instability and import dependence often result in very volatile market prices.

Research on the demand for animal food using the QUAIDS approach has previously been carried out in various cities in various countries, both developed and developing countries (Elijah Obayelu, Okoruwa, & Ajani, 2009) in Nigeria, (Mittal, 2010) in India, and (Korir, Rizov, & Ruto, 2018) in Kenya. However, similar research is still rarely found especially in Jakarta. Therefore, this study wants to analyze the impact of price changes on demand for animal food in urban areas in Jakarta. Through this research we will obtain price elasticity and animal food income, whether animal food is normal or luxury goods, whether animal food is a substitute or complementary. This illustrates the consumption behavior and purchasing power of households for animal foods so that these results can be used to develop a protein fulfillment policy in Jakarta.

METHODS OF RESEARCH

Quadratic Almost Ideal Demand System. Estimating demand impact of rising food prices requires reliable price and income elasticities that could be commonly derived from utility-based demand models. The (Okrent & Alston, 2011), Linear Expenditure System (LES) and Theil (1965) Rotterdam model are among the first attempts to derive utility-based demand models. The AIDS model has been the most commonly used spesification in applied demand analysis for more than two decades as it satisfies a number of desirable demand properties. Moreover, it allows a linear approximation at estimation stage and has budget shares as dependent variables and logarithm of prices and real expenditure/income as regressors. (Banks, Blundell, & Lewbel, 1997), however, observed the existense of nonlinearity in the budget shares for some, if not all, commodities and subsequently introduced an extension to permit non-linear Engle Curves. They proposed a generalized Quadratic Almost Ideal System (QUAIDS) model which has budget shares that are quadratic in log total expenditure.

The AIDS as well as QUAIDS models are derived from indirect utility function (V) of the consumer given by:

inV = A(p)p (1)

Where x is total food expenditure, p is a vector a prices, a(p) is a function that is homogenous of degree one in prices, and b(p) and A(p) are function that are homogenous of degree zero in prices; In a(p) and In b(p) are specified as translog and Cobb-Douglass equations as originally specified in Deaton and Muellbauer's AIDS model. Note also that A(p) is set to zero in Deaton and Muellbauer's AIDS model.

1 na(p) = a0 + E"=i a11np1 + -YI¡=i Yij 1 npi 1 npj (2)

b(p) = nUvf1 (3)

Kp) = Z?=1 1npi (4)

Where = 1,.., n represent commodities.

After application of the Roy's identity to equation [1], the QUAIDS expressed in budget shares form is given by (Banks, et al., 1997):

X

w; =«1+ ^=17,1npj + + fa ^n^)} + £i,i = 1......n (5)

Where wi is budget share for good i,a1,yij and fa are the parameters to be estimated, ei is error term.

The demand theory requires that the above system to be estimated under restrictions of adding up, homogeneity and symmetry.

The adding up is satisfied if Y^w = 1 for all x and p which requires.

£=1 «1 = 1,£1=1 p = 0,21=1 Yij = 0,£=1 k = 0 (Adding-up) (6) £7=i m = 0 (Homogeneity) (7)

m = Yn (Slutsky symmetry) (8)

These conditions are satisfied by dropping one of the n demand equations from the system and recovering parameters of the omitted equations from the estimated equations. Household demand for animal food consumption depends not only on their income and product prices but also on household preferences as well as socio-demographic characteristics (Banks, et al., 1997, Poi, 2012). Household demographic factors can be incorporated (in the demand model) using demographic transition method (Pollak and Wales, 1981). The QUAIDS can then be modeled after specifying the constant terms, s,a1, as follows:

«i = Si + £sj=1 Sa Dj ,& £j=1Sij = 0 i = 1.....n (9)

Where Si and Sij's are parameters to be estimated and Dj are demographic attributes including household size. In the letter approaches, zero consumption is modeled in the following system of demand equation with limited dependent variables.

W* = f(Xi, Ui) + Ui, d* = z'idi + Vi, (10)

Where is budget share of good i (as specified above) and di is a binary outcomes that take one if household consumes food item of the considers aggregate, and zero otherwise, and w* and d* are the corresponding unobserved (latent) variables, xi are household expenditure (income) and prices andzi are household demographic and related variables; Ui and di vectors of parameters to be estimated ui and v^ are the random errors.

Assuming error terms ( ui and vi) have bivariate normal distribution with cov (ui,vi) = 0, for each commodity, Shonkwiller and Yen (1999) correct for inconsistency in the demand system by defining the second-stage regression as;

w* = 0(z'di)f(Xi ,Ui) + Si0(z'di) + ei (11)

Where z'idi) and 0(z'idi) are the probability density function (PDF) and the cumulative distribution function, respectively, which are obtained, in theory, from a probit model using equation (10) in the first step for each of food commodity.

The QUAIDS model for animal food demand with household demographic in the second-step in then modified as (Poi, 2012):

w* = «i (p(z'idi) + m=1 Yij 1npj <p(zidt) + Pi (p(z'idi)1n(^-)) + ^ $(z'idi) [1n(^^)]2 + rj^SijDj $(fa) +

Si0(z[di) + ,i=1,...,n (12)

In order to derive conditional expenditure on food prices elaticities, equation (12) is differentiated with respect to lnm and lnp, such that:

= £ = t№i)(Pi + f) l^) 03)

-i = S; = {YiJ - («j + £n=1 Yj> 1nPk) - ^l1n ky}2) (14)

Where p is a price index calculated as the arithmetic mean of prices for all k animal food groups in the system. The conditional expenditure elasticities are then obtained by ei =

V/w*) +1.

Marshallian (uncompensated) price elasticities are derived as = (vi/w*)-di], where d^ is the Kronecker delta equating one when i=j, and zero otherwise. Using the Slutsky equation, the conditional, Hicksian (compensated) price elasticities are given by eij = (v i/w*) + eiWj. Estimating system using Brain P Poi 2008 "demand-system estimation: update, Iterated Non-linear Seemingly Unrelated Regression (Itnsur) model" (Poi, 2012), written in STATA 14. We based on Poi's Itnsur and developed a program that has taken into account the two-stage probit model for zero comsumption expenditure and household demographic.

The data used in this research is secondary data in the form of Central Bureau of Statistics of the Republic of Indonesia, March 2016. The data analyzed include socio-demographic data, household residence status, number of household member, household income, household consumption, price and total expenditure. The animal foods in this study include eggs (chicken eggs, local chicken eggs, and duck eggs), chicken (local chicken meat and chicken meat), beef, fish (fresh fish and shrimp including fish, shrimp, squid, and shellfish) as well as milk (milk powder and infant milk). The sample size is 4,298 households. Data processing proved challenging because many households do not consume animal foods, so many zero observations.

RESULTS AND DISCUSSION

Parameter estimates. Almost all parameters in the animal food demand system in Jakarta are significant at alpha 1 to 5%. The parameters of income and square of income are very significant, as well as the parameters of the number of household members are also very significant. This parameter will be used to calculate the income elasticity, its own-price elasticity, and the Marshallian and Hicksian cross price prices. Table 1 shows the parameter estimates of factors affecting animal food demand in Jakarta.

Table 1 - Parameter estimates animal food demand in Jakarta, 2016

Parameter (Coefficient and SEM) Eggs (1) Chicken (2) Beef (3) Fish (4) Milk (5)

Constant

a 1,611414 -2,640143 1,453717 0,027303 0,547709

(0,080396) (0,110112) (0,096427) (0,080502) (0,093668)

Expenditure

P 0,192306 -0,543867 0,260958 -0,006042 0,096646

(0,012357) (0,017106) (0,016601) (0,014685) (0,017742)

Price

Y_1 0,496095 -0,500526 0,135477 -0,043424 -0,087622

(0,016039) (0,031534) (0,020627) (0,014526) (0,017905)

f,t 1 -0,500526 1,131309 -0,536707 0,114023 -0,208099

(0,031534) (0,088904) (0,056729) (0,036292) (0,046279)

q 0,135477 -0,536707 0,233135 -0,009269 0,177364

Y_3 (0,020627) (0,056729) (0,045186) (0,018845) (0,020366)

•. A -0,043424 0,114023 -0,009269 0,016307 0,102051

Y _4 (0,014526) (0,036292) (0,018845) (0,009273) (0,021122)

-0,087622 -0,208099 0,177364 0,102051 0,102051

Y 5 (0,017905) (0,046279) (0,020366) (0,021122) (0,021122)

Square expenditure

X 0,021583 -0,027715 0,009273 -0,001496 -0,001645

(0,000422) (0,000949) (0,000838) (0,000678) (0,000817)

Demography

^_hhm_tot 0,001392 -0,002057 0,000654 -0,000118 0,000130

(0,000910) (0,000906) (0,000270) (0,000168) (0,000319)

Demography

P_hhm_tot 0,161776 0,161776 0,161776 0,161776 0,161776

(0,026703) (0,026703) (0,026703) (0,026703) (0,026703)

Source: Authors' computation from Susenas, 2016.

Income and own-price elasticity. Table 2 present the income elasticities, uncompensated own-price elasticities, and compensated own-price elasticities. All animal foods have positive income elasticity. It is consistent with the economic theory that when income increases, households will increase consumption of animal food as a source of protein (Akaichi & Revoredo-Giha, 2014). A 1% increase in household income will increase the demand for eggs, chicken, beef, fish and milk by 0.38, 1.07, 2.19, 1.44 and 1.84% respectively. Eggs are normal items. It is indicated by the value of income elasticity of less than 1. Beef and milk are luxury items. It is indicated by the value of the elasticity of income of more than 1. Chicken meat and fish are luxury items but tend to be normal items. It is indicated by the value of income elasticity closed to 1 (Cupak, Pokrivcak, & Rizov, 2015, Bilgic & Yen, 2013).

Table 2 - Income elasticity, Marshallian and Hicksian Own-price elasticity

Animal food groups Income elasticity Price elasticities Number of household member

Marshallian Hicksian

Eggs 0,38180 -0,63816 -0,48355 0,001392

(0,00812) (0,03787) (-0,03719) (0,000910)

Chicken 1,07257 -1,64344 -1,29633 -0,002057

(0,01282) (0,05539) (-0,05521) (0,000906)

Beef 2,19585 -2,60731 -2,47695 0,000654

(0,04045) (0,24623) (-0,24644) (0,000270)

Fish 1,44415 -2,48026 -2,40039 -0,000118

(0,03900) (0,15617) (-0,15624) (0,000168)

Milk 1,83761 -1,22798 -0,93993 0,000130

(0,02409) (0,06618) (-0,06641) 0,000319)

Source: Authors' computation from Susenas, 2016.

All animal foods have negative price elasticity both Marshallian and Hicksian. It is also in accordance with the economic theory that when there is an increase in prices, households will reduce consumption of a bundle of commodities (Matsuda, 2006). Beef is most sensitive to prices, followed by fish, chicken, fish, and milk (Table 2). Marshallian price elasticity has a greater value (in absolute terms) compared to Hicksian elasticity. It is because the Marshallian price elasticity contains the effect of changes in prices and income, while the elasticity of Hicksian prices only contains the effect of price changes (Demeke & Rashid, 2012, (Weber, 2015).

Demographic effects. The household member includes each of the persons who form household regardless of whether he or she is present or temporarily absent at the date of enumeration. However, a household member who on a journey for six months or longer, or less than six months but intended to move away, is not regarded as a household member (Bellemare, Barrett, & Just, 2013). The number of household members (HH size) influences the demand for household animal food in Jakarta statistically high significance at the 1% level. HH size has a positive relationship with the animal food demand for eggs, beef, and milk, but a negative relationship with chicken and fish. The increase in the number of household members one person will reduce the consumption of chicken and milk meat by 0.09% and 0.017% (Table 2).

Cross-price elasticity. Table 3 shows cross-price elasticity between household animal foods in Jakarta. If the relationship between animal food is positive means, there is a substitution relationship, and if it is negative, then there is a complementary relationship (Matsuda, 2006)(Mittal, 2010), Korir, Rizov, & Ruto, 2018). Marshallian cross-price elasticity for egg groups is negative with all animal food, chicken, beef, fish and milk. It means that among all animal foods complement each other. In other words, households in Jakarta consume animal food simultaneously. If there is an increase in animal food prices, households in Jakarta will reduce consumption of eggs, chicken meat, and milk. Conversely, if there is a decline in animal food prices, households in Jakarta will increase consumption of eggs, chicken, and milk together. The increase in income followed by the decline in milk

prices will increase the demand for eggs, chicken and beef by 5.98%, 16.91%, and 0.49% respectively.

Table 3 - Cross-price elasticity of animal food demand

iНе можете найти то, что вам нужно? Попробуйте сервис подбора литературы.

Cross-price elasticity of Marshallian (Uncompensated Elasticity)

Animal food groups Eggs Chicken Beef Fish Milk

Eggs Chicken Beef Fish Milk -0,63816 0,22027 -0,00373 0,00830 0,03152

(0,03787) -0,02341 (0,03276) -1,64344 (0,02416) 0,27591 (0,01968) 0,29199 (0,02054) 0,02639

(0,04191) -0,69261 (0,05539) 1,10445 (0,03132) -2,60731 (0,02569) -0,22684 (0,03093) 0,22647

(0,16866) -0,36622 (0,17106) 1,57711 (0,24623) -0,19269 (0,14089) -2,48026 (0,11884) 0,01790

(0,14728) -0,49490 (0,15083) -0,21526 (0,15121) 0,11668 (0,15617) -0,01615 (0,10333) -1,22798

(0,05496) (0,06459) (0,04550) (0,03666) (0,06618)

Cross-price elasticity of Hicksian (Compensated Elasticity)

Animal food groups Eggs Chicken Beef Fish Milk

Eggs Chicken Beef Fish Milk -0,48355 0,34383 0,01894 0,02941 0,09137

(0,03719) 0,41092 (0,03271) -1,29633 (0,02420) 0,33958 (0,01969) 0,35131 (0,02064) 0,19452

(0,04124) 0,19659 (0,05521) 1,81509 (0,03132) -2,47695 (0,02569) -0,10541 (0,03102) 0,57068

(0,16451) 0,21859 (0,17077) 2,04448 (0,24644) -0,10696 (0,14085) -2,40039 (0,11921) 0,24428

(0,14395) 0,24923 (0,15033) 0,37945 (0,15130) 0,22577 (0,15624) 0,08548 (0,10369) -0,93993

(0,05326) (0,06413) (0,04543) (0,03664) (0,06641)

Source: Authors' computation from Susenas, 2016.

CONCLUSION

This study uses the QUAIDS model approach to see the impact of price changes on animal food demand in urban Jakarta. The number of samples is 4,298 households. The results of the study show that all animal food income elasticity in Jakarta is positive. All price elasticities are either Marshallian or Hicksian were negative. Eggs are normal goods, while chicken, beef, fish, and milk are luxury items. Eggs are substitute with chicken, beef and milk. Households in Jakarta consume animal food simultaneously because it is seen from the cross elasticity of prices that are mostly negative. If there is an increase in animal food prices, households in Jakarta will reduce consumption of eggs, chicken meat, and milk. Conversely, if there is a decline in animal food prices, households in Jakarta will increase consumption of eggs, chicken, and milk together. The increase in income followed by a decrease in milk prices will increase the demand for eggs, chicken, and beef.

ACKNOWLEDGMENTS

Acknowledgments are submitted to the Central Bureau of Statistics of the Republic of Indonesia which has served the process of the data purchasing and to the Ministry Research and Technology and Higher Education for the funds through the Doctoral Program of Doctoral Dissertation 2018.

REFERENCES

1. Abdulai, A., & Aubert, D. (2004). A cross-section analysis of household demand for food and nutrients in Tanzania. Agricultural Economics, 31(1), 67-79.

2. Akaichi, F., & Revoredo-Giha, C. (2014). The demand for dairy products in Malawi. African Journal of Agricultural and Resource Economics.

3. Banks, J., Blundell, R., & Lewbel, A. (1997). Quadratic Engel curves and consumer demand. Review of Economics and Statistics, 79(4), 527-539.

4. Bellemare, M. F., Barrett, C. B., & Just, D. R. (2013). The welfare impacts of commodity price volatility: evidence from rural Ethiopia. American Journal of Agricultural Economics, 95(4), 877-899.

5. Bharumshah, A. Z., & Mohamed, Z. A. (1993). Demand for Meat in Malaysia: An Application of the Almost Ideal Demand System Analysis. Pertanika Journal of Social Sciences & Humanities, 1, 91-99.

6. Bilgic, A., & Yen, S. T. (2013). Household food demand in Turkey: A two-step demand system approach. Food Policy, 43, 267-277.

7. Cupak, A., Pokrivcak, J., & Rizov, M. (2015). Food demand and consumption patterns in the new EU member states: The case of Slovakia. Ekonomicky Casopis, 63(4), 339-358.

8. Demeke, M., & Rashid, S. (2012). Welfare impacts of rising food prices in rural Ethiopia: a Quadratic almost ideal demand system approach.

9. Elijah Obayelu, A., Okoruwa, V. O., & Ajani, O. I. Y. (2009). Cross-sectional analysis of food demand in the North Central, Nigeria: The quadratic almost ideal demand system (QUAIDS) approach. China Agricultural Economic Review, 1(2), 173-193.

10. Fabiosa, J. F. (2005). Growing demand for animal-protein-source products in Indonesia: trade implications.

11. Korir, L., Rizov, M., & Ruto, E. (2018). Analysis of household food demand and its implications on food security in Kenya: an application of QUAIDS model. 92nd Annual Conference, April 16-18, 2018, Warwick University, Coventry, UK. Agricultural Economics Society.

12. Matsuda, T. (2006). Linear approximations to the quadratic almost ideal demand system. Empirical Economics, 31(3), 663-675.

13. Mittal, S. (2010). Application of the QUAIDS model to the food sector in India. Journal of Quantitative Economics, 8(1), 42-54.

14. Okrent, A. M., & Alston, J. M. (2011). Demand for food in the United States. A Review of Literature, Evaluation of Previous Estimates, and Presentation of New Estimates of Demand. Giannini Foundation Monograph, 48.

15. Poi, B. P. (2012). Easy demand-system estimation with quaids. The Stata Journal, 12(3), 433-446.

16. Weber, R. (2015). Welfare Impacts of Rising Food Prices: Evidence from India. 2015 Conference, August, 9-14.

17. Wood, B. D., Nelson, C. H., & Nogueira, L. (2012). Poverty effects of food price escalation: The importance of substitution effects in Mexican households. Food Policy, 37(1), 77-85.

18. Zellner, A. (1962). An efficient method of estimating seemingly unrelated regressions and tests for aggregation bias. Journal of the American Statistical Association, 57(298), 348-368.

i Надоели баннеры? Вы всегда можете отключить рекламу.