pith. machine review for the scientific record. sign in

arxiv: 2604.19969 · v1 · submitted 2026-04-21 · 💰 econ.GN · q-fin.EC

Recognition: unknown

Educational Mobility Across Multiple Generations in Indonesia

Sarah Cattan , Antonio Dalla-Zuanna , Jan Stuhler , Po Yin Wong
Authors on Pith no claims yet

Pith reviewed 2026-05-10 00:29 UTC · model grok-4.3

classification 💰 econ.GN q-fin.EC
keywords educational mobilitymultigenerational correlationsIndonesiafinancial constraintsmarital sortingAsian financial crisisdeveloping countries
0
0 comments X

The pith

In Indonesia, multigenerational educational mobility exceeds what parent-child correlations alone predict.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Standard intergenerational measures have been shown to understate long-run persistence of advantages in developed countries. This paper studies whether the same pattern holds in Indonesia using education data across three generations from the Indonesian Family Life Survey and Census. Contrary to prior findings, the authors observe greater mobility over multiple generations than two-generation correlations would suggest. They introduce a theoretical framework centered on financial and credit constraints plus cultural norms around marital sorting. Regional differences in exposure to the 1997-98 Asian financial crisis and in local marital customs are used to trace the relevance of these two channels.

Core claim

The paper establishes that in Indonesia the correlation in education between grandparents and grandchildren is weaker than the product of successive parent-child correlations would imply, producing more multigenerational mobility than two-generation data alone indicate. This pattern is explained by a model in which binding financial constraints limit the transmission of educational advantages and in which norms governing who marries whom further weaken persistence across generations. Variation in crisis exposure across regions and differences in marital customs provide evidence that both mechanisms operate.

What carries the argument

A theoretical framework that isolates the roles of financial and credit constraints and of cultural norms on marital sorting in shaping multigenerational education transmission.

Load-bearing premise

Regional differences in exposure to the 1997-98 Asian financial crisis and in marital customs can be treated as exogenous identifiers of the causal effects of financial constraints and cultural norms.

What would settle it

Finding that the gap between two-generation and three-generation mobility estimates shrinks to zero once regional crisis exposure and marital customs are controlled for, or observing no systematic difference in mobility between areas more and less affected by the crisis.

read the original abstract

Standard intergenerational measures have been shown to understate the long-run persistence of socioeconomic advantages in developed countries. We study theoretically and empirically whether this pattern extends to less developed settings, using Indonesia as a case study. Using the Indonesian Family Life Survey (IFLS) and Census data, we study multigenerational correlations in education across three generations. Contrary to previous findings, we observe greater multigenerational mobility than parent-child correlations alone would suggest. We develop a theoretical framework to highlight two key factors influencing multigenerational dynamics in developing countries: (1) financial and credit constraints, and (2) cultural norms related to marital sorting. To confirm their relevance, we exploit regional variations in exposure to the 1997-98 Asian financial crisis and in marital customs.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper studies multigenerational educational mobility in Indonesia using IFLS and Census data. It reports that grandparent-grandchild education correlations are lower than the persistence implied by parent-child correlations alone, indicating greater multigenerational mobility than observed in developed countries. A theoretical framework identifies financial/credit constraints and cultural norms around marital sorting as key mechanisms; these are tested via regional variation in exposure to the 1997-98 Asian financial crisis and in marital customs.

Significance. If the central empirical comparison holds, the result provides evidence that the understatement of long-run persistence documented in rich countries does not generalize to this developing-country setting, and it isolates two concrete channels (credit constraints and assortative mating) that can be tested with regional shocks. The reliance on public datasets (IFLS, Census) and the explicit theoretical derivation of the two mechanisms are strengths that facilitate replication and extension.

major comments (2)
  1. [Abstract / empirical results] The central claim (greater multigenerational mobility than parent-child correlations would suggest) requires that the parent-child education correlation measured in the three-generation subsample equals the benchmark correlation used for the AR(1) extrapolation. The abstract and empirical strategy do not report this comparison or any test for differential selection into the three-generation sample (e.g., by family stability, location, or survival). If the subsample correlation differs, the 'greater mobility' conclusion does not follow.
  2. [Theoretical framework and identification] The identification of financial constraints and marital-sorting effects rests on the assumption that regional variation in 1997-98 crisis exposure and in marital customs is exogenous to unobserved determinants of multigenerational mobility. No balance checks, pre-trend tests, or discussion of potential confounders (e.g., differential regional trends in schooling supply) are referenced in the provided abstract or theoretical setup.
minor comments (2)
  1. [Data section] Clarify the exact construction of the three-generation education measures (e.g., how missing or imputed values are handled and whether years of schooling or attainment categories are used).
  2. [Results] Add a table or figure showing the parent-child correlation both in the full sample and in the three-generation subsample to make the key comparison transparent.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed comments. We address each major comment below and indicate the revisions we will make to strengthen the manuscript.

read point-by-point responses

  1. Referee: [Abstract / empirical results] The central claim (greater multigenerational mobility than parent-child correlations would suggest) requires that the parent-child education correlation measured in the three-generation subsample equals the benchmark correlation used for the AR(1) extrapolation. The abstract and empirical strategy do not report this comparison or any test for differential selection into the three-generation sample (e.g., by family stability, location, or survival). If the subsample correlation differs, the 'greater mobility' conclusion does not follow.

    Authors: We agree that confirming the parent-child correlation in the three-generation subsample matches the benchmark correlation is necessary for the central claim to hold. In the revised manuscript we will explicitly report this comparison using the IFLS data. We will also add tests for differential selection into the three-generation sample along the dimensions of family stability, location, and survival, and discuss any implications for the multigenerational mobility results. revision: yes

  2. Referee: [Theoretical framework and identification] The identification of financial constraints and marital-sorting effects rests on the assumption that regional variation in 1997-98 crisis exposure and in marital customs is exogenous to unobserved determinants of multigenerational mobility. No balance checks, pre-trend tests, or discussion of potential confounders (e.g., differential regional trends in schooling supply) are referenced in the provided abstract or theoretical setup.

    Authors: We recognize the importance of documenting support for the exogeneity assumption. In the revised paper we will add balance checks on observable pre-crisis characteristics across regions with different crisis exposure and marital customs. We will also report pre-trend tests where the data permit and include a discussion of potential confounders such as regional trends in schooling supply, together with evidence or arguments addressing their possible influence on the estimates. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical correlations derived from external data without self-referential reduction

full rationale

The paper is an empirical study that measures multigenerational education correlations directly from IFLS and Census microdata and compares them to parent-child benchmarks within the same samples. The theoretical framework invokes standard economic mechanisms (credit constraints, marital sorting) and uses exogenous regional variation in the 1997-98 crisis and customs for identification; none of these steps define the outcome variable in terms of itself or rename a fitted parameter as a prediction. No self-citation chain, ansatz smuggling, or uniqueness theorem is load-bearing for the central claim. The derivation chain therefore remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests on empirical data from IFLS and Census plus a theoretical framework highlighting credit constraints and marital sorting. No new entities postulated; relies on standard intergenerational mobility measurement approaches.

axioms (2)
  • standard math Intergenerational mobility can be measured using correlations in education levels across generations.
    Standard approach in the field of intergenerational mobility studies.
  • domain assumption Regional variations in crisis exposure and marital customs affect financial constraints and norms independently of other confounding factors.
    Invoked to confirm relevance of the two key factors using the 1997-98 crisis and local customs.

pith-pipeline@v0.9.0 · 5430 in / 1187 out tokens · 37025 ms · 2026-05-10T00:29:28.106615+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

29 extracted references

  1. [1]

    Dynastic Human Capital, Inequality, and Intergenerational Mobility

    Adermon, Adrian, Mikael Lindahl, and M ˚ arten Palme.2021. “Dynastic Human Capital, Inequality, and Intergenerational Mobility.”American Economic Review, 111(5): 1523–48. Ahsan, Md Nazmul, M. Shahe Emran, and Forhad Shilpi.2023. “Public Primary School Ex- pansion, Gender-Based Crowding Out, and Intergenerational Educational Mobility.” The World Bank Polic...

    2021

  2. [2]

    The Transmission of Inequality Across Multiple Generations: Testing Recent Theories with Evidence from Germany

    Braun, Sebastian Till, and Jan Stuhler.2018. “The Transmission of Inequality Across Multiple Generations: Testing Recent Theories with Evidence from Germany.”The Economic Journal, 128(609): 576–611. 37 Brunori, Paolo, Francisco HG Ferreira, and Vito Peragine.2013. “Inequality of Opportu- nity, Income inequality, and Economic Mobility: Some International C...

    2018

  3. [3]

    Like (Grand) Parent, Like Child? Multigenerational mobility across the EU

    Clark, Gregory.2014.The Son Also Rises: Surnames and the History of Social Mobility.Prince- ton University Press. Colagrossi, Marco, B´ eatrice d’Hombres, and Sylke V Schnepf.2020. “Like (Grand) Parent, Like Child? Multigenerational mobility across the EU.”European Economic Review, 130: 103600. Collado, M. Dolores, Ignacio Ortu˜ no-Ort´ ın, and Jan Stuhle...

    2014

  4. [4]

    Economic and Mechanical Models of Intergenerational Transmis- sion

    Frankenberg, Elizabeth, Lynn A. Karoly, Paul Gertler, Sulistinah Achmad, I. G. N. Agung, Sri Harijati Hatmadji, and Paramita Sudharto.1995.The 1993 Indonesian Fam- ily Life Survey: Overview and Field Report.Santa Monica, CA:RAND Corporation. Goldberger, Arthur S.1989. “Economic and Mechanical Models of Intergenerational Transmis- sion.”The American Econom...

    1995

  5. [5]

    Do Labor Market Opportunities Affect Young Women’s Work and Family Decisions? Experimental Evidence from India

    38 Jensen, Robert.2012. “Do Labor Market Opportunities Affect Young Women’s Work and Family Decisions? Experimental Evidence from India.”Quarterly Journal of Economics, 127(2): 753–

    2012

  6. [6]

    Multigenerational Mobility Among Males in India

    Kundu, Anustup, and Kunal Sen.2023. “Multigenerational Mobility Among Males in India.” Review of Income and Wealth, 69(2): 395–418. Kusnanto, Hari.2002. “Regional Differences in the Impact of the Economic Crisis and Social Safety Net on Child Nutrition in Indonesia.” Takemi Program in International Health Working Paper RP197. Lindahl, Mikael, M ˚ arten Pa...

    2023

  7. [7]

    Family Background, Neighborhoods, and Inter- generational Mobility

    Mogstad, Magne, and Gaute Torsvik.2023. “Family Background, Neighborhoods, and Inter- generational Mobility.”Handbook of the Economics of the Family, 1(1): 327–387. Narayan, Ambar, Roy Van der Weide, Alexandru Cojocaru, Christoph Lakner, Sil- via Redaelli, Daniel Gerszon Mahler, Rakesh Gupta N Ramasubbaiah, and Stefan Thewissen.2018.Fair Progress?: Econom...

    2023

  8. [8]

    Dynastic Inequality Compared: Multigenerational Mobility in the United States, the United Kingdom, and Germany

    Neidh¨ ofer, Guido, and Maximilian Stockhausen.2019. “Dynastic Inequality Compared: Multigenerational Mobility in the United States, the United Kingdom, and Germany.”Review of Income and Wealth, 65(2): 383–414. 39 Neidh¨ ofer, Guido, Mat´ ıas Ciaschi, Leonardo Gasparini, and Joaqu´ ın Serrano.2024. “Social Mobility and Economic Development.”Journal of Eco...

    2019

  9. [9]

    Drivers of Mobility in the Global South

    Elsevier. Piraino, Patrizio.2021. “Drivers of Mobility in the Global South.”Social Mobility in Developing Countries, 35–53. Poppele, Jessica, Sudarno Sumarto, and Lant Pritchett.1999. “Social Impacts of the Indonesian Crisis: New Data and Policy Implications.” Working Paper. Raza, Syed Hasan, and Ugur Aytun.2021. “How Far the Apple Falls from the Tree: In...

    2021

  10. [10]

    Whose Intergenera- tional Mobility? A New Set of Estimates for Indonesia by Gender, Geography, and Generation

    Sakri, Diding, Andrew Sumner, and Arief Anshory Yusuf.2022. “Whose Intergenera- tional Mobility? A New Set of Estimates for Indonesia by Gender, Geography, and Generation.” WIDER Working Paper Series 2022/12. Solis, Alex.2017. “Credit Access and College Enrollment.”Journal of Political Economy, 125(2): 562–622. Solon, Gary.2004. “A Model of Intergeneratio...

    2022

  11. [11]

    Multigenerational Inequality

    Stuhler, Jan.2024. “Multigenerational Inequality.”Research Handbook on Intergenerational In- equality, 100–121. Thomas, Duncan, Kathleen Beegle, Elizabeth Frankenberg, Bondan Sikoki, John Strauss, and Graciela Teruel.2004. “Education in a Crisis.”Journal of Development Eco- nomics, 74(1): 53–85. Torche, Florencia.2015. “Analyses of Intergenerational Mobil...

    2024

  12. [12]

    The Effects of Grandparents on Children’s Schooling: Evidence From Rural China

    Zeng, Zhen, and Yu Xie.2014. “The Effects of Grandparents on Children’s Schooling: Evidence From Rural China.”Demography, 51(2): 599–617. 40 A Appendix A.1 The grandparent coefficient under non-stationarity In a multivariate regression of child outcomey it on parent outcomey it−1 and grandparent outcome yit−2, the coefficient on the latter is positive if ...

    2014

  13. [13]

    parents”, “self

    4 Marital traditions.The surveys elicit information on the mechanism of marriage formation. In particular, we focus our attention on the question about who in the household selects the spouse. Possible answers include “parents”, “self” or “family”. 5 Education expenditures.Each household is asked the total school-related expenditures during the relevant s...

    2016

  14. [14]

    Approximately what was the total monthly expenditures (e.g., tuitions, pocket money)?

    However, the 1993 wave asks different questions compared to the following waves. 6 The resulting distribution of school expenditures is very different in the different waves. In the analysis in Section 5, where we compute the expenditure share at the province level, we consider the average expenditure within each province only including the expenditures f...

    1993

  15. [15]

    Figure A5 plots the estimated density of the birth years of parents, which shows that the youngest parents are born in the late 1960s. Figure A6 plots the same estimated distribution for grandparents: they are born between the late 19th century and the late 1940s and, as expected, paternal grandfathers are the oldest group on average, while maternal grand...

    1995

  16. [16]

    Each regression controls for birth cohort of the wife fixed effects, standard errors are clustered at the family level

    or the mother (columns 2 and 3). Each regression controls for birth cohort of the wife fixed effects, standard errors are clustered at the family level. *** p<0.01, ** p<0.05, * p<0.1. xvii Table A9:Crisis Impact on Multigenerational Links in Education (Table 4 cont.) (1) (2) (3) (4) (5) (6) Specification 1 Specification 2 Pooling treated cohorts Separati...

    1940

  17. [17]

    xviii Table A10:Crisis Impact on Multigenerational Links in Education: Placebo Test (1) (2) (3) Sample All Males Females GP edu 0.0328 0.0541 0.0136 (0.0729) (0.0874) (0.0988) GP edu x crisis 0.0226 0.0799 -0.0558 (0.0414) (0.0586) (0.0847) Born 77–78 x GP edu 0.0078 -0.0562 0.0665 (0.0667) (0.1050) (0.1240) Born 77–78 x GP edu x crisis -0.0532 0.0009 -0....

    2030

  18. [18]

    OLS regression of years of education on average parental education and average grandparental education, fully interacted with an indicator for whether the child was born in 1977 or 1978 (as opposed to 1975 or

    1977

  19. [19]

    and the variable “crisis”, which measures exposure to the crisis as the proportionate change in nighttime lights in the kabupaten of residence in 1993 between 1996 and

    1993

  20. [20]

    This measure is then standardised to have mean 0 and standard deviation 1 in the analysis sample. All specifications also control for kabupaten fixed effects and cluster the standard errors at the Karesidenan level, which is a geographical unit larger than kabupaten but smaller than province. ***p <0.01, **p <0.05, *p <0.1. xix Table A11:Crisis Impact on ...

    1900

  21. [21]

    This measure is then standardised to have mean 0 and standard deviation 1 in the analysis sample. All specifications also control for kabupaten fixed effects and cluster the standard errors at the Karesidenan level, which is a geographical unit larger than kabupaten but smaller than province. Cohorts are binned such that the potentially affected cohorts a...

    1997

  22. [22]

    xxii Figure A3:Education by Cohort for Grandparents in IFLS and Census Note: The top panel plots the average years of education per cohort (born between 1900 to

    in the 1995 or 2005 Census and in the IFLS panel. xxii Figure A3:Education by Cohort for Grandparents in IFLS and Census Note: The top panel plots the average years of education per cohort (born between 1900 to

    1995

  23. [23]

    The bottom panel plots the proportion of respondents with no education per cohort (born between 1900 to

    in the 1995 or 2005 Census and in the IFLS panel. The bottom panel plots the proportion of respondents with no education per cohort (born between 1900 to

    1995

  24. [24]

    in the 1995 or 2005 Census and in the IFLS panel. xxiii Figure A4:Years of Birth, Cohorts 1975-1988 0 2 4 6 8 10 Percent 1975 1980 1985 1990 Years of Birth Men Women Notes: The figure plots the distribution of birth cohorts in our grandchildren sample by gender. Figure A5:Distribution of Years of Birth, Parents 0 .02 .04 .06 1900 1920 1940 1960 1980 Years...

    1995

  25. [25]

    xxiv Figure A6:Distribution of Years of Birth, Grandparents 0 .01 .02 .03 .04 1800 1850 1900 1950 2000 Years of Birth Paternal GF Maternal GF Paternal GM Maternal GM Note: Kernel density estimation for the distribution of years of birth of grandfathers and grandmothers by lineage. Figure A7:Distribution of Years of Education by Generation in the IFLS 0 .2...

    1900

  26. [26]

    NS refers to North Sumatra, WS refers to West Sumatra, SS refers to South Sumatra, LP refers to Lampung, JKT is DKI Jakarta, WJ is West Java, CJ is Central Java, YKT is DI Yogyakarta, EJ is East Java, WNT is West Nusa Tenggara, SK refers to South Kalimantan, and SS refers to South Sulawesi. xxvii Figure A10:Validating Nighttime Lights as a Measure of Cris...

    1996

  27. [27]

    Subfigure (d) is a scatter plot of this variable against the predicted change in kabupaten-level GDP between 1996 and

    1996

  28. [28]

    We construct this Bartik-style predictor of economic shock as a weighted average of the change in GDP in 9 one-digit industries between 1996 and 1998 weighted by the employment share in each of the 8 industries in

    1996

  29. [29]

    The industry-specific GDP figures are taken from the 1996 and 1998 Indonesia’s Statistical Yearbooks (Statistik Indonesia), which reports gross domestic product by industrial origin at constant 1993 prices (Badan Pusat Statistik , BPS,B). The employment shares are constructed using microdata from the 1995 Intercensal Population Survey (SUPAS), obtained fo...

    1996