Looking for Factors Affecting Food Security in the Province of Yogyakarta Special Region (DIY) in 2019

Food security is closely related to community welfare and poverty. The ultimate goal of food security is to improve human welfare which can be seen from the fulfillment of one's right to food. The type of data used in this study is the data from March 2019 Susenas DI Yogyakarta. The research includes measuring the degree of food security using indicators of household income and nutritional consumption. This study uses a factor analysis of the CFA (Confirmatory Factor Analysis) method. Resilience at the national level is based on resilience at the household level. For this reason, a study on household variables is the focus of the government in realizing food security. This study identifies 13 variables of household characteristics, identification using Confirmatory Factor Analysis (CFA) to classify the variables into factors/dimensions of food security. Based on the results of research, 13 household characteristics variables are divided into 4 factors, namely the Dimension of Food Availability including the level of consumption of calories/capita/day, household BPNT acceptance in quintile 2, rice prices, and expenditure groups per capita with a loading factor of 0.80 are the dominant variables. The Food Stability dimension includes the sex of the head of the household, the number of household members, and the variable of the marital status of the head of the household is the dominant variable with a factor loading of The dimensions of access to food to get food consist of the status of the area of residence, the status of the head of the household the formal sector, the age of the head of household, and the dominant variable of the education of the head of the household with a loading factor of 0.64. The dimensions of use include the number of household members experiencing health as the dominant with a loading factor value of 0.89 and the of household members who have been outpatient in the last month. This is an open access article under the CC–BY-SA license.

Food security is closely related to community welfare and poverty. The ultimate goal of food security is to improve human welfare which can be seen from the fulfillment of one's right to food. The type of data used in this study is the data from March 2019 Susenas DI Yogyakarta. The research includes measuring the degree of food security using indicators of household income and nutritional consumption. This study uses a factor analysis of the CFA (Confirmatory Factor Analysis) method. Resilience at the national level is based on resilience at the household level. For this reason, a study on household variables is the focus of the government in realizing food security. This study identifies 13 variables of household characteristics, identification using Confirmatory Factor Analysis (CFA) to classify the variables into factors/dimensions of food security. Based on the results of research, 13 household characteristics variables are divided into 4 factors, namely the Dimension of Food Availability including the level of consumption of calories/capita/day, household BPNT acceptance in quintile 2, rice prices, and expenditure groups per capita with a loading factor of 0.80 are the dominant variables. The Food Stability dimension includes the sex of the head of the household, the number of household members, and the variable of the marital status of the head of the household is the dominant variable with a factor loading of 0.96. The dimensions of access to food to get food consist of the status of the area of residence, the status of the head of the household in the formal sector, the age of the head of the household, and the dominant variable of the education of the head of the household with a loading factor of 0.64. The dimensions of food use include the number of household members experiencing health complaints as the dominant variable with a loading factor value of 0.89 and the number of household members who have been outpatient in the last month. 10.12928/optimum.v10i2.15012 ensure a person can live and do daily activities. The guarantee for humans to be able to live healthily and work productively is better known as food security (Rahman, et al, 2002).
In 2019, the Food Security Index (IKP) of the Special Region of Yogyakarta (DIY) reached 83.63 or the best food security status (Food Security Agency of the Ministry of Agriculture, 2019). The IKP DIY is the second-highest after Bali Province. The food security index is measured based on aspects of food availability, aspects of food affordability, and aspects of food utilization. Each of them is measured by a variable that has different weights.
When viewed from the minimum calorie intake to be able to live a healthy and active life, the DIY Prevalence of Undernourishment (POU) data in the same period was 87.13. POU DIY ranks 13th. POU is measured by adjusting individual calorie needs according to gender, age to height, and certain body weight and activities undertaken.
On the other hand, in the last 10 years, the DIY poverty rate has always been lower than the national poverty rate and is still the highest in Java. The poverty rate in DIY in 2019 was 11.70 percent or only 5.13 percentage points decreased from 2009. On average, the poverty rate decreased by 0.57 percentage points per year. This condition is an initial illustration that the slow poverty reduction will hamper the ability of the population to obtain adequate and nutritious food. Income in household members can affect consumption and food security such as energy and protein consumption of household members (Kusumawati, TD, (2013) et al.,). In other words, poverty becomes a problem to achieve food security. Furthermore, apart from poverty, income inequality is still quite high. One measure of income inequality is shown by the magnitude of the Gini ratio. During the last five years, the DIY ratio (2015-2019) has always been higher than the National. In 2019, the DIY ratio was 0.423, while at the national level it was 0.382. The issue of income inequality provides a signal that there is still a fairly high-income gap between the population in the lowest economic group and the population in the upper economic group. This shows that apart from poverty, income inequality is an obstacle to the ability of the lower class population to access food. The stability of food availability that is evenly distributed among the lower classes is of concern to the government at least.

Figure 2 Development of Gini Ratio DIY and Indonesia, 2015-2019
At the international level, food security is still a world concern and is discussed in the main points in the Sustainable Development Goals (SDGs) document. The countries that are members of the United Nations have agreed to end hunger and achieve food security by 2030. This was included in the SGD's 2nd target. Realizing food security is also a development priority in Indonesia. Resilience at the national level is based on resilience at the household level. If a household is food insecure, it is difficult to see it safe at the national level. Household-level food security refers to the ability to access adequate food at any time (Warr, 2014) The concept of food security is quite diverse. Law of the Republic of Indonesia Number 18 of 2012 concerning food security includes three aspects, namely availability of quantity, safety, and affordability of prices. The distribution of the pillars for food security based on the Indonesian Food Law is availability, accessibility, and stability. Meanwhile, FAO (Food and Agricultural Organization), 1992 defines food security as a situation when all people at all times have sufficient amounts of safe and nutritious food for a healthy and active life.
Food security is explained in 4 pillars, namely food availability, physical and economic access to food, stability of supply and access, and food utilization. (Nurhemi (2014). the fulfillment of food in an even condition, meaning that the distribution of food must support the availability of food at all times and evenly throughout the country.
the fulfillment of food in affordable conditions, that is, it is easy for everyone to obtain at an affordable price.
Hapsari, NI, & Rudiarto, I. (2017) in a study "Factors Affecting Food Insecurity and Security and Policy Implications in Rembang Regency" explains the dimensions of food security from various sources, namely first, food availability depends on food crop production and food production. This is also influenced by the availability of water in a place and food and food production are also influenced by the availability of water one place and a time (Hammer et al., 2001). Second, access to food according to Hanani (2000) consists of economic, physical, and social access. Economic access consists of income received, job vacancies, and food prices. Physical access is related to infrastructure in the distribution process and social access is more likely to be the selection of household foodstuffs. Third, the dimension of food use consists of two elements, namely the use of food that is accessible to households and the body's ability to absorb nutritional content. The use of this food is strongly influenced by several things, including household food storage and processing facilities, customs, culture, and the level of knowledge of housewives in deciding the type of food, food availability in the family, and individual health in a household (Food Security Council, 2009). Fourth, population density in an area also affects the established food security. In the most marginalized and densely populated areas, households are trapped in a Malthusian situation characterized by low resources, high dependence on external inputs, poor accessibility, Furthermore, Peter Warr (Australian National University, 2014) distinguishes food security at four levels, namely the global level, the national level, the household level, and the individual level. Peter Warr further stated that resilience at the national level is based on resilience at the household level. If a household is food insecure, it is difficult to see it safe at the national level. Household-level food security refers to the ability to access adequate food at any time. Not only the adequacy of food intake for today but also includes the expectation of future problems and it is not a problem at this time.
Several studies on food security in Indonesia have been carried out. This research, among others, is based on research on the factors that influence food insecurity and security and the implications of its policies in the Rembang Regency. By Hapsari, NI, & Rudiarto, I.
(2017) it is found that the main factor causing food security is the availability of food and the main factor causing food insecurity is socio-economic factors. Then

Research methods
The data used in this study came from secondary data from the National Socio-Economic Survey (Susenas) in the Special Region of Yogyakarta in 2019. The unit of analysis in this study was households with a sample size of 3,734 households. A household is a person or group of people who live in part or all of a physical building or census, and usually, live together and eat from one kitchen (BPS, 2016: 41). The list and types of variables used in the study are as follows:

Analysis Model Specifications
Santoso 2012 in Aini etc, 2019 states that factor analysis is a method used to find interrelationship between several variables that are initially independent of each other so that one or several sets of variables that are less than the initial number of variables can be created.
The Var (ai) is a commonality or variance value that represents the common factor and Var (ei) represents the error term factor of the variable Var (Xi)

Rotation
Factor rotation was carried out to obtain a better interpretation of the processed data by factor analysis. The rotation method used is the orthogonal rotation method. Orthogonal rotation is rotation with an axis of rotation perpendicular to each other or at right angles.
Using this method, each factor is not correlated or independent between one factor and 10.12928/optimum.v10i2.15012 another. While the orthogonal rotation method is the Varimax method as follows:

Extraction
Several variables are likely to measure the same factors, the extraction procedure is carried out to squeeze or extract the variables until only a few factors are found at the core.
In factor analysis, variables are grouped based on their correlation. Highly correlated variables will be in a certain group to form an actor, while variables in other groups have a relatively small correlation. In this study, the approach used for factor analysis is Confirmatory Analysis Factor (CFA), which is to confirm statistically the model that the researcher has built.
CFA analysis goes through a series of stages as follows: Determine the indicators to be analyzed In this study, there are 13 indicators/variables that will be analyzed to classify into 4 dimensions of food security.

Perform partial correlation calculations using the Kaiser Meyer Oikin test (KMO). The
KMO test aims to determine whether all the data that has been taken are sufficient to be analyzed and CFA. KMO value greater than 0.5, will meet the CFA requirements. KMO is calculated using the formula:

Rij = correlation between variables i and j aij = partial correlation between variables i and j
Conducting a variable feasibility test using the Barlett test of Sphecirity method.
Barlett's Test of sphericity is a statistical test to test whether the variables involved are correlated. The hypothesis used is that there is no correlation between variables for the null hypothesis and there is a correlation between variables for the alternative hypothesis. The KMO and the Barlett test have given the correct conclusions, then the factor analysis is carried out.
Checking the Measure of Sampling Adequacy (MSA). This aims to determine whether the sampling process is adequate or not, which can be seen from the Anti-Image Correlation.
The numbers in this matrix represent partial correlations between variables, that is, correlations that are not influenced by other variables. If the MSA value continues <5 then the variable is excluded from the CFA analysis.
Calculates anti-image matrices. Choose the eigenvalue. The ability of each factor to represent the analyzed variables is indicated by the amount of variance described (eigenvalue). In Total Variance Explained, what needs to be considered is the factor whose eigenvalues are more than 1, this is because the eigenvalues less than 1 are not significant.
Group the variables into 4 factors. The next step is to determine the factor matrix. The factor matrix is presented in tabular form, in this table contains the factor loading or correlation value between each factor and the analysis variables. Not all variables in the factor matrix can be membership determined, therefore rotation is required.
Interpretation of results. Variable grouping based on the loading factor value. The largest loading factor value will determine whether a variable is included in a factor. Valid criteria in CFA analysis or it can be said to be valid if the loading factor is> 0.30. The method section structure should: describe the materials used in the study, explain how the materials were prepared for the study, describe the research protocol, explain how measurements were made and what calculations were performed, and state which statistical tests were done to analyze the data.

Discussion
This study uses the CFA method. In detail, the CFA method will confirm the food security model/concept proposed by FAO (Food and Agricultural Organization), 1992, and the Law of the Republic of Indonesia Number 18 of 2012. 13 household characteristic variables will explain the food security factor, namely food availability ( Food availability), food stability, access to food, food utilization. Each of these factors cannot be measured directly

Validity test
Validate test consists of 2 things, namely the adequacy test of the number of observations (data) and the correlation test between variables (multivariate). From the KMO and Bartlet's Test tables, the Kaiser-Meyer-Olkin value is 0.676. Thus, the KMO is greater than 0.5 so that from the data adequacy perspective, it has been fulfilled for the factor analysis to be carried out. Next is the value of Bartlett's Test of Sphericity. Table   1 shows the value of Barlett's Test Sig = 0.0000 or less than alpha, namely 0.05. This means that we reject Ho so that it can be concluded that there is a correlation between multivariate variables so that factor analysis is feasible to use.
The sample collection process is seen by the value of the Measure of Sampling Adequacy (MSA), this value is to determine whether the sampling process is adequate or not, by looking at the Anti-Image Correlation. Table 4 shows the MSA values of the 13 research variables. The MSA testing criteria are as follows: MSA = 1, this indicator can be predicted without error by other indicators.
MSA> 0.5, the indicator is still predictable and can be analyzed further.

MSA <0.5, the indicator is unpredictable and cannot be analyzed further
The Anti Image Matrices  1 which will later be formed will be able to explain the variation of data by 23.544 percent.
Factor 2 will be able to explain the data variation by 15.66 percent. Factor 3 can explain the variation of data by 11.18 percent and factor 4 will be able to explain the variation in data by 10.125 percent. CFA is used when the main purpose is to identify the factors that underlie a construct.
The basic principle of factor analysis (Marisca, 2014) is to extract several common factors from the original group of variables so that: a. There are fewer factors than many of the original X variables.
b. Most of the information (variety) of the original variable X is stored in some factors.
With rotation, the value of the factor loading or the correlation value between each factor and the analysis variables will be obtained. The value of the biggest loading factor for each variable will determine where the variable will be incorporated into a particular factor. Table 7 shows the value of the factor loading before and after rotation. The rotation used is the varimax orthogonal rotation. The numbers on the rotation factor matrix, 4 new factors are formed. Each of these variables has 4 loading factor values spread over 4 factors.  In detail, it can be seen that the variables of the marital status of the head of the household, the sex of the head of the household, and the number of household members, the biggest factor loading value is in the factor column 1. This shows that the three variables are included in factor 1. Then factor 2 includes the percentile average. per capita expenditure average, calorie consumption, and BPNT revenue. Factor 3 consists of the variable number of household members who experience health complaints and the variable number of households who are outpatient. Factor 4 includes the education variable of the head of the household, the status of the area of residence, the work status of the head of the household, and the age of the head of the household.
Based on the variables incorporated in the factors, the names of each factor that affect food security in Yogyakarta Province are as follows.

No.
Factor Name Variable The dimension of food stability, the most dominant variable is the status of the head of the household with a factor loading of 0.96. The dimension of food availability, the per capita expenditure group with a factor loading of 0.80 is the dominant variable. In the dimension of food use, the most dominant variable is the number of household members who experience health complaints with a loading factor of 0.89. The dimension of access to food is the education of the head of the household with a loading factor value of 0.64.

Conclusion
The concept and scope of dimensions regarding food security are quite diverse. The dimension of food security requires a variable that can explain the concept of food security in more detail. Meanwhile, realizing food security at the national level begins with efforts to improve food security at the household level. Thus, increasing food security will be more targeted if it is based on household characteristics that can explain each food security factor.
The results of the food security analysis using the CFA method obtained a set of variables incorporated in the 4 dimensions of food security coverage. Here are the variables in each factor/dimension: 1. The dimensions of food stability include the marital status of the head of the household, the sex of the head of the household, and the number of household members. The dimension of food availability, the most dominant variable is the status of the head of the household with a factor loading of 0.96. 2. Food Availability Dimensions include household expenditure groups Calorie consumption/capita/day) Household BPNT revenues in quintile 2 and rice prices. Dimensions of food access, per capita expenditure group with a loading factor of 0.80 are the dominant variables 3. The dimensions of access to food to obtain food consist of the status of the area of residence, the status of the head of the household in the formal sector, the age of the head of the household, and the education of the head of the household. The dimension of access to food with the dominant variable of education for the head of the household has a loading factor of 0.64 4. The dimensions of the use of food to obtain food consist of the number of household members experiencing health complaints, the number of household members who have been outpatient in the last month. In the dimension of food use, the most dominant variable is the number of household members who experience health complaints with a loading factor of 0.89.

Suggestion
Efforts to achieve food security in DIY in 2030 can be achieved through realizing food security from the household level. This can be done by: 1. Optimizing food security targets by taking into account household characteristics that can explain the dimensions of food security. These characteristics include the following:  The dimension of food stability includes household characteristics, namely the marital status of the head of the household, the sex of the head of the household, and the number of household members.  Dimensions Food availability includes household expenditure groups Calorie consumption/capita/day) Household BPNT receipts in quintile 2 and rice prices.  The dimensions of access to food to obtain food consist of the status of the area of residence, the status of the head of the household in the formal sector, the age of the head of the household, and the education of the head of the household.  The dimensions of the use of food to obtain food consist of the number of household members experiencing health complaints, the number of household members who have been outpatient in the last month. 2. Explanation of the results of this research can be combined with the mapping of the food security status of each region.