Animal products consumption patterns of rural households in Turkey

C Akbay

Kahramanmaras Sutcu Imam University, Faculty of Agriculture,
Department of Agricultural Economics, Kahramanmaras, Turkey
cakbay@ksu.edu.tr

Abstract

This paper analyzes socioeconomic and demographic factors influencing the households' preferences and consumption of red meat, poultry, fish, fluid milk, yogurt, cheese, butter, and eggs in rural areas of Turkey by using 2003 household survey data.

According to results, the shares of the total food consumption in total household's expenditure are about 38.0 percent in rural areas and 25.8 percent in urban area. Also, as household income increases, total food consumption increases, but the share of the food consumption in total expenditure decreases. Results show that rural households' animal products consumption are highly influenced by socioeconomic and demographic characteristics of the households such as household income, sizes of household, education levels of head of households and wife, age and gender of household head, employment status of household head and wife and regional differences. Animal product consumption patterns for lower- and higher- income households in rural areas are highly different. Increases in income received by higher-income households do increase their animal product consumption and expenditures. Response to changes in income is higher for meat which indicates that increasing income increases meat consumption share more than other items. Moreover, increasing educational levels increase animal consumption of meat, milk, cheese, butter, and eggs.

Key words: Cheese, consumption patterns, eggs, meat, milk, rural household, Turkey

Introduction

Food consumption patterns are rapidly changing in Turkey, as well as in other developing countries. Per capita consumption of milk and red meat products has decreased and meat consumption has shifted over time from beef, veal, lamp, mutton and goat to a greater consumption of poultry and fish due to changing tastes, costs and income (Akbay and Boz 2005). For example, annual per capita consumption of milk fell from 175 kg in 1960 to 98.1 kg in 2002. Per capita total meat consumption has increased from 21.0 kg in 1960 to 35.0 kg in 2002 even though red meat consumption has decreased from 14.2 kg to 9.42 kg in the same period (Akbay and Boz 2005; FAO 2005). These changes occurred both in urban and rural areas. However, these changes become more important for rural areas because of the growing food distribution channels and increasing household income and education levels of consumers.

Rural areas consist of small villages, district or sub districts where population in these areas is mainly involved in agricultural activities. Because of cultural and educational factors, food consumption patterns of households in rural areas are different from urban households. According to 2000 data, 35% of the population lives in rural settlements, which accounted 43% of employment (SIS 2000).

This paper focuses on households' animal products consumption levels and patterns of Turkey's rural households across the income and different categorical groups. Although many studies exist on Turkey's animal product consumption, most of them focus on urban household animal consumption in relatively small provinces or cities (Hanta and Yurdakul 1995; Albayrak 1999; Koc and Tan 1999; Oguz and Kucukcongar 2002). This paper represents a first attempt to address the issue of animal products consumption of rural households in Turkey.

Data used in the study come from the 2003 Turkish Household Consumption Expenditure Survey (HCES) collected and provided by the Turkey State Institution of Statistics. The HCES was designed to be nationally representative of Turkish households.

The next section describes the data and outlines the methodology to analyze rural household' animal products consumption. This section is followed by an overview of the rural household' animal products preferences and consumption by income groups. Then, households' consumption of each animal product by different socioeconomic and demographic groups are analyzed and the last section summarizes the results and conclusions.

Materias and methods

The data used in this study comes from a cross-sectional survey data of 7488 households in Turkey's rural areas and environs. The survey was conducted between January 1 and December 31, 2003 and implemented by Turkey State Institution of Statistics (SIS 2003). In survey, 624 households were interviewed in each month rotely. For example, 624 sample households were interviewed in January to get data on their consumption expenditures and socioeconomic and demographic characteristics of the household members; in February another 624 households represent to households that of January were interviewed. This alternate process continued until December 2003. In the survey, the stratified multi-stage systematic cluster sampling method was used. 12 geographical regions stratification was used for survey. Using the latest definition of the State Planning Organization, the settlements with a population of 20 000 and less were taken as rural in every region. Income was used to classify households into income quintiles. The data allocated to the five strata using monthly disposable incomes of all households surveyed were ordered from less to more separately and quintiles were constituted by dividing the households to 5 equal parts.

This study includes 9 animal originated products by assuming weak separability between these products and all other goods. These product groups are namely (1) red meat, (2) poultry, (3) fish, (4) other meat products such as sausage, salami and sucuk (garlic-flavoured sausage), (5) butter, (6) fluid milk, (7) cheese, (8) yogurt and (9) eggs. Beside households' food consumption quantities and total expenditures, data also include socioeconomic and demographic characteristic for each sample household such as household income and expenditures, household size and distribution, and head of household's employment, age and education.

Data was analyzed using descriptive statistics, Chi-square test, F-test and t-test statistics. Since the variable gender, household size, marital status, employment status of household head and wives had each two categories, the t-test procedure is used to find out if there is a significant difference among the groups of each demographic variable. Since the variables, age, education, income and geographical locations, were measured in more than two groups, the analysis of variance procedure (F-test) was used to find out if significant differences exist in animal originated product consumption among the groups of each demographic variable. Moreover, the Tukey's Post-hoc Multiple comparisons test was used to find out which groups are significantly different than others. Finally, Chi-square procedure was used to test if households' animal products consumption decisions (consumption, non-consumption) are statistically different between income groups
 

Results

Table 1 gives the basic socioeconomic and demographic characteristics of the rural households. The average household has 4.6 people. Household size increased with income. For example, there were 4.1 people in poorest household and 5.2 people in the wealthiest households. Data indicated adult relatives and domestic workers boosting household sizes among the higher income households. For example, in lowest income households only 9.3% of households have people aged more than 50 years old. This percentage for higher income households was 19.1%. Men head represented 93.2% of the rural households. Men household head were over represented among the wealthiest households. Among household heads, about 19.0% of the sample were illiterate or literate without a diploma, 69.7% have elementary level of education and 3.4% have graduated college or university. Similarly, about 26.0% of wives of the household heads were illiterate or literate without a diploma, 59.9% have elementary level of education and 0.9% graduated from college or university. 18.5% of wives were working and 59.2% of households have children less than 14 years of age.

Table 2 shows the household average monthly income, total expenditure and food consumption expenditures for rural households by income strata. Across the sample for total household, expenditure averages 183.6 US $ to 631.2 US $ from the lowest to highest income quintile. Total household expenditure decreases as a percentage of income as income rises. The share of the total household income spent on all expenditure fell from 124.7% of income among the poorest quintile to 58.0% among the wealthiest income group. Results showed that poorest household spent more than their household income.

There is a strong link between food expenditures and income. Engel's law states that, as income increases, the share of a household's budget spent on necessities, such as food, will decrease. However, it also suggests that the actual amount spent on food may increase because lower-income households may alter the composition of their food bundle as their income rises. Information in this section also examines differences in food expenditure by income quintile. In general, analysis of data confirms Engel's law, which states that the proportion of income spent on food declines as income increases. Across the sample for food expenditure averages 88.2 US $ to 190.8 US $ from the lowest to highest income quintile. Not surprisingly, food consumption decreases as a percentage of income as income rises. The expenditure on food averages 38.0% of household expenditure across the sample. Relatively poorer household spend a larger proportion of their expenditure on food peaking at 48.0% of expenditure among the poorest quintile and falling off to 42.0% among the 3^rd quintile and 30.2% among the wealthiest households.

Related to the animal originated product consumption, the highest-income households spent significantly more on animal products, per households, than did the lowest-income households. In fact, the difference is relatively large: 180.8% more. Average household expenditure on animal products items in rural Turkey was 41.7 US $ with quintiles means ranging from a low of 23.0 US $ among the poorest households to 64.5 US $ among the wealthiest group. Share of the animal product consumption in total food expenditure was average 30.7% ranged between 26.1% for the lowest-income households and 33.8% for the highest-income households (Table 2).

In this study, households' animal products consumption is broke into its components so the relative importance of the different food items can be shown. Differences in animal products consumption levels and consumer preferences can be observed across income groups in Turkey. According to results in Table 3, as income increases the proportion of households purchasing each animal products rises.

Red meat was consumed by only 47.2% of households, particularly among the rich. In the top income quintile, 65.2% of households consumed red meat against 26.0% in poorest household group. Poultry was consumed by 57.8% of households and fish by only 30.9% of households. While relatively more rich households consumed butter, it is consumed across all income groups. 73.2% of households in lowest income groups and 88.9% of households in highest income group consumed fluid milk. Moreover, average 69.8% of households consumed yogurt which is a very famous Turkish food, 79.1% of household consumed cheese which is popular breakfast food and finally 88.8% of household consumed eggs which again popular food in breakfast.

From the table 4 and 5, we hypothesize that if lower-income households actually consume less animal products than higher-income households, then, they must also have significantly lower expenditures on these items.

For most animal products analyzed in this study, higher-income households consumed more per household than did lower income households. Univariate analysis shows that animal food expenditure patterns vary across income levels. Per household red meat consumptions is average 2.49 kg ranged from 0.97 kg from lowest income household to 3.98 kg for highest income households. Per household poultry consumption averaged 1.16 kg in lowest income household groups, compared to 2.86 kg in highest income households groups, a difference of 146.6%. Similarly, per household consumption of fish, butter, fluid milk, yogurt, cheese and eggs are average 1.07 kg, 0.52 kg, 13.63 lt, 9.67 kg, 2.78 kg and 37.52 units, respectively, and these quantities increases as household income increases.

Not only would average expenditures be significantly different between income groups, but expenditures share of each animal products in total animal product expenditure between income groups would also be significantly different. As shown in Table 5, expenditure share of meat and meat products excluding poultry and fish accounted for 24.77% of animal product consumption for lowest income households, but 41.80% for highest income ones. Although quantity amount of animal products increases with income, expenditure share for these animal products, except meat, decreases as income increases. For example quantity share of fluid milk is 14.40% for lowest income households, but only 9.64% for highest income households.

The univariate analysis in Table 6 present differences in households' animal products consumption by socioeconomic and demographic characteristics of the households. Results show that larger households spend more animal products than smaller households. Since larger households have more people and spent more time preparing meals, it is not surprising that animal product consumption are higher for larger households. For example, a high positive relationship has been found between numbers of members in the household and consumption of dairy products.

Data also was able to give a gender dimension. Surprisingly, male-headed households consumed an additional 2.19 kg of meat, 1.8 kg butter, 3.64 lt milk, 3.59 kg yogurt, 0.85 kg cheese and 7.82 units of eggs per month compare to female household head. Thus, one can say that males in rural areas are not much concerned about healthy diet as they consume more red meat. However, male head of households also consumed more from other animal products compare to females.

The level of education influences households' animal product consumptions. We hypothesized that more educated consumers consume more animal products than less educated consumers. Meat is the preserve of the most educated groups with consumption increasing with the education to peak at household where the head graduated from university or college consumed average 8.79 kg meat and meat products. This decreases to 6.63 kg among those who finished secondary and high school 5.73 kg among those who complete elementary school and decreases to 4.98 kg among household where the head never went to school or is literate without a diploma. Consumption of butter among households where the head is illiterate or literate without a diploma is high at 0.49 kg. Consumption rises to 0.56 kg among those who completed elementary school and falls from there as education levels increases. This pattern is much stronger in the case of education level of wife. Increases education level of wife decreases amount of butter consumption an important amount from 0.64 kg for households who are illiterate or literate without a diploma to 0.20 kg for university-graduated wife. Amount of yogurt is lowest among those who graduated from university or college and increases from 4.73 kg per household to peak at 12.18 kg among households where the head never went to school or not-completed elementary school. Eggs is the preserve of the most educated groups with consumption increasing with the education to peak at 42.15 units per household among household headed by a university graduate.

Even thought married household head are slightly greater consumption of animal products, except eggs, the differences are not found to be statistically significant at 0.05 levels. Married-headed households show a greater consumption of egg than do single-headed households. The effect of the age of the head of the households was examined for each animal product groups and found statistically significant for all commodity groups. Households with older head of household consumed more meat and butter than younger ones. Households with the head of household between 41 and 50 years old consumed more milk, yogurt, cheese and eggs than younger age groups. However, as age of the head of household increases more than 50, consumption of most animal products decreases. This behavior seems to be related to health issues; older people tend to follow healthier diets.

Since employment of household head and wife may reflect lifestyles and economic well-being of the consumers these factors have significant effects on households animal product consumptions. Our results showed that employed household head and wife consume more meat, butter, milk, yogurt and eggs. Finally regional differences are included because it reflects cultural and lifestyle differences and economic well-being. Regional differences are found to be very important factor for all animal product groups which investigated.

The key results from this study are the following. First, socioeconomic and demographic factors have high level of influences on Turkish households animal product consumption and demand. Second, Turkish households animal product consumption will continue to be sensitive to per capita incomes. Increasing income will cause large increases in demand for animal products especially for meat and meat products.

Conclusions and recommendations

• This paper investigates households' animal product consumption patterns and socioeconomic and demographic factors influencing the households' preferences and consumption of red meat, poultry, fish, fluid milk, yogurt, cheese, butter, and eggs in rural areas of Turkey by using 2003 household survey data. As expected the results obtained in this research suggest that higher income households devote a much higher proportion of their budget to total food. According to results, rural households animal products consumption are highly influenced by socioeconomic and demographic characteristics of the households such as income, education, age, regional differences, employment status of household head and wife and household size. Animal product consumption patterns for lower and higher income households in rural areas are highly different. The proportion spent in animal products in lower income households (26,1%) is smaller than higher income ones (33.8%). These differences can also be observed across regions.

• This study has shown that the magnitude, and even directions in many cases and taste changes for several major product categories vary substantially among demographic groups in the rural areas. This information has considerable practical significance for the food industry. It is important for policy and marketing purposes to know whether an observed change in food consumption is being affected by socioeconomic and demographic factors and income changes. There is relatively not much thing policy makers or livestock sector can do to influence socio-demographic characteristics of the households such as household size, age, gender, marital status and regions. But education level of consumers and the incomes that also results from education are factors over which policy makers have an influence. Increases in incomes will be the biggest factor influencing per capita animal product consumption in the rural area of Turkey.

• Future research is needed to determine whether or not the level of animal product consumption is associated with the variables attitudes of consumer behavior such as health and believes about animal products. Moreover, future research should analyze income and price elasticity of animal products.
   

References

Akbay C and Boz I 2005 Turkey's livestock sector: Production, consumption and policies. Livestock Research for Rural Development 17(09) 2005 http://www.cipav.org.co/lrrd/lrrd17/9/akba17105.htm

Albayrak M 1999 Analyzing of the Changes of Animal Products Consumption Types in Turkey, International Livestock'99 Congress, September 21-24, Izmir-Turkey.

FAO 2005 Statistical Database, FAOSTAT WEB page, Rome. http://faostat.fao.org/

Hanta B and Yurdakul O 1995 Adana ilinde kentsel alanda hayvansal gida tuketim yapısı. CU. Ziraat Fakultesi Dergisi 10(2): 169-184.

Koc A and Tan S 1999 Household demand for dairy products in Turkey: The impact of Household's composition on consumption. Agricultural Economic Research Institute, Ankara.

Oguz C and Kucukcongar M 2002 Konya ili Selcuklu ilcesi hanehalkinin sut ve sut urunleri tuketim ve satin alma davranislari. Turkiye V. Tarim Ekonomisi Kongresi, Eylul 2002.

SIS 2000 Statistical Yearbook of Turkey. State Institute of Statistics, Ankara, Turkey.

SIS 2003 Household Budget Survey. State Institution of Statistics, Ankara, Turkey