1 / 44

Evaluation: Experiments, Matching and D-in-D

Evaluation: Experiments, Matching and D-in-D. Lecture 5. Agenda. Work our way through a bit of Kremer and duFlo … Experiments Matching DD Migration, Mothers and Money …. Migration, Mentoring and Mothers: The Effect of Migration on Children’s Educational Performance in Rural China .

galena
Download Presentation

Evaluation: Experiments, Matching and D-in-D

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Evaluation: Experiments, Matching and D-in-D Lecture 5

  2. Agenda • Work our way through a bit of Kremer and duFlo … • Experiments • Matching • DD • Migration, Mothers and Money …

  3. Migration, Mentoring and Mothers: The Effect of Migration on Children’s Educational Performance in Rural China Stanford University QQ Huang (U. of Minn.) and Linxiu Zhang (CCAP) July 2007 Scott Rozelle and Xinxin Chen

  4. Introduction/Motivation Migration is one of the main ways of alleviating poverty in developing countries Migration itself, however, is not costless. • For example: There may be an adverse effect of migration on the educational achievement of the children of migrants (McKenzie et al. on Mexico)

  5. Overall Increase in Off-farm Work In 2000: 45% of rural labor force have jobs off the farm … more than 80% of households have at least 1 person working off the farm 51% 2005 In 1980: only 4% worked full time off the farm

  6. Migration-fastest growing segment Migrant Wage Earners TVE/Local wage earning local Self employed migrant

  7. Summary: Migration in China • Migration is rising fast, surpassing 100 million individuals (deBrauw et al., 2002) • Migrants also are moving further away from home and leaving for a longer period of time (Rozelle et al., 1999). • Most of China’s migration is by individuals instead of entire households, in most cases the school-aged children of the migrant parents are being left. • Work in China: Migration  higher income / poverty alleviation

  8. Introduction/Motivation Migration is one of the main ways of alleviating poverty in developing countries Migration itself, however, is not costless. • For example: There may be an adverse effect of migration on the educational achievement of the children of migrants (McKenzie et al. on Mexico)

  9. Results from current literature • School performance of the migrant children is being adversely affected by migration since parental care falls with migration (Wang and Wu, 2003; Tan and Wang, 2004; Li, 2004; Zhou and Wu, 2004). • These results are all based on casual observation • Are they true? • Is there anything about migration that can offset this effects?

  10. The Trade-off: Parents can stay at home and can take care of their kids …

  11. But often good jobs are hard to find and people are very poor … sometimes they can not supply their kids with warm clothes and enough nutrition

  12. Although migrants to China’s cities typically earn more money than those left in the village (Giles and deBrauw / Park and Du)

  13. Parents are often far from their kids …

  14. … and, their children are left behind … sometimes with their grandparents … and sometimes at boarding schools …

  15. Objectives Examine the effect of migration activities of men and women on the educational performance of their children. • Compare the distribution of children’s scores for different types of rural households and describe how the grades vary over time. • Examine whether migration negatively affects the school grades of rural children. • Explore how migration will affect children’s educational performance in different types of households in terms of wealth or demographic composition.

  16. Limitations / Contribution • Only one region of China • Counterfactual limited … do not observe children that migrate with their parents … • May be mainly looking at correlations • Still say something about what happens to grades of children of migrants when they leave them in the countryside

  17. Data • A data set collected in 2006, with information of changes in school performance of children before and after their parents outmigrated. • 1649 fifth grade students in 36 primary schools in 6 counties in Shaanxi province • Random sample of towns and schools within the counties and classes within the schools … but surveyed ALL students within each class …

  18. The Sample (6 counties)

  19. Two Key Variables • Grades of school achievement • math and Chinese language scores • scores from 2001/2 (first grade) to 2005/6 (fifth grade) • From records kept by students + schools • standardized scores (second term scores) [all scores from standardized tests corrected by joint grading panel of teachers] • Migration status • migration histories of each parent between 2002 and 2006

  20. Time lines of academic calendars from 2001/2 to 2005/6 Grade 1 second term scores Grade 5 second term scores Never Migrant Households 2001 2002 2003 2004 2005 2006 Grade 1 second term scores Grade 5 second term scores New Migrant Households 2001 2002 2003 2004 2005 2006 New migrant = parents were both home in 2002; at least one or both parents outmigrated by 2005

  21. Different types of new migrant households by migration status Any Parent Migrated households: households in which both parents lived at home in 2002 and at least one parent –either the father; mother or both parents—outmigrated in 2006; Father Migrated Only (or mother-stayed-at-home) households: households in which only the father outmigrated in 2006 but was at home in 2002; Father Migrated (Unconditional ) households: households where the father was at home in 2002 but outmigrated in 2006 (including households in which the mother was either at home or not at home in 2006); Mother Migrated Only (or father-stayed-at-home) households: households where only the mother outmigrated in 2006 but was at home in 2002; Mother Migrated (Unconditional) households where the mother was at home in 2002 but outmigrated in 2006 (including households in which the father was either at home or not at home in 2006); Both Parents Migrated households: households where both parents were at home in 2002, but outmigrated in 2006. New Migrant households Never Migrant households.

  22. Patterns of Migration for Sample Households in China, 2002 and 2006 Migration status, 2006 Migration status, 2002 Never migrants, 2006 Never migrants, 2002 131 35 54 1138 1594 Total sample size

  23. School Performance Figure 1 Average Yearend Test Scores in China, 2002 and 2006 The average grades of all children fell between 2002 and 2006 … one interpretation is that grades fell as more migration … alternative explanation: fifth grade is graded “harder” than first grade?

  24. The effect of migration on school performance seems to be complicated and that care needs to be exercised in any interpretation. Figure 2. Differences in Yearend Test Scores between First Grade Students from Migrant and Non-migrant Households in Rural China, 2006 If one went out and found a family in 2006 in which neither parent outmigrated in 2002 and both parent outmigrated in 2006 … and compared to Never Migrant … might be inclined to blame migration!

  25. The effect of migration on school performance seems to be complicated and that care needs to be exercised in any interpretation. Figure 2. Differences in Yearend Test Scores between First Grade Students from Migrant and Non-migrant Households in Rural China, 2006

  26. Figure 3. Kernel Density Plots of Distributions of Average Test Scores in Never Migrated Households and Both Parents Migrated Households,2002 and 2006 the gap was narrowing

  27. Figure 4. Kernel Density Plots of Distributions of Average Test Scores in Households that Vary by Wealth Category and Household Composition in China, 2002 and 2006

  28. Methodology (1) Difference in Difference (DID) Model (1), Restricted & Unadjusted: ΔScorei = α + δMIGi + εi Model (2), Unrestricted & Unadjusted: ΔScorei = α +δMIGi +γScore_02i+ εi Model (3), Restricted & Adjusted: ΔScorei = α +δMIGi +βXi+ εi, Model (4), Unrestricted & Adjusted: ΔScorei = α +δMIGi +γScore_02i +βXi+ εi where, i is an index for the student, ΔScorei is the change of the second term score of student i between 2002 and 2006 (that is the final grade from the fifth grade minus the final grade from the first grade); MIGi is the treatment variable (which makes δ the parameter of interest). Finally, the term Xiis a vector of covariates that are included to capture the characteristics of students, parents and households and also includes a set of 12 town indicator or dummy variables.

  29. Model #2: Difference-in-differences • Compares before-after changes of participants vs. before-after change of non-participants • Any common trends get differenced out. • Limitation: only common trends between two groups get differenced out • We control for base value of Yi and observables

  30. Our equation of choice (the full model) • Model (4), • Unrestricted & Adjusted: ΔScorei = α +δMIGi +γScore_02i +βXi+ εi

  31. Methodology (2) • Propensity Score Matching (PSM) ----Basic matching ----Multi-dimensional matching • Difference in Difference Matching (DDM) ----Basic matching ----Multi-dimensional matching

  32. Results • DD results • PSM results • DDM results

  33. Table 2. DD Regression Results Analyzing the Effect of Migration on School Performance of Students in China

  34. Table 2. DD Regression Results Analyzing the Effect of Migration on School Performance of Students in China

  35. Table 2. DD Regression Results Analyzing the Effect of Migration on School Performance of Students in China Highest R-square: > 0.40

  36. Table 3. DD Regression Results Analyzing the Effect of Migration on School Performance of Students in China by Household’s Migration Status

  37. Table 4. PSM and DDM Estimators and the Effect of Migration on the School Performance of Students in Rural China, 2002 and 2006

  38. Table 4. PSM and DDM Estimators and the Effect of Migration on the School Performance of Students in Rural China, 2002 and 2006

  39. Table 4. PSM and DDM Estimators and the Effect of Migration on the School Performance of Students in Rural China, 2002 and 2006

  40. Summary • There is no evidence that migration in our sample of households has hurt school performance. • In fact, when the father outmigrates (either by himself or with others), migration appears to have a small, positive effect on the school performance of migrant children.

  41. Heterogeneous effects • Heterogeneous Effects from Wealth Model (5): ΔScorei = α +δ1MIGi +δ2*MIGi *poor+γScore_02i+βXi+ εi, • Heterogeneous Effects fromHousehold Composition • Model (6): ΔScorei = α +δ1MIGi +δ2*MIGi *nosibling+γScore_02i +βXi+ εi,

  42. Table 5. DD Regression Results with Heterogeneous Effects from Wealthand Household Composition

  43. Conclusion • We can reject the hypothesis that migration harms the grades of their children . • In fact, the migration of some migrant households has a statistically significant and positive effect on the performance of the children. • There is neither a systematically different effect of migration between the children of more wealthy and less wealthy households nor between the children from families that have one and more than one child.

  44. Policy implications • It is not that migrants do not need to have special attention in education … their grades are lower (but they are always lower) … increased education will raise their productivity (other studies) … • Point of our paper: migration by itself does not cause this … although we have not identified that exact mechanism, may be that the income effect of migration is offsetting the parental care effect • So should build better schools … have high quality boarding facilities … increase mentoring inside schools (e.g., by small classes) … and promote the admittance of rural students to urban schools (for low or no tuition) … but, don’t do it because believe migration leads to lower education achievement … there is no evidence from our sample

More Related