Educational attainment EA is represented by a categorical variable where each individual is assigned a particular level of educational attainment, beginning with the value of 0 for those who have not completed primary education, the value of 1 for those who completed primary education, and so on, up to the value of 8 for those who have completed a doctoral degree.
In Indonesian formal education system, there are eight levels of education, beginning from primary schools with a duration of six years. However, for the purpose of this study, individuals with uncompleted primary education are also included. These individuals may also participate in the labor force.
The next level is junior high schools ISCED level 2 with a length of study period of three years. The twelve-year education up to high schools level is a free compulsory education with government funding.
This twelve-year compulsory education begun in has successfully resulted in a high net enrollment rate at Marital status MAR is represented by a binary variable indicating the marital status of an individual. The value of 1 is assigned to married individuals and 0 otherwise. The number of children under five KID variable is a proxy for fertility rate of the household. The health status is proxied by several types of health problems suffered by individuals. They consist of difficulties in vision, difficulties in hearing, difficulties in mobility, i.
Thus, DIF variable indicates the total number of types of disability suffered by an individual. Building upon Mincer that states wage is a function of education and experience, we construct the hypothesis that education, training, and experience determine a decent standard of living. This is based on the stylized facts about the positive effects of education and training Black and Lynch ; Himaz and Aturupane as well as of experience Eraut ; Medoff and Abraham In this study, the decent standard of living is treated as the dependent variable DSL, defined as the ratio of an individual's monthly earnings distributed to each household member to the amount of the family living wage distributed to a reference family.
DSL has a value greater than or equal to 1 if the amount of the individual monthly earnings distributed to each household member is greater or equal to the family living wage distribution with a four members household, in accordance with the reference family size suggested by Anker and Anker It is important to note that decent standard of living is, in accordance to Adams , achieved through living wages and not minimum wages. Trading Economics, a provider of economic data and information, discloses an estimate of the value of family living wages in Indonesia in at 2,, rupiahs per month.
This means an average distributed income of , rupiahs per month per family member. Thus the DSL variable is valued at1 if an individual is able to distribute his monthly income to each of his family member in the amount of at least , rupiahs per month. In this study the term monthly income refers to the value of money from income earned in a month by individuals whose employment status falls within the category of own-account workers and casual workers, or alternatively the value of money from wages and benefits received in a month by individuals who belong to the status of employees.
One should note that with DSL variable being so defined, one implication is that it is observed only if the individual is working, but unobserved if otherwise.
As a determinant of decent living standards, education is aggregated so that it is categorized into three groups of educational attainment by level; primary education level ISCED level 1 , secondary education level ISCED level 2 to 4 , and tertiary education level ISCED level 5 to 8. Adopting the method of Greene , the educational attainment variable at each level is defined as a binary variable which is assigned the value of 1 for the highest level of education attained by an individual.
Furthermore, training is defined as a form of skills acquisition, so that individuals are equipped with specific skills specific to certain individuals, and therefore they gain competitive advantages in the labor market. Training can take place in the workplace to improve work-related skills for individuals. In this study we define the training variable TRA as the one of the above-mentioned type. The purpose is to distinguish it from more generally defined training. In this context, training is based on an individual's participation in work-related training programs, either certified or not.
Finally, with regards to experience variable, this study directly follows Gardeazabal and Ugidos , where experience is proxied by age, assuming that individuals tend to gain more work experience with age.
The Model The achievement of a decent standard of living involves a two-step decision, namely the individual's decision to participate in the labor force, and actually work to earn income that meets the needs for a decent living.
The Heckman method is employed to correct the bias caused as is commonly used when non-randomly selected samples is existent in an estimation of behavioral relationships. The model - also called the Heckit model - consists of two equations. The first is the selection equation to determine whether the variable of interest is observed. The first sample of this study, denoted by N, consists of , observations, but since the variable of interest was observed only for , individuals, so the latter, denoted by n, is less than N.
The selection equation is then expressed in the latent variable, which in this case indicates the individual's decision to participate in the labor force. It is bound to several explanatory variables. Based on the literature cited in Section 2, we argue that an individual's decision to participate in the labor force is determined by their educational attainment EA , marital status MAR , number of children under five in the household KID , and disability status DIF.
In addition, age cohort is also considered a demographic factor that may influence differences in the patterns of work participation of men and women. We therefore added a categorical variable ACO that represents individual's age cohort to investigate the likely variations in the labor force participation due to gender differences for various age groups.
We used 9 age cohorts, consisting of age cohort 1 for individuals aged years, and so on up to age cohort 9 for individuals aged years. Opportunities for having a decent standard of living for individuals are assumed to be affected by regional economic conditions where an individual lives. Following Grunewald that uses regional poverty rates to proxy for regional economic conditions, we add a categorical variable POV, assumed to vary across provinces.
Cluster 1 consists of the regions with low poverty rates, cluster 2 medium, and cluster 3 high. We used the Stata software which specifically provides the Heckit module and this is important to avoid incorrect standard errors and t-statistics that is generated after estimation. Following Hardy and Reynolds , we employed a regression technique for an estimation that involves categorical information in the model.
Results and Discussion 5. Descriptive statistics Table I presents a brief summary descriptive statistic of the variables used in the study. Since most of the dependent variables and independent variables are binary and categorical, they are observed to have relatively small standard deviations, implying that the observed values are less spread out from their means.
Using large sample size also implies a small variability in observed values of most of those variables. Descriptionofthe variables Variables Mean S. Labor Force Participation: the first stage analysis Tables II and III and Figure 1 present the predictive value of individual labor force participation decisions based on the results of Probit regression estimation analysis for the selection equation 1 by gender and for various age cohorts. The educational attainment positively explains the labor force participation of men aged 15 to 74 years.
The positive effect of educational attainment was found to be weakest 0. We also found that for men aged 75 years and older, the educational attainment does not have a positive influence on their work participation. Table II. The predicted value of male labor force participation for each determinant by the age cohort Indep. Male Var.
The predicted value of female labor force participation for each determinant by the age cohort Indep. Female Var. In general, the educational attainment is a major determinant of women's work participation in Indonesia and this is consistent with the finding of Psacharopoulos and Tzannatos that education is a potential driver of women's labor participation in developing countries. Furthermore, we find that marital status is a major determinant of male work participation in all age cohort.
The magnitude of the positive effects of marital status on male work participation was found to be highest 2. In line with Hill , this finding reflects the strong influence of patriarchal culture in Indonesia which gives men a role as the main breadwinner for their household. Marital status has a positive effect on women's work participation when they enter the age range of years, but with a smaller magnitude of effect compared to those of men.
The participation of married women in the labor market could indicate that their support is needed to improve family income that are deemed inadequate if they only rely on their spouses, as in Belle and Tebbets This argument is also based on the compensating differentials hypothesis proposed by Grossbard-shechtman and Neuman which states that married women will participate in the labor force if their material needs are not satisfied by marriage.
Variation in the predictive value of male and female work participation for each determinant by age cohort The presence of children under five KID in the household positively explains the work participation of men in the age range of years, but it is negative for men aged 75 years and older.
This finding implies that marriage motivates men's labor participation. It can be influenced by a patriarchal culture in which men are viewed as breadwinners responsible for the material needs of their dependents, especially those who are in an economically productive age range. Nevertheless, there is still a positive effect of the presence of children under five in the household on the labor participation of women, especially of those in the age range of years.
These results corroborate the finding of Duncan, Prus, and Sandy that the presence of toddlers in the household would prevent women's work participation. However, it does not necessarily hold true for women of any age group. Furthermore, health problems or disabilities negatively affect the labor force participation of men and women aged years.
This negative effect was also found in men over 65 years, and women over 75 years. In other words, for men in the age range of years and women in the age range of years, these health problems do not hamper their work participation. The positive sign of the influence of health problems on work participation is higher in men than in women. This again implies that men tend to be more responsible relative to women in providing livelihood support for their families even in the event of health problems they may face.
In general, an inverted U shape patterns Figure 1 describes the relationships between the predictive value of labor force participation and its determinants. A similarity in the relationship patterns for men and women can also be seen, except the fact that there exist differences at which age cohort men's and women's maximum values occur.
Wald test results which show significant values indicate good suitability of the model. The correlation between error terms of the selection equation and the outcome equation, which is indicated by the significance of the value of the Wald test results on the independence of the two equations, provides justification for the use of the Heckman selection method.
The estimation results of the Heckman model can be seen in Appendix. The negative sign of the IMR coefficient indicates that the estimation of the determinants of a decent standard of living would be downward biased without this kind of correction. Table IV presents the predicted values of the ratio of decent standard of living for each determinant according to gender and poverty rates. Figure 2 shows the patterns of the effects of various determinants of decent standard of living for men and women by provincial poverty rates.
Table IV. Predicted values of the ratio of decent standard of living for each determinant by provincial poverty rates Indep.
Male Female Var. The effect becomes smaller as they live in regions with higher poverty rates. For male individuals living in low poverty rates regions, the effect of secondary education attainment on the ratio of decent standard of living is positive and higher at 0.
It is also higher for males living in higher poverty rates regions. The highest predicted value of the ratio of the decent standard of living is contributed by the tertiary education level at 1. This is true for men living in regions regardless of poverty rates.
We find a somewhat similar pattern for women in terms of the effect of education on the ratio of decent standard of living that can be achieved.
The higher the level of education the higher the standard of living of women. However, the positive predicted value of the ratio of decent living standards explained by tertiary 0.
The regional condition as reflected by regional poverty rates negatively affects the benefits of educational attainment. The regional condition - reflected by provincial poverty rates - where women live negatively affects the benefits of women's educational attainment. Primary educated women living in the regions with medium and high poverty rates have more than three times larger negative effects on their decent standard of living. The above-mentioned findings emphasize the importance of education for any effort to achieve a decent standard of living, both for men and women.
Participation in trainings has a significant positive effect on efforts to achieve a decent standard of living with men 0. However, the effect of acquired trainings decreases with increasing poverty rates of regions where they live. The effect of experience becomes very small for those who live in regions with medium and high poverty rates.
This finding implies that experience does not provide a meaningful influence on decent standard of living if they do not invest in education and training. Figure 2. Variation in the predicted value of the ratio of a decent standard of living for men and women for each determinant by regional poverty rates 6.
The global average rate of return to schooling, estimated at 10 per cent, is used as a global benchmark. Empirical … Expand. The paper examines the role of education in economic growth from both a theoretical and historic perspective, addresses why education has been the limiting factor determining growth historically, … Expand. Heterogeneous Returns to Education in the Labor Market. Since the development of human capital theory, countless estimates of the economic benefits of investing in education for the individual have been published.
While it is a universal fact that in all … Expand. View 3 excerpts, cites background. This study examines the developments in educational attainment and education returns of wage earners in Thailand over the period — In so doing, several potential weaknesses of previous … Expand. Typically estimates of the benefits of education investments show average private rates of return for the average individual.
The average may not be useful for policy. An examination of the … Expand. View 4 excerpts, cites background. Comparing two approaches to the rate of return to investment in education. The economic value of investment in education has typically been measured by its rate of return, frequently estimated by the internal rate of return or the earning function approach. Given the … Expand. Assessing the Returns to Education in Georgia.
The economic returns to education in transition countries have been extensively evaluated in the literature. The present study contributes to this literature by estimating the returns to education in … Expand. View 1 excerpt, cites results. Background This paper was prepared to contribute to a special issue on the value of education.
Purpose The paper examines the role of education in economic development from both a theoretical and a … Expand. The role of education in economic growth: theory, history and current returns. Background This paper was prepared to address the issue of whether current levels of public expenditures on education are cost-effective in countries with widely differing average levels of … Expand.
View 2 excerpts, cites background. Economic Volatility and Returns to Education in Venezuela: Preliminary evidence suggests that the rates of return to education in Venezuela have been declining since the s. This study rigorously estimates the returns to education in Venezuela for the … Expand. View 1 excerpt, cites background. This paper uses data from three household surveys in , and to investigate the relationship between earnings and education in Mexico.
The time series feature of the data allows us to … Expand. Returns to investment in education: A global update. The author updates compilations of rate of return estimates to investment in education published since - and discusses methodological issues surrounding those estimates. Some key patterns: among … Expand. View 2 excerpts, references background. Education in Production. Journal of Political Economy. There have been several studies of the demand for education as an investment good1 which generally take input and product prices as given and concentrate on computing internal rates of return to … Expand.
View 1 excerpt, references background. Abstract The accepted wisdom is that rates of return to lower levels of schooling are higher than to university and tend to stay higher even as a country develops.
But increasing evidence suggests … Expand.
0コメント