Türkiye'de ortaöğretime kayıt olma kararının belirleyenleri

Becker indicates in Human Capital (1964) that ?education and training are the most important investments in human capital?. The more the people are educated, the more the human capital improves and every economical system in development needs more human capital to improve itself. This study investigates Turkey?s secondary level education as it is the second level of the education that requires to be examined. In 1997, Turkey has accepted to change the primary education system into eight years of compensatory education to improve the rate of school enrollment. Since then, high level of primary school enrollment has been reached. However, secondary school enrollment is still unsatisfactory. A large number of children drop out after completing primary school and do not continue to high school. In this study, the determinants of high school enrollment are examined.This paper?s aim is to analyze the determinant factors of high school enrollment in Turkey. The analysis is based on data of Turkish National Ministry of Education data from 2010/2011 education year. The data involves the high school enrollment of the children, their age, the information about the education and the occupation of children?s parents, the number of siblings, the income of household, the entrance test results, the region and place where the children live.The data acquired from Turkish National Ministry of Education includes the information about household and individual characteristics of children graduated from primary school in 2010 / 2011 school year. The dependent variable in this study is that if the children are ?enrolled to high school? or ?drop out after graduating from primary school?. This is not an ordinal, numerical data; it takes the values ?0? for the children who are drop out after graduating primary school and ?1? for the children who enroll to high school. This type of dummy variable data is considered as binary outcome. The binary outcomes can not be estimated with Ordinary Least Squares (OLS) estimation; the OLS assumes that the dependent variable is continuous and unlimited. The independent variables of this study?s estimation, such as the education level of parents, the occupation of parents, the household income quintiles, the regions, et al are also dummy variables.To estimate the factors those affect the high school enrollment / drop out after primary school, logistic regression technique is used. Logistic regression method is a generalized linear model to model binary outcome. This type of regression has been used numerous times by the researchers who analyzed the factors determining the high school enrollment decision (Tansel, 2000; Hoşgör and Smits, 2006; Göksel, 2008; Moyi, 2009). Logit and probit regression models can be used to estimate the determinants of school enrollment, in our case the logit regression technique is applied. Moyi (2009) calculated the odds ratio (which is a measure to compare the effects of independent variables) in his study of Malawian delayed school enrollment. Our study computes the odds ratio of household and individual factors affecting the high school enrollment in Turkey.The dependent variable may take the value 1 with a probability of success p (in this study it is high school enrollment) or the value 0 with a probability of failure (1 ? p (it is the drop out after primary school in this study)). This type of variable (binary variable) is called a Bernoulli variable . In the statistics literature, a common interpretation of the coefficients of logistic regression is in terms of marginal effects on the odds ratio. (p / (1 ? p)) measures the probability that y takes the value 1 relative to the probability that y = 0 and is called odds ratio or relative risk. With the odds ratio, logistic regression may predict the probability of success or failure, which is in our study school attainment or drop out after primary school. In this study, the odds ratios for each factor are calculated and are interpreted considering the last researches of school enrollment and educational participation in Turkey (Tansel, 2000; Hoşgör and Smits, 2006; Göksel, 2008).Our analyses show that in 2011, about %20 of the children drop out the education after graduating from primary school. This is one fifth of 1.189.154 children graduated from primary school which is a very important and large number of children that requires attention.Since 1997, after the compensatory eight years policy for primary school has been accepted by Turkish government, primary school enrollment has been increasing and almost reached 98% rate of enrollment. In 1997, primary school and middle school has been combined and became primary education. Since then, high school became the secondary level education. In 2006, the educational years of high school has been extended to four years, thus the national education became twelve years of education.Even though high levels of primary school enrollment are reached, high school enrollment rate in Turkey is not satisfactory. Therefore the aim of our study is to examine the reasons of the drop out rates after graduating from primary school. The household characteristics and regional determinants are mostly significant on high school enrollment decision.We find out that parental education and parents? occupation are determinative factors on high school demand, number of sibling are also determinative on high school demand in Turkey contrarily to Patrinos and Psacharapoulos? study in Peru (1997). However more research is needed to control the decision of family to enroll their child in high school; cultural, demographic and regional reasons may affect the decision of children?s high school enrollment.other?s education status is one of the most prominent determinants of their daughter?s high school enrollment decision. If their education level is higher than eight years, their children?s probability of getting enrolled to high school is higher than the children of illiterate or primary school graduated mother?s. This is also the case for the children of fathers who are educated for more than eight years. However the effects of fathers? education levels are not as prominent as mothers? education level.The occupation of parents is a decisive factor. Every occupation type other than being unemployed has positive effects on children?s high school enrollment except the mothers working on private sector. This may be due to working hours of private sector and mothers may not take care of their children enough. The working hours of private sector in Turkey are very extensive; in service sector the shift are long and in some occasions both parents may work in long hours. Considering the results of this study, the odds ratio for mothers in private sector indicates that, working in private sector may affect negatively the daughters? and all the children? high school enrollment.Father?s occupation except the ?other? section is positively affecting the children?s high school enrollment. If the father is self employed, this may affect negatively his daughters? high school enrollment but not his sons? enrollment. According to Tansel?s study (2002), if father is self employed he might also be less educated and he may decide not send his children to high school. However in our results, it only affects the decision of daughter?s high school enrollment.The place where the children live is important. Urbanization is an outstanding indicator of development and the children living in town centers or cities have more probability to get enrolled in high school than the children living in rural areas and squatter settlements. Regional differences are also remaining; however the children from Southeastern Anatolia region and the boys from Black Sea region have more probability of getting enrolled to high school if they have the same features as the children in Eastern Anatolia (the reference group in the analysis). Other regions (Marmara, Aegean, Mediterranean and Central Anatolia) do not have the same results.The income of household is a determinative factor on high school enrollment. The children from families other than the lowest income quintile have more probability to getting enrolled to high school. However the boys from second wealth quintile have less probability than the boys from lowest wealth quintile, this may be due to support their family income. The children living in lowest wealth quintile and probably the boys living in second wealth quintile may prefer to work instead of going to high school, mostly to support the family income.The families with numerous siblings do not take care equally of their children. It can be said that if the number of children in a family increases, the education of every child will be different. This study?s results show that the more the siblings in a family the probability of a child to get enrolled to high school decreases. The older siblings in those families may also work to support their family income or may take care of their younger siblings if both of their parents are working.Understanding these results may help policy makers to improve the situation of the children living in the regions where the education is insufficient and to invest in education.