Emigration and remittances are not linked to youth school attendance

As discussed earlier, emigration and remittances can affect children's education in several ways. These links are investigated here for Georgia.

The empirical literature on the link between migration and education in Georgia is limited and shows somewhat mixed effects. One study using data from the early 2000s found no association between migration and the level of household spending on education in Tbilisi (Dermendzhieva, 2011), while other studies using more recent data found a positive relationship between remittances and educational expenditures in Georgia (Gugushvili, 2013; Chappell et al., 2010).

As shown in the first section of this chapter, primary school enrolment rates are high in Georgia. The analysis of the link between remittance receipt and education therefore focuses on school attendance for the 15-17 and 18-22 age groups (Figure 6.3). Young people in the 15-17 year old group living in households receiving remittances are slightly less likely to be in education, while the pattern is reversed for youth in the 18-22 year old group. These differences are however not statistically significant.

Figure 6.3. Remittances show little effect on youth school attendance

Share of youth attending school by household remittance status

Note: Results that are statistically significant (using a chi-squared test) are indicated as follows: ***: 99%, **: 95%, *: 90%. Source: Authors’ own work based on IPPMD data.

Sta.tLink^^2 http://dx.doi.org/10.1787/888933457972

The link between migration and educational attendance was analysed in more depth using regression analysis that controls for other individual and household characteristics (Box 6.2). The results confirm that migration and remittances are not linked to youth school attendance. The results are not statistically significant for either age group (15-17 years and 18-22 years; Table 6.4).

Box 6.2. The links between migration, remittances and youth school attendance

A regression framework was used to estimate the effect of migration and remittances on education attendance using the following equation:

Prob(edu _ attendancej) = f}0 +ft1emighh + fl2remithh + y1controlsj + y2controlshh + 5r + sj (2)

Where the dependent variable edu _ attendancei is education attendance of youth in the two age groups: 1) 15-17 years old and 2) 18-22 years old. emighh represents a binary variable for emigration, where “1” denotes if the youth lives in a household with at least one emigrant and “0” if not, while remithh represent a binary variable for remittances taking the value “1” if the household in which the youth lives is receiving remittances and 0 if not. controlsj and controlshh are sets of observed individual and household characteristics influencing the outcome, and ej is the randomly distributed error term. controlsj and controlshh are sets of observed individual and household characteristics believed to influence the outcome.1 Sr represents regional fixed effects and is the randomly distributed error term.

Table 6.4. Migration and remittances do not influence school attendance

Dependent variable: Youth education attendance

Main variables of interest: Household has emigrant/receive remittance/has return migrant

Type of model: Probit

Sample: Youth aged 15-17 and 18-22

Sample

Variables of interest

(1)

(2)

Youth

Youth

aged 15-17

aged 18-22

Household has at least one emigrant

-0.078

0.060

(0.058)

(0.060)

Household receives remittances

-0.023

-0.023

(0.066)

(0.065)

Number of observations

242

498

Note: Results that are statistically significant are indicated as follows: ***: 99%, **: 95%, *: 90%. Standard errors are in parentheses and robust to heteroskedasticity. Specification including a sample of youth aged 15-22 years old was also carried out but did not generate any statistically significant results.

1. The set of independent variables includes age and sex of the youth, a binary variable indicating if the household in which the youth lives is located in the capital, the household’s dependency ratio (i.e. the share of teenagers, children and elderly in the household in relation to members of working age), the total number of children in the household, the number of children in the age ranges 6-14 and 0-14 respectively, the male-to-female ratio and a household asset wealth index (measured through an asset index using principal component analysis).

 
Source
< Prev   CONTENTS   Source   Next >