Parental Education and Children’s Old-Age Mortality: Evidence Using College Expansions

Hamid Noghanibehambari; Jason Fletcher

doi:10.1016/j.socscimed.2026.119685

. Author manuscript; available in PMC: 2026 Aug 23.

Published in final edited form as: Soc Sci Med. 2026 Aug 10;406:119685. doi: 10.1016/j.socscimed.2026.119685

Parental Education and Children’s Old-Age Mortality: Evidence Using College Expansions

Hamid Noghanibehambari ², Jason Fletcher ³

PMCID: PMC13499677 NIHMSID: NIHMS2204457 PMID: 42585715

Abstract

This paper examines the intergenerational effects of parental college education on children’s old-age mortality, focusing on health benefits flowing from parents to children. We utilize historical data of college expansions in the late 19^th and early 20^th centuries, employ Social Security Administration death records linked with the full count 1940 census, and implement a difference-in-difference approach to exploit parental cross-cohort variations in exposure combined with spatial and temporal expansions in colleges. We document that parental exposure to new college openings during adolescence is associated with significant increases in children’s later life old age longevity, with treatment-on-treated effects ranging from 4.2 to 15.4 months. Heterogeneity analyses uncover dynamic complementarities with educational policies whereby the effects of parental exposure to college expansions amplify in states with more stringent compulsory schooling and child labor laws. Furthermore, we provide evidence of improvements in parental education, income, and socioeconomic status, along with significant increases in children’s education as key mechanism channels. Monetizing the implied gains in children’s life expectancy using standard value-of-statistical-life estimates suggests aggregate benefits on the order of $6–8 billion per additional college opening.

Keywords: Mortality, Longevity, College Opening, Education

JEL Codes: I18, J18, N51, N52

1. Introduction

Understanding the factors that explain variations in longevity and mortality remains a central and, at times, controversial question (Fletcher et al., 2022; Lleras-Muney & Moreau, 2022). Recent research across various settings documents the role of early life and childhood conditions in shaping later-life health and mortality outcomes (Aizer et al., 2016; Almond et al., 2018; Almond & Currie, 2011; Barker, 1992, 1994, 1997; Hayward & Gorman, 2004; Lindeboom et al., 2010; J. Montez & Hayward, 2011; Van Den Berg et al., 2006). Within this literature, several studies focus on individuals’ educational attainments as an important determinant of health and longevity (Fletcher, 2015; Galama et al., 2018; Lleras-Muney et al., 2022). While these studies have focused on individuals’ own educational attainments, evidence on intergenerational influences remains more limited. In particular, fewer studies have examined the role of parental education, especially higher education, on their children’s long-term health outcomes. This paper contributes to this literature by investigating the intergenerational effects of parental college education on children’s old-age longevity, leveraging the historical expansion of colleges across the US in the late 19^th and early 20^th centuries.

The expansion of higher education in the United States during the late 19^th and early 20^th centuries was driven by a number of demographic, economic, and cultural factors. The Morrill Land Grant Act of 1862 played a pivotal role in establishing state colleges focused on agriculture, science, and engineering, leading to a significant increase in the number of public universities, particularly in the Midwest and West (Duemer, 2007; Sorber, 2018). Private institutions also experienced substantial growth during this period, often founded by religious groups or philanthropic individuals seeking to promote specific educational philosophies (Geiger, 2000; Turpin, 2016). The perceived economic benefits of higher education, coupled with legislative support, contributed to a steady rise in both public and private institutions (Rudolph, 2021), thereby expanding access to higher education across the nation during this period (Figure 1 and Figure 2).

Figure 1 - — Geographic Distribution of College Expansions During the Sample Period

Figure 2 – — Time-Series Evolution of College Expansions across the Final Sample Years

The historical expansion of colleges during this period provides a particularly well-suited context for examining intergenerational health effects. Focusing on children’s old-age mortality allows us to capture the cumulative consequences of early-life and childhood conditions, which life-course research suggests are critical for shaping health trajectories over many decades. By linking parental exposure to expanded higher education opportunities with children’s survival into old age, this setting offers a unique opportunity to assess whether expansions in educational access generate long-lasting health benefits that extend beyond the directly treated generation.

The contributions of this paper to the literature are threefold. First, this study advances the literature on education and health across the life course by shifting the focus from individuals’ own educational attainment to the intergenerational longevity consequences of parental exposure to higher education expansions. A large body of U.S.-based research demonstrates that higher educational attainment is associated with improved health and lower mortality (Galama et al., 2018; Hamad et al., 2018; Lleras-Muney et al., 2022; Xue et al., 2021). For example, Lleras-Muney (2005) exploits early twentieth-century compulsory schooling laws in the United States and finds significant reductions in adult mortality associated with additional education, while Cutler & Lleras-Muney (2010) show that education improves health behaviors and health knowledge that translate into lower mortality risk. Using U.S. mortality data, Montez et al. (2012) and Sasson (2016) document large and persistent educational gradients in adult longevity, and Halpern-Manners et al. (2020) and Lleras-Muney et al. (2022) show that these gradients extend well into older ages. At the same time, evidence suggests that the mortality returns to education have been historically uneven, with Montez et al. (2019) documenting muted health and longevity benefits of education among non-White adults.

Other studies examine the relationship between college education and health and mortality outcomes. For example, Hayward et al. (2015) document a steep and increasingly pronounced mortality gradient at higher levels of educational attainment in the United States, showing that adults with college and post-college education experience substantially longer life expectancy than those with less schooling, particularly in more recent cohorts. Meara et al. (2008) provide descriptive evidence of large educational gaps in U.S. mortality and life expectancy, documenting that in 2000 life expectancy at age 25 was approximately seven years higher for individuals with some college education than for those with only a high school diploma. Using draft deferments during the Vietnam War as a source of exogenous variation, Buckles et al. (2016) find that increases in college completion significantly reduce cumulative mortality among men, with income and health insurance identified as key mechanisms. Fletcher & Noghanibehambari (2023) show that historical college expansions increase the longevity of directly exposed cohorts, while Savelyev (2020) finds that among high-IQ individuals, college education is associated with significantly longer life expectancy.

Despite this extensive evidence on the individual health returns to education, far less is known about whether expansions in access to higher education generate health benefits that persist across generations. While prior research examines the effects of college expansions on infants’ health (Currie & Moretti, 2003), crime (Noghanibehambari & Tavassoli, 2022), and individuals’ own mortality (Fletcher & Noghanibehambari, 2023)¹, little is known about the extent of the effects of educational expansions on health outcomes of the next generations, specifically children’s longevity in old age. By documenting effects on later-life mortality rather than early-life or midlife outcomes, this paper highlights higher education as a source of long-run, multigenerational health externalities.

Second, the paper contributes to the literature on the social and familial determinants of health by documenting the mechanisms through which parental education shapes children’s long-term mortality outcomes. A growing literature documents that the health returns to education extend beyond the individual to other household members. For instance, studies show that spousal education improves partners’ health and reduces mortality through shared resources and behaviors (Brown et al., 2014; Fletcher, 2009, 2018; Fletcher & Marksteiner, 2017; Jaffe et al., 2006; Monden et al., 2003; Spoerri et al., 2014). Parental education has also been shown to improve children’s health and early-life environments, with lasting consequences for later-life outcomes (Currie & Moretti, 2003; Hamad et al., 2018; Huebener, 2019, 2025; Noghanibehambari & Fletcher, 2024a). More recently, evidence suggests reverse spillovers, whereby children’s educational attainment improves parents’ health and survival through informational support, caregiving, and healthcare navigation (Barr et al., 2025; Dennison & Lee, 2021; Friedman & Mare, 2014; Lee, 2018; Torssander, 2013; Yahirun et al., 2020). Together, this literature highlights education as a household-level determinant of health operating across generations.

Building on this literature, our study contributes directly to research on familial determinants of health by showing how parental education operates through family-level processes to shape children’s long-term mortality outcomes. Our proposed mechanisms suggest a multifaceted pathway in which parental education influences children’s long-term outcomes not only through intergenerational human capital transmission, but also via improvements in early-life and childhood environments, including parental socioeconomic status, fertility behavior, and children’s educational attainment. By linking educational access to family conditions during critical developmental periods, the study extends existing work on early-life environments and later-life mortality and highlights parental education as a key familial determinant of health operating across generations (Almond et al., 2018; Hayward & Gorman, 2004; Lawlor et al., 2006; Montez & Hayward, 2014; Smith et al., 2009; Van Den Berg et al., 2006).

Third, this paper contributes to research on intergenerational transmission of education and policy complementarities by leveraging historical college expansions as an exogenous shock to parental higher education in a period of rapid institutional change. Using linked census and administrative data, we show that exposure to college expansions increases both parental and children’s educational attainment, providing direct evidence of intergenerational educational spillovers. Moreover, we demonstrate that these effects are amplified in states with more stringent compulsory schooling and child labor laws, underscoring the importance of institutional context in shaping the intergenerational health returns to education. This evidence extends prior work on intergenerational transmission of education by highlighting how education policies interact over the life course to generate long-term health benefits (Black et al., 2005; Chalfin & Deza, 2018; Oreopoulos, 2006).

The rest of the paper is organized as follows. Section 2 details the data and sample selection process. Section 3 outlines the econometric approach. Section 4 presents the main findings, heterogeneity in the results, robustness checks, balancing tests, and placebo tests. Section 5 discusses potential mechanism channels. In section 6, we discuss the economic significance and implications of our findings. We conclude the paper in section 7.

2. Data and Sample Selection

The primary source of data comes from the Social Security Administration’s Numerical Identification (Numident) death records. The data is extracted from the Censoc project (Breen & Osborne, 2022; Goldstein et al., 2021). The Censoc-Numident data reports deaths that occurred between the years 1988–2005 and is linked to the full count 1940 census, extracted from Ruggles et al. (2020).² The linked data offers three benefits that are essential in our setting. First, it allows us to observe parents’ characteristics, including educational outcomes and unique historical identification numbers. As explained below, we use this identification number to match individuals across historical censuses to infer their city of birth/childhood, a crucial piece of information for college assignments. Second, the initial dataset contains millions of observations, and even after additional data merging and sample selection, we still retain a considerably large sample size. Since our source of variation comes from changes in city-level college openings across several decades of cohorts, it is essential to have a large sample size to be able to identify effects. The combination of these two aspects of the data makes it superior to alternative data sets that contain information on mortality outcomes in addressing our desired question.³ Given that the death records are drawn from the US population, the data is fairly diverse, which allows us to examine heterogeneity analysis based on sociodemographic characteristics.

We employ cross-census linking rules provided by Price et al. (2021) in order to link parents in the 1940 census to historical censuses 1850–1930 and observe the city of their birth/childhood. In so doing, we start by linking fathers’ and mothers’ HISTID variables to the 1850 census. For individuals that could be linked, we assign the city of 1850 as the city of birth/childhood. We then repeat this process for censuses 1860–1930. Using this method, we are able to identify the city of birth/childhood for roughly 84% of parents. For the remainder of the sample, we use the city of residence in 1940 as a proxy for the city of birth/childhood.⁴

The college inventory data is extracted from Goldin & Katz (2011). The data reports the total number of colleges that were actively operating at the city level in 1897, 1924, 1934, or 1980. The data also reports the opening year of these colleges. Therefore, we are able to construct a balanced panel of the total number of operational colleges in each city and year between the years 1850 and 1940. We merge this data with the Numident data based on the parental city of birth/childhood and the year they turned age 20.⁵ We implement these data linking and merging processes for both mothers and fathers separately. Throughout the paper, we refer to these samples as the Father Sample and Mother Sample. All the analyses are done and reported for both samples separately.⁶

Summary statistics of the mother and father samples are reported in the left and right panels of Table 1. The average age at death in the mother (father) sample is 884.2 (880.5) months. About 93% (6.3%) and 95% (5.2%) of observations in the mother and father samples are white (black), respectively. On average, 4.1 and 3.9 colleges are operating in the city of birth/childhood of mothers and fathers, respectively. Roughly 6.1% and 9.2% of mothers and fathers have a college education, respectively.

Table 1 -.

Summary Statistics

	Mother Sample		Father Sample
	Mean	SD	Mean	SD
*Child Characteristics:*
Death Age (Months)	884.24	94.32	880.47	93.32
Birth Year	1924.32	7.12	1924.71	6.98
Death Year	1998.01	4.87	1998.08	4.85
White	.93	.25	.94	.23
Black	.06	.24	.05	.22
Female	.45	.5	.44	.5
Hispanic	.02	.13	.02	.13
Other Race	.002	.05	.002	.05
Years of Schooling	11.09	2.38	11.15	2.35
Any College Education	.16	.37	.16	.37
*Parental Characteristics:*
City-Level College Exposure	4.07	3.57	3.91	3.48
Birth Year	1906.56	7.67	1902.99	7.98
Age in 1940	42.35	9.65	45.75	9.77
Years of Schooling	8.11	3.4	8.14	3.7
Education < High School	.63	.48	.66	.47
Education = High School	.31	.46	.25	.43
Education: College	.06	.24	.09	.29
Observations	1,954,706		1,721,971

Open in a new tab

Notes. The sample covers birth cohorts of 1890 – 1940 and parental birth cohorts of 1876 – 1926. Parental city of birth/childhood covers 414 major US cities. Education outcomes of children are conditional on age above 17.

The top and bottom panels of Figure 1 show the geographic distribution of college inventory in the US in the years 1880 and 1940, respectively. Although college expansions during this period occurred across all census regions and states, they had a higher concentration in the West and Northeast regions. Figure 2 illustrates the time-series evolution of college expansions across different census divisions over the years 1880–1940. Figure 3 shows the density distribution of age at death in the final sample for individuals whose parents were raised in cities with below-median and above-median college exposure. Visually, the distribution for individuals from families with greater college availability in the parents’ generation is shifted slightly to the right, indicating a longer average age at death compared to those from families with lower college availability.

Figure 3 - — Density Distribution of Age at Death for Children of Parents Who Were Born in Below and Above Median College Expansions Cities

3. Econometric Method

The identification strategy exploits temporal and spatial variations induced by college expansions across cities and parental cohorts based on their age at the time of a new college opening. We operationalize this difference in difference comparison using the following specification:

{D A}_{icsb} = α_{0} + α_{1} {Coll}_{c s, b + 17} + α_{2} Z_{i} + ξ_{c} + ζ_{s b} + ε_{icsb}

(1)

Where DA is the death age (in months) of individual i whose parents are born in city c in state s and year b. The variable Coll measures the total number of colleges that each parent was exposed to in their city of birth/childhood at age 20, a common college-going age threshold. In matrix Z, we include individual i’s county-of-birth and birth-year fixed effects to absorb place-specific time-invariant and secular time-varying unobservable characteristics (e.g., economic, environmental, and policy exposures) of counties that influence long-term health and longevity. This extensive set of fixed effects also controls for the child generation’s exposure to college expansions to allow the variation to come from only the parental generation’s exposure to college openings. This matrix also includes race and gender dummies as individual covariates.

The parameter ξ includes parental city fixed effects to account for place-specific time-invariant factors that might have intergenerational spillovers. The parameter ζ represents parental birth state by birth year fixed effects. These fixed effects account for all state-level policy changes and sociodemographic/socioeconomic compositional changes in state-level population that affect parental outcomes with possible spillovers for their children’s long-run longevity. Finally, ε is a disturbance term. Standard errors are clustered on the parental city to account for serial correlations in error terms.

4. Results

4.1. Main Results

The main results of the paper are reported in the left and right panels of Table 2 for the mother and father samples, respectively. Columns 1 and 3 report regression results without the matrix of the child’s controls (i.e., Z_i) and columns 2 and 4 report full specification models. The results appear to be quite robust with or without child covariates. Maternal and paternal exposure to each additional college in the city when they were 20 is associated with 0.3- and 0.4-months higher longevity among children, respectively.⁷

Table 2 -.

The Effects of Parental Exposure to College Opening and Children’s Later-Life Old-Age Longevity

	Outcome: Age-at-Death (Months)
	Mother Sample		Father Sample
	(1)	(2)	(3)	(4)
Parental College Exposure	.27^** (.11)	.28^** (.11)	.38^*** (.13)	.39^*** (.13)
Observations	1954706	1954706	1721971	1721971
R-squared	.63	.63	.62	.63
Mean DV	884.239	884.239	880.470	880.470
Parental City FE	✓	✓	✓	✓
Child/Parental Birth Year FE	✓	✓	✓	✓
Parental Birth-State-by-Birth-Year FE	✓	✓	✓	✓
Child Birth-County		✓		✓
Child Controls		✓		✓

Open in a new tab

Notes. Standard errors, clustered on parental city, are in parentheses. Controls include dummies for race, ethnicity, and gender.

^***

p<0.01,

^**

p<0.05,

p<0.1

4.2. Heterogeneity by Race/Ethnicity and Gender

There is evidence that the effects of parental education on children’s outcomes could be heterogeneous based on parental sociodemographic characteristics (Adermon et al., 2021; Card et al., 2018). Therefore, one might expect differential impacts of parental college exposure based on race/ethnicity and gender. In columns 1–5 of Table 3, we examine these heterogeneities. The results provide mixed evidence. In the mother (father) sample of panel A (B), we find a roughly 0.33-month (0.39-month) increase in longevity among white children. However, we observe negative coefficients among non-Hispanic Blacks and positive coefficients among Hispanics. Since most of these coefficients are statistically insignificant, they limit further interpretations. The only significant coefficient is for the father sample and among Hispanics, suggesting an increase in longevity of roughly 2.4 months.

Table 3 -.

Examining Heterogeneity in the Effects of Parental College Exposure and Children’s Later Life Old-Age Mortality

	Outcome: Age-at-Death (Months), Subsamples:
	Non-Hispanic Whites	Non-Hispanic Blacks	Hispanics	Males	Females	Compulsory Attendance Index > 8	Compulsory Attendance Index ≤ 8	Child Labor Index ≥ 8	Child Labor Index < 8
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
*Panel A. Mother Sample*
Parental College Exposure	.33^*** (.12)	−1.06 (.65)	.65 (.78)	.48^*** (.17)	.08 (.18)	.47^*** (.14)	.1 (.21)	.39^** (.17)	.31^* (.17)
Observations	1791914	122555	35081	1083797	870540	1135144	718312	635821	1217683
R-squared	.63	.62	.66	.61	.65	.59	.55	.5	.54
Mean DV	886.530	855.130	869.055	876.083	894.389	863.717	896.143	833.619	898.581
P-Value of equality of coefficients	Columns (1)–(2) 0.00		Columns (1)–(3) 0.00	Columns (4)–(5) 0.23		Columns (6)–(7) 0.06		Columns (8)–(9) 0.17
*Panel B. Father Sample*
Parental College Exposure	.39^*** (.13)	−.09 (.68)	2.43^** (1.14)	.42^** (.2)	.35 (.22)	.8^*** (.15)	−.16 (.21)	.63^*** (.23)	.33^** (.16)
Observations	1598576	89831	28637	966689	754874	1027717	622114	587881	1062024
R-squared	.62	.61	.65	.61	.64	.59	.55	.5	.55
Mean DV	882.617	848.554	860.865	873.386	889.534	861.800	894.445	832.623	897.095
P-Value of equality of coefficients	Columns (1)–(2) 0.00		Columns (1)–(3) 0.00	Columns (4)–(5) 0.30		Columns (6)–(7) 0.03		Columns (8)–(9) 0.16

Open in a new tab

Notes. Standard errors, clustered on parental city, are in parentheses. Regressions include parental city fixed effects, parental birth state by birth year fixed effects, child’s birth county by Birth year fixed effects, and covariates. Controls include dummies for race, ethnicity, and gender.

^***

p<0.01,

^**

p<0.05,

p<0.1

In columns 4–5, we show the results by gender. In the mother sample, we find a significantly larger effect among male children and a small and statistically insignificant effect among female children (panel A). Although we observe a similar pattern of effects based on the child’s gender in the father sample of panel B, the coefficient on female longevity is comparable to the main results but statistically insignificant.

To formally assess whether estimated effects differ across groups, we report p-values from tests of equality of coefficients across subsamples in the last row of each panel in Table 3. The equality tests strongly reject equality of coefficients between Whites and both Blacks and Hispanics (p-values < 0.01), indicating meaningful differences in estimated effects across racial groups, despite imprecision in some subsamples. In contrast, gender differences are not statistically significant in both mother and father.

Later in the paper, we examine heterogeneity in the first-stage effects of parental college exposure on both parental and children’s educational outcomes (Appendix Table H-1). The patterns closely mirror those observed here: the effects are concentrated among whites and are stronger for males, particularly in the father sample. These findings suggest that the heterogeneity in longevity responses largely reflects differences in how college exposure translated into educational gains across groups. In other words, the mortality improvements documented in Table 3 appear to operate partly through stronger intergenerational transmission of education among whites and male offspring.⁸

4.3. Dynamic Complementarity with Educational Policies

There is a recently emerging theoretical and empirical research that suggests various exposures and policies may dynamically interact to influence children’s outcomes in the short run and long-run, especially in educational policy settings with an emphasis on human and health capital (Cunha & Heckman, 2007; Dioikitopoulos, 2014; Hendren et al., 2020; Johnson & Jackson, 2019). Therefore, another source of heterogeneity may arise from the stringency of compulsory schooling and child labor laws in the child’s generation. We use the compulsory schooling index and child labor law index provided by Acemoglu & Angrist (2000). We then divide the sample into states with more/less stringent compulsory attendance and child labor index and replicate the results across mother and father samples. These results are reported in columns 6–9 of Table 3. The effect of maternal and paternal college exposure increases to 0.5 and 0.8 for those in states with a more stringent compulsory attendance index. The effects become smaller and statistically insignificant for those families in less strict states (columns 6–7). The coefficient of maternal college exposure rises to 0.39 for those in states with more strict child labor policies. It becomes smaller in size but still statistically significant for those in states with less stringent policies (columns 8–9 of panel A). In the father sample of panel B, we observe a similar pattern across states with differential child labor index. Further, the fact that the mean of age at death in the subsample of more strict child labor law is smaller suggests an even larger implied effect in this subsample, a pattern that also holds true in the mother sample.

4.4. Robustness Checks

In Table 4, we examine the robustness of the results to alternative specifications. To have a benchmark comparison, we replicate the fully parametrized regression results of Table 2 in column 1 for the mother and father samples in panels A and B, respectively. In column 2, we add a city specific linear trend in parental birth year to account for secular evolution in unobservable factors specific to each parental birth city. The results suggest larger point estimates than those of column 1. In column 3, we interact child’s birth year by child’s birth county fixed effects to account for all place specific time-variant unobservable factors that influence long-term health and longevity of children. This set of fixed effects also controls for child’s generation exposure to colleges and better isolates the effects of parental college exposure. We observe slightly larger coefficient sizes. To account for place-specific cohort-specific factors during adolescent and early adulthood years that affect long-term health outcomes, we add county of residence in 1940 by birth cohort fixed effects into our regressions. The results, reported in column 4, are quite comparable to those of column 1.

Table 4 -.

Robustness Checks

	From the Full Model of Table 2	Adding Parental City Trend in Parental Birth Year	Adding Child’s Birth County by Birth Year FE	Adding 1940 County by Child’s Birth County FE	Adding Parental City by Race and Child’s Birth County by Race FE	Adding Birth Month and Death Month FE	Adding Parental Birth City Controls	Outcome: Log Age at Death	Accelerated Failure Time Model	Heckman (1979) Two-Stage Method	Sun & Abraham (2021) Method
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
*Panel A. Mother Sample*
Parental College Exposure	.28^** (.11)	.45^** (.2)	.3^** (.12)	.28^** (.12)	.28^** (.11)	.26^** (.11)	.26^** (.12)	.0003^** (.0001)	.0003^** (.0002)	.61^*** (.17)	.26^*** (.10)
Observations	1954706	1954706	1952722	1902161	1954224	1954706	1946132	1954706	1954706	11734235	2422387
R-squared	.63	.63	.64	.64	.63	.63	.63	.628	---	---	.68
Mean DV	884.239	884.239	884.117	883.813	884.239	884.239	884.156	6.779	884.239	884.239	884.239
*Panel B. Father Sample*
Parental College Exposure	.39^*** (.13)	.53^** (.23)	.46^*** (.14)	.41^*** (.13)	.4^*** (.13)	.38^*** (.13)	.36^*** (.13)	.0005^*** (.0002)	.0005^*** (.0002)	.63^*** (.19)	.39^*** (.13)
Observations	1721971	1721971	1719929	1674487	1721460	1721971	1712690	1721971	1721971	9421421	1721971
R-squared	.63	.63	.63	.63	.63	.63	.63	.622	---	---	.62
Mean DV	880.470	880.470	880.359	880.028	880.470	880.470	880.370	6.775	880.470	880.470	880.470

Open in a new tab

^***

p<0.01,

^**

p<0.05,

p<0.1

To allow for place-specific unobservables (such as school resources and segregation) to have differential effects on longevity based on race (Aaronson & Mazumder, 2011; Carruthers & Wanamaker, 2017), we interact parental city fixed effects and child’s county of birth fixed effects with race dummies and include them in the regressions. The results, reported in column 5, point to robust coefficient sizes. Moreover, there is evidence of the influence of the season of birth and season of death on mortality outcomes (Marti-Soler et al., 2014; Vaiserman, 2021). To account for these confounders, we include in our regressions a series of fixed effects for birth month and death month. The results, reported in column 6, provide comparable estimates to those of column 1.

In column 7, we add a wide range of parental birth city controls extracted from historical censuses and interpolated for inter-decennial years. These controls include share of different race groups, share of females, share of different age groups, share of different occupation groups, literacy rate, labor force participation rate, school attendance rate, and average measures of socioeconomic status and occupational income scores. The point estimates remain quite comparable to the main results.

In column 8, we replace the outcome with the log age at death. We observe a 0.033% and 0.047% change in longevity in the mother and father samples, respectively. These estimates are quite comparable to the implied percentage changes of the coefficients in column 1 with respect to the mean of the outcome. To complement the analysis regarding robustness to functional form, we examine the effects on binary outcomes indicating survival beyond various ages. We report these results for longevity beyond ages 65 – 80 in top panels of Figure 5. To provide context for these estimates, the bottom panels of this figure present the implied percentage change for each coefficient relative to the mean of the outcome in each sample. The results show that the effects increase in magnitude as longevity extends to older ages.

Figure 5 - — Notes. Point estimates and 95% confidence intervals are depicted. Standard errors are clustered on parental city. Regressions include parental city fixed effects, parental birth state by birth year fixed effects, child’s birth county and Birth year fixed effects, and covariates. Controls include dummies for race, ethnicity, and gender.

In columns 9–11, we examine the robustness to alternative estimators. In column 9, we implement an Accelerated Failure Time hazard model with exponential functional form (i.e., logarithmic outcome).⁹ The estimated effects are almost identical to those of linear models reported in column 8. In column 10, we account for the selection of the final sample from the original population using the two-stage method developed by Heckman (1979). This method corrects for sample selection bias when the sample is not randomly chosen, and the outcome is only observed for a subset of the population. It implements the correction process in two steps. First, the model estimates the probability of being selected in the final sample from the original population as a function of observable characteristics, based on which it calculates the Inverse Mills Ratio (IMR). In the second step, the IMR is added to the longevity equation to correct for selection bias. The estimated coefficients using this method (column 10) suggest significantly larger point estimates, implying that selection might add a downward bias in our regressions and the true effects might be larger.¹⁰

Finally, to account for negative weights and counterintuitive comparisons of ordinary least square estimations, we employ the method developed by Sun & Abraham (2021). The results, reported in column 11, provide comparable point estimates to those of column 1, suggesting very limited concerns regarding the use of the ordinary least squares method.

4.5. Balancing Tests

One concern in interpreting the main results is the possibility of correlations between college expansions and other sociodemographic characteristics. For instance, college expansions might occur in areas with a higher share of white residents and better socioeconomic status, potentially leading to higher enrollment. Alternatively, families might migrate to areas with new college openings, overall expansion, and increasing urbanization. Since the ability and willingness to migrate correlate with sociodemographic characteristics, then such migrations may confound longevity equations as these sociodemographic characteristics may also influence health and longevity. To examine these sources of endogeneity, we regress predetermined observable individual characteristics (i.e., race, ethnicity, and gender) on measures of parental college exposure, conditional on fixed effects. These results are reported in Table 5. The overall picture emerging from the estimated coefficients does not support this endogeneity concern for two reasons. First, most coefficients are statistically insignificant. Further, the point estimates are quite small economically and with respect to the mean of the outcomes. For instance, the respective coefficients on white and female in the mother sample imply a change of 0.27% and 0.018%, respectively. Second, the sign of the coefficients suggests an underestimation of the results. We observe negative correlations with the share of whites (although small and insignificant) and positive correlations with the share of Blacks (although economically small). Since the longevity of Blacks is, on average, lower than whites, the overrepresentation of Blacks in the final sample adds a downward bias in the estimated coefficients, if any.

Table 5 -.

Balancing Tests: Parental College Exposure and Changes in Sociodemographic Characteristics in the Final Sample

	Outcomes:
	White	Black	Female	Hispanic	Other races
	(1)	(2)	(3)	(4)	(5)
*Panel A. Mother Sample*
Parental College Exposure	−.0027^*** (.001)	.0032^*** (.0008)	.0008 (.001)	.0001 (.0011)	−.0006 (.0003)
Observations	1954706	1954706	1954706	1954706	1954706
R-squared	.1487	.1538	.0347	.2347	.0449
Mean DV	0.934	0.064	0.445	0.018	0.002
*Panel B. Father Sample*
Parental College Exposure	−.0012 (.001)	.0022^** (.0008)	−.0002 (.0012)	.0004 (.001)	−.001^** (.0004)
Observations	1721971	1721971	1721971	1721971	1721971
R-squared	.1349	.1399	.0311	.2279	.0543
Mean DV	0.945	0.053	0.438	0.017	0.002

Open in a new tab

Notes. Standard errors, clustered on parental city, are in parentheses. Regressions include parental city fixed effects, parental birth state by birth year fixed effects, and child’s birth county by Birth year fixed effects.

^***

p<0.01,

^**

p<0.05,

p<0.1

Moreover, we examine the association between city-level socioeconomic and sociodemographic characteristics and the prevalence of colleges over the years 1880–1940. We report and discuss these results in Appendix C. We document estimated correlations that are mostly insignificant and economically indistinguishable from zero.

4.6. Placebo Tests

Another concern in the main results is that college expansions occur in areas that are already improving in many social and economic aspects with their potential effects on population health and longevity. Therefore, rises in children’s longevity pick up on these pre-existing trends among parents in expanding cities. To investigate this concern, we implement a series of placebo tests in which we examine the effects of new college openings on children’s longevity where colleges are assigned at different parental ages. If the effects operate through improvements in parental education rather than pre-existing city-specific trends, we should be able to detect sizable point estimates for parents around college-going ages and younger. In addition, we do not expect to observe detectable changes in the longevity of children when we assign colleges to older ages. In so doing, we flexibly assign colleges at different ages between 10–30 and examine the effects on children’s longevity. These results are reported in the top and bottom panels of Figure 4 for the mother and father samples, respectively. We observe a fairly similar pattern across both samples. Specifically, we observe larger impacts around ages 13–21 and insignificant coefficients for ages 22–30.

Figure 4 - — Notes. Point estimates and 95% confidence intervals are depicted. Standard errors are clustered on parental city. Regressions include parental city fixed effects, parental birth state by birth year fixed effects, child’s birth county and Birth year fixed effects, and covariates. Controls include dummies for race, ethnicity, and gender.

4.7. Spillover Effects

A further concern in interpreting the main results is the potential spillover influence of college expansions in neighboring cities/counties. Using our college inventory database, we construct a measure for total number of colleges in neighboring cities and counties. We then add this measure into our regressions to allow for the effects of own-city college exposure to compete with neighboring cities/counties’ college exposure. The results are reported in Appendix E. The estimated coefficients of own-city college exposure remain comparable to the main results while the effects of neighboring cities/counties’ colleges are indistinguishable from zero.

5. Mechanism Channels

So far, the results suggest sizable and significant associations between parental exposure to college expansions and children’s later-life longevity. In this section, we explore the mechanisms underlying this intergenerational relationship by systematically examining a range of educational, socioeconomic, demographic, and family-based pathways. Parental college exposure and higher educational attainment may influence children’s long-run health through multiple, potentially complementary channels, including improvements in early-life and childhood environments, parental socioeconomic status, fertility behavior, residential location, and investments in children’s education. These channels can shape access to nutrition, healthcare, housing quality, and neighborhood conditions during sensitive developmental periods, with cumulative and long-lasting consequences for health and longevity (Almond et al., 2018; Currie, 2009).

Existing evidence highlights that policy exposures influencing early-life family environments are particularly consequential for long-term outcomes, providing a clear rationale for why parental education, shaped by higher-education expansions, can translate into healthier household environments through greater health literacy and access to health-related information (De Buhr & Tannen, 2020; Hendren et al., 2020; Johnson & Jackson, 2019). In addition, parental education is strongly linked to children’s educational attainment, which itself plays a central role in shaping adult health and mortality risk (Black et al., 2005; Fletcher, 2015; Lleras-Muney, 2005; Oreopoulos, 2006). Beyond educational transmission, parental education may also facilitate upward socioeconomic mobility, residential sorting into more advantaged local environments, and changes in family structure and fertility decisions, all of which can influence children’s long-term health trajectories.

Based on this theoretical background, we examine several mechanism channels. We first analyze the effects of parental college exposure on parental and children’s educational outcomes, providing evidence on intergenerational human capital transmission. We then investigate broader socioeconomic and environmental pathways, including children’s later-life socioeconomic status and residential mobility, which capture dimensions of family advantage and access to healthier social and physical environments. Finally, we explore changes in spousal education and parental fertility as additional channels through which parental education may shape early-life conditions and long-run health. While fully disentangling these mechanisms is inherently challenging, this analysis provides a more comprehensive picture of the processes linking higher education expansions to intergenerational improvements in longevity.

5.1. Parental/Children’s Educational Outcomes

We use information from the 1940 census to explore mechanism channels.¹¹ These results are reported in Table 6. Each additional college in the city of residence of the mother is associated with an increase in maternal years of schooling of about 0.16 years (column 1). The respective coefficient for the father sample suggests an increase of 0.08 years additional schooling (column 7). In addition, although there are some associations between maternal college exposure and lower levels of educational attainment, these estimates remain infinitesimal with respect to the mean. Further, the fact that we do not observe these associations in the father sample are inconsistent with endogenous changes in educational outcomes at the lower tail of educational distribution (columns 2–3 and 8–9). This evidence supports our claim that the effects do not operate by picking up on pre-existing changes in educational outcomes.

Table 6 -.

Investigating Mechanism Channels: Parental College Exposure and Parental/Children’s Educational Outcomes

	Outcomes and Samples:
	Maternal Years of Schooling	Maternal Education < High School	Maternal Education = High School	Maternal Education: Any College	Child’s Years of Schooling (Mother Sample)	Child’s Education: Any College (Mother Sample)	Paternal Years of Schooling
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Parental College Exposure	.15916^*** (.03038)	−.00824^*** (.00259)	.00415^* (.00212)	.0041^*** (.00105)	.04606^*** (.01029)	.00603^*** (.00155)	.07555^** (.03731)
Observations	1954706	1954706	1954706	1954706	930169	936577	1721971
R-squared	.14456	.12113	.09014	.04224	.13069	.08103	.11614
Mean DV	8.111	0.634	0.305	0.061	10.916	0.141	8.142
	Paternal Education < High School	Paternal Education = High School	Paternal Education: Any College	Paternal log wage income	Paternal socioeconomic index	Child’s Years of Schooling (Father Sample)	Child’s Education: Any College (Father Sample)
	(8)	(9)	(10)	(11)	(12)	(13)	(14)
Parental College Exposure	−.00031 (.00361)	−.00203 (.00288)	.00234^* (.00141)	23.52841^*** (6.0362)	.36954^*** (.14104)	.03022^*** (.01012)	.00268^* (.00162)
Observations	1721971	1721971	1721971	1666688	1595014	785980	791637
R-squared	.09442	.07007	.03384	.0628	.07412	.12755	.08533
Mean DV	0.656	0.251	0.092	1242.553	34.168	10.963	0.144

Open in a new tab

^***

p<0.01,

^**

p<0.05,

p<0.1

Additionally, we observe a significant increase in maternal college education (column 4). A new college opening is associated with 41 basis points higher probability of any college education in the mother sample, off a mean of 0.061. This is equivalent to a change of 6.7% with respect to the mean. In the father sample, we observe an implied percentage change of 2.5% with respect to the mean of the outcome (column 10).

In columns 5–6, we restrict the mother sample to individuals above age 17 and examine the effects on children’s schooling outcomes. A new college opening in the mother’s city of residence is associated with 0.05 higher years of schooling among children. It is also associated with 60 basis points higher probability of any college education, off a mean of 0.14. Columns 13–14 replicate these estimates for the father sample. We find comparably smaller coefficients. For instance, a new college opening in the father’s city of residence is associated with a 27 basis points increase in the probability of the child’s college education, off a mean of 0.14. The improvements in educational outcomes through intergenerational transmission of human capital may affect children’s socioeconomic status and health outcomes later in life. Several studies docusment the association between educational attainment and later life health and mortality outcomes (Fletcher, 2015; Galama et al., 2018; Lleras-Muney, 2005; Lleras-Muney et al., 2022).

Although the effect of new college openings on paternal college education is noisily estimated, we observe statistically significant associations with paternal wage income and socioeconomic index. A new college opening in the father’s city of residence is associated with roughly $23.5 higher wage income, a change of approximately 1.9% with respect to the mean. The overall improvements in education and socioeconomic status may provide an advantageous early life environment for children through improvements in residential location, nutrition, healthcare access, and general health environment. Several strands of research document associations between early life conditions and later life mortality (Lindeboom et al., 2010; J. Montez & Hayward, 2011, 2014; Noghanibehambari & Fletcher, 2023, 2024b; Van Den Berg et al., 2006).

Appendix Table H-1 reports heterogeneity in the effects of parental college exposure on both parental and children’s educational outcomes, disaggregated by gender and race/ethnicity. The patterns broadly align with the heterogeneity observed in the reduced-form mortality estimates presented in Table 3.

Among whites, who drive most of the reduced-form mortality effects, parental college exposure substantially raises both parents’ and children’s education. For example, one additional college exposure increases mothers’ schooling by about 0.17 years and the probability of attending college by roughly 0.45 percentage points, while for fathers the corresponding gains are around 0.09 years. The children of these parents also experience meaningful educational improvements—about 0.05 years of schooling and a 0.7 percentage-point increase in college attendance. These magnitudes are consistent with the pronounced longevity effects among whites in Table 3. On the other hand, for Black families, the estimated educational effects are weaker and statistically insignificant for parents, mirroring the null mortality effects in Table 3. There is some evidence of modest positive effects on children’s years of schooling in the maternal sample, but these do not translate into significant differences in later-life mortality. This pattern likely reflects historical constraints on educational access and returns to schooling in Black communities during the period studied.

Turning to gender heterogeneity, the first-stage results indicate that while parental exposure improves education for both sons and daughters, the gains are somewhat larger for males, particularly in the paternal sample. This corresponds closely to the mortality findings in Table 3, where longevity effects are concentrated among male offspring. The consistency between the first-stage and reduced-form patterns supports a mechanism whereby parental educational gains translate into stronger intergenerational benefits for sons’ schooling and, ultimately, longevity.

5.2. Socioeconomic Conditions and Residential Mobility

To further probe non-educational pathways through which parental college exposure may shape children’s long-run health, we examine indicators of children’s later-life socioeconomic conditions and residential mobility. For the analysis in this section, we use information on children observed in the 1940 Census, restricting the sample to individuals aged at least 15 at the time of enumeration. We examine the effects of parental college exposure on a composite socioeconomic index, an indicator of cross-county migration—defined by differences between county of birth and county of residence in 1940—and an indicator of employment in a white-collar occupation. Appendix Table I-1 shows that parental exposure to college expansions is associated with improvements in a composite socioeconomic index and a higher likelihood of cross-county migration, outcomes that are commonly interpreted as markers of upward mobility and access to more advantaged local environments (Aaronson et al., 2021; Baran et al., 2024; Finkelstein et al., 2021; Hajat et al., 2011). We further find positive coefficients for having a white-collar occupation, although the coefficients are statistically insignificant and limit further interpretations. These patterns are consistent with a mechanism in which parental education facilitates improved residential sorting, labor market opportunities, and exposure to healthier social and physical environments during childhood.

5.3. Spousal Education

Another potential mechanism channel is through changes in spousal quality. Better educated individuals (as a result of higher parental education due to college expansions) may choose partners with higher educational levels under assorted matching (Barban et al., 2021; Lewis & Oppenheimer, 2000). There is a narrow strand of research that suggest spillover effects of spousal education on own health (Brown et al., 2014; Guo et al., 2020; Kravdal, 2017). We examine the effects of parental college exposure on children’s later life spousal education using the information in the 1940 census. These results are reported in appendix table F. We observe significant increases in daughters’ spouses’ years of schooling due to maternal exposure to colleges. However, we failed to find additional evidence for paternal college exposure on daughters or for parental college exposure on sons’ spousal education.

5.4. Parental Fertility

There is evidence that education may reduce fertility and family size. Smaller family sizes could result in greater resource allocation per child, such as time, nutrition, and investments in human capital (Blake, 1981; Conley & Glauber, 2006; J. M. Fletcher & Kim, 2019; Frenette, 2011). Reduced fertility also lowers the cost associated with enhancing child quality, incentivizing parental investments (Becker & Lewis, 1973). Therefore, fertility reducing the effects of education may enhance children’s health capital with long-term implications for their longevity. In Appendix G, we directly examine the effects of parental education on their fertility. We observe significant reductions in the number of children in the household across both samples. Further, we observe that these effects are more pronounced in the right tail of fertility, suggesting greater reductions in family size among larger families.

6. Discussion

This study examines whether expanding access to higher education generates health benefits that persist across generations. Leveraging historical college expansions in the United States, we show that parental exposure to higher education opportunities during adolescence is associated with meaningful improvements in children’s survival into old age. Specifically, exposure to an additional college is associated with increases in children’s longevity of approximately 0.3–0.4 months, with implied treatment-on-the-treated effects of roughly 4 to 15 months when accounting for changes in parental educational attainment. These intergenerational longevity gains are not uniform: they are concentrated among populations historically able to translate educational access into socioeconomic advantage and are larger for sons than for daughters. Moreover, the effects are amplified in institutional contexts characterized by more stringent compulsory schooling and child labor laws, highlighting the importance of policy complementarities across the life course. Beyond educational attainment itself, our findings point to broader improvements in family socioeconomic conditions and early-life environments, including gains in parental education, children’s education, and household economic status, which together provide plausible pathways linking higher education expansions to long-run health outcomes.

In this section, we discuss the economic significance and the magnitude of the estimated coefficients in light of previous research. Halpern-Manners et al. (2020) and Lleras-Muney et al. (2022) examine the education-longevity associations using twin fixed effect and sibling fixed effect strategies using similar data and settings, respectively. They document that each additional year of education is associated with 3.4–4.9 months. Combining their estimates and columns 5 and 13 of Table 6, one expects to observe reduced form effects in the range of 0.11–0.23 months. These estimates are roughly 28–82% of the observed effects in Table 2. Therefore, although a significant portion of the effects can be explained by intergenerational transmission of human capital, a similar claim could be made based on general improvements in early life conditions and alternative pathways between parental education/socioeconomic status and children’s outcomes.

Another way to understand the magnitude of the intent to treat estimates is to compare them with the findings of other early life shocks on longevity in similar settings. For instance, Noghanibehambari & Fletcher (2023b) examine the effects of in-utero and childhood exposure to the Dust Bowl of the 1930s on old-age longevity and document a reduction of roughly 1 month. Fletcher & Noghanibehambari (2024) examine the effects of pesticide exposure during in-utero and early life induced by the emergence of cyclical cicadas on later-life male longevity and document a reduction of roughly 2.2 months. Linde & Egede (2024) examine the effects of exposure to residential redlining, a discriminatory practice denying loans to minority neighborhoods, on later life mortality and document that higher early life exposure to redlining is associated with a reduction in longevity of about 5.9 months. Vu et al. (2023) examine the effects of in-utero and early-life exposure to lynching among African Americans and document a reduction in later-life old age longevity of about 3 months. Although the intent-to-treat effects of parental exposure to college expansions on children’s longevity appears smaller than these studies, we should note that these studies pick up on extreme early life shocks, such as the environmental disasters during the Dust Bowl or the extremely tragic and horrifying incidence of lynching among African Americans. The fact that parental exposure to college expansions is in a comparable range to these extreme early life shocks points to the relevance and significance of their magnitude.

We should emphasize that the overall impacts documented in Table 2 reveal intent to treat influences across the whole population regardless of a college education. The documented effects of columns 4 and 10 of Table 6 suggest a relatively small first-stage effect, implying fairly large treatment on treated effects of parental exposure to college expansions on children’s longevity. We may use the coefficients of these columns as the first stage coefficient to inflate the reduced form effects of maternal college exposure in Table 2. We estimate that maternal (paternal) college education induced by opening new colleges raises children’s longevity by roughly 4.2 (15.4) months.

Fletcher & Noghanibehambari (2023) document that college education induced by college expansions increases individuals’ age at death by roughly 1.3 years. This treatment on treated estimates is in the same range as treatment on treated calculations of the previous paragraph, especially for the effects of paternal college education on children’s longevity. Further, combining their estimates and the effects reported in columns 6 and 14 of Table 6, we can estimate that parental exposure to a new college opening may increase the longevity of children by roughly 0.29–0.66 months through increases in children’s college education. These back-of-an-envelope calculations are in the same range as those documented in the paper.

We can also use the Value of Statistical Life (VSL) estimates to monetize gains in life expectancy implied by the reduced form estimates of Table 2. Studies suggest a VSL of roughly $10 million for US residents (Kniesner & Viscusi, 2019; Viscusi, 2018). The average longevity in the mother and father samples are 884.2 and 880.5 months, respectively. Therefore, each additional month corresponds to $11.3K and $11.4K. Based on the reduced form estimates of Table 2 and these values, the monetary equivalent of maternal and paternal exposure to a new college opening at age 20 in terms of improvements in the longevity of children is roughly $3.2K and $4.5K for each individual.

Using the number of individuals in the final sample and the estimated effects of Table 2, we calculate gains in longevity of about 45.6K and 55.9K years for the mother and father samples, respectively. Using the monetary values calculated above and these gains in longevity, we estimate that each additional college opening in the mother and father generation is associated with monetary gains of roughly $6.2 and $7.7 billion.

7. Conclusion

This paper provided new insights into the long-term intergenerational effects of parental education, particularly at higher educational levels, on children’s old-age mortality. By leveraging historical college expansions in the US and Social Security Administration death records, we find that parental exposure to college openings is associated with significant increases in children’s longevity. Using the estimated coefficients on parental college education to adjust the intent-to-treat effects, we estimate that parental college education, induced by exposure to new college openings during the late adolescence years, is associated with an increase of 4.2 to 15.4 months in children’s longevity. We provide extensive comparisons of these effects in light of previous research on early life and childhood exposures, arguing for their economic significance. We implement a wide array of robustness checks, balancing tests, and placebo tests to demonstrate that the effects are not confounded by selection bias or unobservables and, if anything, may underestimate the true impact. Furthermore, we document dynamic complementarity with educational policies, showing that the effects of college expansions are larger in states with more stringent compulsory schooling and child labor laws, suggesting that the benefits of parental education are amplified in environments with stronger educational policies. This finding underscores the importance of policy context in maximizing the intergenerational benefits of higher education.

Parental exposure to college openings increased children’s later-life longevity.
Treatment-on-treated effects range from 4.2 to 15.4 months in lifespan gains.
Effects are amplified in states with stricter schooling and child labor laws.
Higher parental education improved family SES and children’s education.
Findings reveal long-run intergenerational health benefits of higher education.

Acknowledgments

The authors claim that they have no conflict of interest to report. The authors would like to acknowledge financial support from NIA grants (R01AG060109, R01AG076830). We thank Claudia Goldin for sharing the college inventory database.

Appendix A

As discussed in section 4.4, one concern in the main results is the influence of endogenous selection of the final sample from the original population. If the selection does not follow a random process, certain characteristics of selected individuals may correlate with their health and longevity. Conditional on that these characteristics also correlate with parental exposure to college openings, then the estimated effects are confounded by these endogenous selections. We can empirically test this concern by exploring the association between parental exposure to college openings and the successful merging of the final sample records with the original population. In so doing, we start with the population of cohorts in the 1940 census whose parents resided in the cities of the final sample. We then merge this population-level data with the final sample records and generate a dummy variable indicating successful merging. We then regress this successful merging dummy variable on the parental college exposure variable, conditional on fixed effects and covariates. The results are reported in Appendix Table A-1. We find negative and significant coefficients, suggesting that selection is negatively correlated with exposure to new college openings. However, the estimated coefficients are quite small in magnitude. For instance, the fully parameterized regressions of the mother and father samples imply a change of 1.2% and 0.8%, respectively (columns 2 and 4). Moreover, as the Heckman (1979) estimate of column 9 of Table 4 suggested, these associations might bias coefficients, and the true effects could be larger than the reported main results.

Appendix Table A-1 -.

Exploring Endogenous Merging Between Original Cohorts of the 1940 Census and Numident Records in the Final Sample

	Outcome: Successful Merging between Numident Records in the Final Sample and the 1940 Census
	Mother Sample		Father Sample
	(1)	(2)	(3)	(4)
Parental College Exposure	−.00362^*** (.00068)	−.0003 (.00042)	−.00516^*** (.00062)	−.00114^*** (.00041)
Observations	11734234	11733321	9421390	9420387
R-squared	.00849	.02014	.00878	.01996
Mean DV	0.106	0.106	0.106	0.106
Parental City FE	✓	✓	✓	✓
Parental Birth Year FE	✓	✓	✓	✓
Parental Birth-State-by-Birth-Year FE	✓	✓	✓	✓
Controls	✓	✓	✓	✓
Child Birth-County-By-Birth-Year FE		✓		✓

Open in a new tab

Notes. Standard errors, clustered on parental city, are in parentheses. Controls include dummies for race, ethnicity, and gender.

^***

p<0.01,

^**

p<0.05,

p<0.1

Appendix B

In this appendix, we examine to what extent the limited death window of the Numident data might affect the estimates. In so doing, we employ the Death Master Files (DMF) data extracted from the Censoc project. Similar to the Numident data, the DMF data can also be linked to the 1940 census. However, the DMF data covers death to male individuals only and contains a relatively wider death window, i.e., the years 1975–2005. We implement a similar sample selection and econometric method to investigate the impacts of parental exposure to new college openings on the longevity of their sons in the DMF data. The results are reported in Appendix Table B-1. In columns 1 and 3, we restrict the sample to death years 1988–2005 (similar to death years of the Numident data) for the mother and father samples, respectively. The coefficient of the mother sample is quite similar to the respective coefficient of column 3 of panel A of Table 3. However, we do not observe a similar comparable coefficient in the father sample. In columns 2 and 4, we report the results across death years 1975–2005 (extending to the DMF death window coverage) for the mother and father samples, respectively. The estimated coefficients increase by 100% and 150% compared with previous columns, respectively. This cross-column comparison suggests that the truncation of the Numident death window may underestimate the true effects.

Appendix Table B-1 -.

Replicating the Main Results Using Death Master Files (DMF)

	Outcome: Age-at-Death (Months)
	Mother Sample, Death Years 1988–2005	Mother Sample, Death Years 1975–2005	Father Sample, Death Years 1988–2005	Father Sample, Death Years 1975–2005
	(1)	(2)	(3)	(4)
Parental College Exposure	.45757^*** (.16183)	.69796^** (.31224)	.26492 (.21261)	.3599 (.3103)
Observations	837630	1198951	739767	1041236
R-squared	.65276	.37421	.64403	.36283
Mean DV	885.503	841.665	881.036	836.578

Open in a new tab

Notes. Standard errors, clustered on parental city, are in parentheses. Regressions include parental city fixed effects, parental birth year by birth state fixed effects, child’s birth year fixed effects, child’s birth county fixed effects, and controls. Controls include dummies for race, ethnicity, and gender.

^***

p<0.01,

^**

p<0.05,

p<0.1

Appendix C

In this appendix, we examine how cities’ characteristics change in the years before and after college expansions. In so doing, we first construct a city-by-year panel using the historical full count censuses 1880–1940. We interpolate linearly for city characteristics in inter-decennial years. We then regress socioeconomic and sociodemographic characteristics on a continuous variable of the total number of colleges in the city, conditional on city fixed effects and state-by-year fixed effects. These regressions are weighted using city-level population. The results are reported in Appendix Table C-1. Similar to Table 5, we observe quite small reductions in the share of whites and increases in the share of Blacks. We do not observe discernible changes in the share of females, school attendance rate, share of farmers, female and male labor force participation rates, share of married individuals, and urbanization rate. Although the coefficients on the share of different childhood age groups are statistically significant, their implied percentage changes are minuscule and economically meaningless. We further observe small changes in literacy rate and female’s socioeconomic index. Compared with the mean of the outcomes, these coefficients correspond to less than 1% change. The overall picture of this table does not point to significant and concerning changes in the socioeconomic and sociodemographic landscape of cities in anticipation of or following college expansions.

Appendix Table C-1 -.

College Expansions and Changes in Sociodemographic and Socioeconomic Characteristics of Cities (1880–1940)

	Outcomes:
	Share of Whites	Share of Blacks	Share of Females	School Attendance Rate	Share of Farmers
	(1)	(2)	(3)	(4)	(5)
Total Number of Colleges s	−.00248^*** (.00046)	.0023^*** (.00049)	−.00035 (.00029)	−.00047 (.00075)	.00012 (.00012)
Observation	17817	17817	17817	17817	17817
R-squared	.99648	.9966	.96442	.92919	.88651
Mean DV	0.922	0.075	0.502	0.408	0.004
	Female Labor Force Participation Rate	Male Labor Force Participation Rate	Share of Married Individuals	Literacy Rate	Share Aged 0–5
	(6)	(7)	(8)	(9)	(10)
Total Number of Colleges	−.00098 (.00137)	−.00121 (.00348)	−.00021 (.00038)	.00728^** (.00321)	−.00059^*** (.00019)
Observations	17817	17817	17817	17817	17817
R-squared	.92281	.77282	.9697	.70804	.95779
Mean DV	0.325	0.876	0.577	0.875	0.088
	Share Aged 5–10	Share Aged 11–18	Urbanization Rate	Female Socioeconomic Index	Male Socioeconomic Index
	(11)	(12)	(13)	(14)	(15)
Total Number of Colleges	−.00043^* (.00022)	−.0005^*** (.00017)	−.00087 (.0006)	.07112^** (.03599)	−.13793^** (.06019)
Observations	17817	17817	17817	17817	17817
R-squared	.95046	.96701	.8397	.98361	.95154
Mean DV	0.104	0.132	0.998	33.686	32.959

	Outcomes and Samples:
	Maternal Years of Schooling	Maternal Education: Any College	Child’s Years of Schooling (Mother Sample)	Child’s Education: Any College (Mother Sample)	Paternal Years of Schooling	Paternal Education: Any College	Child’s Years of Schooling (Father Sample)	Child’s Education: Any College (Father Sample)
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
*Panel A. Females*
Parental College Exposure	.16415^*** (.03659)	.00435^*** (.00117)	.03839^*** (.01206)	.00414^** (.00187)	.0775^* (.04107)	.00196 (.00162)	−.0003 (.0126)	−.00112 (.00196)
Observations	870540	870540	433371	436323	754874	754874	358384	360961
R-squared	.14638	.04662	.13196	.0774	.11985	.03903	.12725	.08225
Mean DV	8.096	0.061	10.947	0.127	8.139	0.093	10.999	0.129
*Panel B. Males*
Parental College Exposure	.15543^*** (.02698)	.00399^*** (.00117)	.04724^*** (.01277)	.00666^*** (.00195)	.07485^** (.03655)	.0028^* (.00156)	.05409^*** (.01236)	.00531^*** (.00186)
Observations	1083797	1083797	496294	499751	966689	966689	427018	430097
R-squared	.14767	.04538	.14272	.09418	.11812	.03579	.14106	.09886
Mean DV	8.123	0.060	10.890	0.154	8.145	0.092	10.934	0.157
*Panel C. Non-Hispanic whites*
Parental College Exposure	.17117^*** (.03166)	.0045^*** (.0011)	.04778^*** (.01051)	.00675^*** (.00164)	.08527^** (.03804)	.00238 (.00147)	.03145^*** (.01047)	.00321^* (.00168)
Observations	1791914	1791914	873956	879991	1598576	1598576	747570	752978
R-squared	.14743	.04717	.11357	.08219	.11209	.03594	.11677	.08645
Mean DV	8.222	0.063	11.006	0.145	8.272	0.096	11.025	0.147
*Panel D. Non-Hispanic Blacks*
Parental College Exposure	−.02387 (.03518)	−.00272 (.00199)	.10712^** (.05366)	.00551 (.00517)	−.02389 (.05138)	.00188 (.0024)	.07229 (.06488)	.00381 (.00665)
Observations	122555	122555	40355	40626	89831	89831	26264	26435
R-squared	.1492	.04283	.19491	.12506	.14505	.05477	.21627	.16078
Mean DV	7.294	0.040	9.572	0.087	6.639	0.047	9.848	0.098

Open in a new tab

^***

p<0.01,

^**

p<0.05,

p<0.1

Appendix I

Appendix Table I-1 -.

Investigating Mechanism Channels: Additional Outcomes

	Outcomes and Sample:
	Socioeconomic Index	Cross-County Migrant	White-Collar Occupation
	(1)	(2)	(3)
Parental College Exposure	.20322^* (.114)	.0058^*** (.00174)	.0008 (.00074)
Observations	526221	1146623	1146623
R-squared	.07659	.29529	.05743
Mean DV	34.535	0.282	0.043
	Socioeconomic Index	Cross-County Migrant	White-Collar Occupation
	(8)	(9)	(10)
Parental College Exposure	.11425 (.12189)	.00412^** (.00179)	.00032 (.00076)
Observations	430342	979017	979017
R-squared	.07649	.2961	.05859
Mean DV	34.489	0.278	0.041

Open in a new tab

^***

p<0.01,

^**

p<0.05,

p<0.1

Appendix J

Appendix Table J-1 -.

Robustness Check: Excluding Individuals for Whom Childhood or Birth Location Is Proxied Using the 1940 City/County

	Outcome: Age-at-Death (Months)
	Mother Sample		Father Sample
	(1)	(2)	(3)	(4)
Parental College Exposure	.23188^* (.13036)	.22541^* (.12802)	.40538^*** (.15315)	.41846^*** (.15009)
Observations	1632723	1632708	1424855	1424838
R-squared	.52571	.52948	.5113	.51516
Mean DV	901.832	901.833	898.873	898.873
Parental City FE	✓	✓	✓	✓
Child/Parental Birth Year FE	✓	✓	✓	✓
Parental Birth-State-by-Birth-Year FE	✓	✓	✓	✓
Child Birth-County		✓		✓
Child Controls		✓		✓

Open in a new tab

Notes. Standard errors, clustered on parental city, are in parentheses. Controls include dummies for race, ethnicity, and gender.

^***

p<0.01,

^**

p<0.05,

p<0.1

Appendix K

Appendix Table K-1 -.

Estimates of the Effects of Parental College Exposure by Parental Cohort

	Outcome: Age-at-Death (Months)
	Mother Sample		Father Sample
	(1)	(2)	(3)	(4)
	Mother Birth Year < 1910	Mother Birth Year ≥ 1910	Father Birth Year < 1910	Father Birth Year ≥ 1910
Parental College Exposure	.41459^** (.20023)	.61542^* (.3663)	.48098^** (.19407)	.21961 (.57062)
Observations	1185153	769302	1340329	381260
R-squared	.61076	.65359	.60961	.66593
Mean DV	891.844	872.522	884.458	866.451
Parental City FE	✓	✓	✓	✓
Child/Parental Birth Year FE	✓	✓	✓	✓
Parental Birth-State-by-Birth-Year FE	✓	✓	✓	✓
Child Birth-County	✓	✓	✓	✓
Child Controls	✓	✓	✓	✓

Open in a new tab

Notes. Standard errors, clustered on parental city, are in parentheses. Controls include dummies for race, ethnicity, and gender.

^***

p<0.01,

^**

p<0.05,

p<0.1

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Fletcher & Noghanibehambari (2023) examine the long-run effects of college expansions on individuals’ own longevity, using historical variation in college availability during adolescence to show that access to higher education increases the age at death of directly exposed cohorts. Their analysis treats education primarily as an individual-level investment, focusing on how increased educational attainment translates into improvements in health and longevity. While their study builds on the same institutional setting of college expansions, the research questions, time period, and contributions are fundamentally distinct from those of the current paper. First, the period of college expansions in their study is the 1940s–1950s, whereas the current study focuses on the late nineteenth and early twentieth centuries. Second, rather than examining the health returns to education for the directly treated individual, the present paper shifts the unit of analysis to the family and asks whether parental exposure to higher education generates intergenerational health benefits for their children. By linking parents’ educational opportunities during adolescence to children’s longevity, this study identifies intergenerational mortality effects that cannot be captured by individual-level exposure in their paper. Moreover, we move beyond reduced-form estimates to examine family-based mechanisms, including parental socioeconomic status, fertility behavior, residential mobility, and children’s educational attainment. As such, this paper reframes college expansions not as a determinant of individual health, as in their study, but as a familial and social determinant of population longevity. Consequently, the current paper also contributes to the literature on early-life conditions and later-life mortality (Hayward & Gorman, 2004; J. Montez & Hayward, 2014; Myrskylä, 2010; Van Den Berg et al., 2011).

One drawback of this data is the truncation death window. In Appendix B, we use alternative data that spans the years 1975 – 2005 to replicate the main results of the paper. We argue that limited and truncated death window might underestimate the true effects, a fact that has been suggested by similar studies (Lleras-Muney et al., 2022).

Other data sources, such as the National Center for Health Statistics death records or the National Longitudinal Mortality Study, cannot be linked to historical censuses and do not provide parental information. Data sources that can be linked to the 1940 census, such as the Health and Retirement Study, offer small sample sizes on the order of several thousand which inevitably results in underpowered statistical tests.

⁴

Appendix Table J-1 evaluates the sensitivity of the results to this proxy by excluding individuals for whom the city or county of residence in 1940 is used to approximate childhood or birth location. The estimates remain positive and statistically significant in both the mother and father samples, suggesting that the main findings are robust to potential measurement error in early-life location.

⁵

The historical age of college entrance does not necessarily reflect contemporary norms. In the 1940 census, those reported at least one year of college education and were attending school have an average (median) age of 19.9 (19). This number could be even larger during the late 19^th and early 20^th century. Although we merge based on age 20, Figure 4 illustrates the effects based on different age assignments.

⁶

We construct separate Mother and Father Samples for both methodological and conceptual reasons. Methodologically, maternal and paternal linkages differ in coverage and data quality in historical census records, and estimating models separately avoids imposing symmetry across parents when linkage success, measurement error, and sample composition may differ. Conceptually, prior work documents that mothers’ and fathers’ education can affect children’s health through distinct channels such as prenatal conditions, caregiving, labor market attachment, and household resources, making it important to allow the intergenerational effects of college exposure to differ by parent (Balaj et al., 2021; Chou et al., 2010; Huebener, 2019, 2020; Rasciute, 2023).

⁷

In Appendix D, we exploit expansions in coeducational colleges. We find fairly larger effects on children’s longevity due to maternal exposure to these coeducational college expansions. The point estimates and the first stage effects on parental educational outcomes are comparable to the main results.

⁸

In addition to these heterogeneity analyses, we examine whether the intergenerational effects of parental college exposure vary across cohorts. In so doing, we estimate cohort-specific effects by splitting the sample based on parental birth cohorts. Appendix Table K-1 reports estimates separately for parents born before and after 1910. The results provide suggestive evidence of cohort variation, although differences are not uniformly statistically precise. For the maternal sample, we find positive effects in both cohorts, with somewhat larger point estimates for mothers born in 1910 or later. In contrast, for fathers, the estimated effects appear stronger for earlier cohorts, while estimates for later-born fathers are smaller and less precisely estimated. Importantly, confidence intervals overlap across cohorts, indicating that differences should be interpreted cautiously.

⁹

This model has been used in a similar setting by Aizer et al. (2016).

¹⁰

In Appendix A, we explore endogenous data merging between the original 1940 records and death records in the final sample. Specifically, we examine the correlations between parental exposure to college expansions and the likelihood of being in the linked sample, conditional on fixed effects and covariates. We observe significant but economically ignorable correlations.

¹¹

One caveat of this analysis is selection due to mortality between birth and 1940. If parental college education resulted in reductions in children’s mortality, then we observe surviving children who are more likely to come from better educated parents, hence revealing higher levels of education. In this sense, these results may partly reflect mortality selection prior to 1940 (Lundborg et al., 2018).

Ethics Approval Statement

This study uses publicly available, de-identified secondary data and does not involve human subjects as defined by applicable regulations. As a result, ethics approval was not required.

References

Aaronson D, & Mazumder B. (2011). The impact of Rosenwald schools on black achievement. Journal of Political Economy, 119(5), 821–888. 10.1086/662962/SUPPL_FILE/2010282DATA.ZIP [DOI] [Google Scholar]
Aaronson D, Mazumder B, Sanders SG, & Taylor EJ. (2021). Estimating the effect of school quality on mortality in the presence of migration: Evidence from the jim crow south. Journal of Labor Economics, 39(2), 527–558. 10.1086/709783/SUPPL_FILE/17462DATA.ZIP [DOI] [Google Scholar]
Acemoglu D, & Angrist J. (2000). How Large Are Human-Capital Externalities? Evidence from Compulsory Schooling Laws. NBER Macroeconomics Annual, 15, 9–59. 10.1086/654403 [DOI] [Google Scholar]
Adermon A, Lindahl M, & Palme M. (2021). Dynastic Human Capital, Inequality, and Intergenerational Mobility. American Economic Review, 111(5), 1523–1548. 10.1257/AER.20190553 [DOI] [Google Scholar]
Aizer A, Eli S, Ferrie J, & Muney AL. (2016). The Long-Run Impact of Cash Transfers to Poor Families. American Economic Review, 106(4), 935–971. 10.1257/AER.20140529 [DOI] [PMC free article] [PubMed] [Google Scholar]
Almond D, & Currie J. (2011). Killing Me Softly: The Fetal Origins Hypothesis. Journal of Economic Perspectives, 25(3), 153–172. 10.1257/JEP.25.3.153 [DOI] [PMC free article] [PubMed] [Google Scholar]
Almond D, Currie J, & Duque V. (2018). Childhood circumstances and adult outcomes: Act II. Journal of Economic Literature, 56(4), 1360–1446. [Google Scholar]
Balaj M, York HW, Sripada K, Besnier E, Vonen HD, Aravkin A, Friedman J, Griswold M, Jensen MR, Mohammad T, Mullany EC, Solhaug S, Sorensen R, Stonkute D, Tallaksen A, Whisnant J, Zheng P, Gakidou E, & Eikemo TA. (2021). Parental education and inequalities in child mortality: a global systematic review and meta-analysis. The Lancet, 398(10300), 608–620. 10.1016/S0140-6736(21)00534-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
Baran C, Chyn E, & Stuart BA. (2024). The Great Migration and Educational Opportunity. American Economic Journal: Applied Economics, 16(3), 354–398. 10.1257/APP.20220079 [DOI] [Google Scholar]
Barban N, De Cao E, Oreffice S, & Quintana-Domeque C. (2021). The effect of education on spousal education: A genetic approach. Labour Economics, 71, 102023. 10.1016/J.LABECO.2021.102023 [DOI] [Google Scholar]
Barker DJP. (1992). Fetal growth and adult disease. British Journal of Obstetrics and Gynaecology, 99(4), 275–276. 10.1111/J.1471-0528.1992.TB13719.X [DOI] [PubMed] [Google Scholar]
Barker DJP. (1994). Mothers, babies, and disease in later life. BMJ publishing group London. [Google Scholar]
Barker DJP. (1997). Maternal nutrition, fetal nutrition, and disease in later life. Nutrition, 13(9), 807–813. 10.1016/S0899-9007(97)00193-7 [DOI] [PubMed] [Google Scholar]
Barr AB, Dennison CR, Lee KS, & Zhang Z. (2025). Adult children’s education, parental cognitive health, and the role of selection. Social Science & Medicine, 384, 118529. 10.1016/J.SOCSCIMED.2025.118529 [DOI] [PubMed] [Google Scholar]
Becker GS, & Lewis HG. (1973). On the Interaction between the Quantity and Quality of Children. Journal of Political Economy, 81(2, Part 2), S279–S288. 10.1086/260166 [DOI] [Google Scholar]
Black SE, Devereux PJ, & Salvanes KG. (2005). Why the Apple Doesn't Fall Far: Understanding Intergenerational Transmission of Human Capital. American Economic Review, 95(1), 437–449. 10.1257/0002828053828635 [DOI] [Google Scholar]
Blake J. (1981). Family size and the quality of children. Demography, 18(4), 421–442. 10.2307/2060941/METRICS [DOI] [PubMed] [Google Scholar]
Breen CF, & Osborne M. (2022). An Assessment of CenSoc Match Quality. 10.31235/OSF.IO/BJ5MD [DOI] [Google Scholar]
Brown DC, Hummer RA, & Hayward MD. (2014). The Importance of Spousal Education for the Self-Rated Health of Married Adults in the United States. Population Research and Policy Review, 33(1), 127–151. 10.1007/S11113-013-9305-6/TABLES/8 [DOI] [PMC free article] [PubMed] [Google Scholar]
Buckles K, Hagemann A, Malamud O, Morrill M, & Wozniak A. (2016). The effect of college education on mortality. Journal of Health Economics, 50, 99–114. 10.1016/J.JHEALECO.2016.08.002 [DOI] [PubMed] [Google Scholar]
Card D, Domnisoru C, Taylor L, Fahey A, Fromson-Ho A, Grogan J, Haegle I, Rose E, Saeme D, Wessel A, & John Heinz H. (2018). The Intergenerational Transmission of Human Capital: Evidence from the Golden Age of Upward Mobility. 10.3386/W25000 [DOI] [Google Scholar]
Carruthers CK, & Wanamaker MH. (2017). Returns to school resources in the Jim Crow South. Explorations in Economic History, 64, 104–110. [Google Scholar]
Chalfin A, & Deza M. (2018). The Effect of Parental Education on Children's Drug and Alcohol Use. AEA Papers and Proceedings, 108, 373–378. 10.1257/PANDP.20181105 [DOI] [Google Scholar]
Chou SY, Liu JT, Grossman M, & Joyce T. (2010). Parental education and child health: Evidence from a natural experiment in taiwan. American Economic Journal: Applied Economics, 2(1), 33–61. 10.1257/app.2.1.33 [DOI] [PMC free article] [PubMed] [Google Scholar]
Conley D, & Glauber R. (2006). Parental Educational Investment and Children’s Academic Risk. Journal of Human Resources, XLI(4), 722–737. 10.3368/JHR.XLI.4.722 [DOI] [Google Scholar]
Cunha F, & Heckman J. (2007). The Technology of Skill Formation. American Economic Review, 97(2), 31–47. 10.1257/AER.97.2.31 [DOI] [Google Scholar]
Currie J. (2009). Healthy, wealthy, and wise: Socioeconomic status, poor health in childhood, and human capital development. Journal of Economic Literature, 47(1), 87–122. 10.1257/jel.47.1.87 [DOI] [Google Scholar]
Currie J, & Moretti E. (2003). Mother’s Education and the Intergenerational Transmission of Human Capital: Evidence from College Openings. The Quarterly Journal of Economics, 118(4), 1495–1532. 10.1162/003355303322552856 [DOI] [Google Scholar]
Cutler DM, & Lleras-Muney A. (2010). Understanding differences in health behaviors by education. Journal of Health Economics, 29(1), 1–28. 10.1016/j.jhealeco.2009.10.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
De Buhr E, & Tannen A. (2020). Parental health literacy and health knowledge, behaviours and outcomes in children: A cross-sectional survey. BMC Public Health, 20(1), 1–9. 10.1186/S12889-020-08881-5/TABLES/9 [DOI] [PMC free article] [PubMed] [Google Scholar]
Dennison CR, & Lee KS. (2021). Adult Children’s Educational Attainment and Parent Health in Mid- and Later-Life. The Journals of Gerontology: Series B, 76(9), 1857–1869. 10.1093/GERONB/GBAB109 [DOI] [PMC free article] [PubMed] [Google Scholar]
Dioikitopoulos EV. (2014). Aging, growth and the allocation of public expenditures on health and education. Canadian Journal of Economics/Revue Canadienne d’économique, 47(4), 1173–1194. 10.1111/CAJE.12113 [DOI] [Google Scholar]
Duemer LS. (2007). The agricultural education origins of the Morrill Land Grant Act of 1862. American Educational History Journal, 34(1), 135–146. [Google Scholar]
Finkelstein A, Gentzkow M, & Williams H. (2021). Place-Based Drivers of Mortality: Evidence from Migration. American Economic Review, 111(8), 2697–2735. 10.1257/AER.20190825 [DOI] [PMC free article] [PubMed] [Google Scholar]
Fletcher JM. (2009). All in the family: Mental health spillover effects between working spouses. B.E. Journal of Economic Analysis and Policy, 9(1). 10.2202/1935-1682.1967/XML [DOI] [PMC free article] [PubMed] [Google Scholar]
Fletcher JM. (2015). New evidence of the effects of education on health in the US: Compulsory schooling laws revisited. Social Science & Medicine, 127, 101–107. 10.1016/J.SOCSCIMED.2014.09.052 [DOI] [PMC free article] [PubMed] [Google Scholar]
Fletcher JM. (2018). Spousal concordance in telomere length: New evidence from older adults in the US. PLOS ONE, 13(11), e0202388. 10.1371/JOURNAL.PONE.0202388 [DOI] [PMC free article] [PubMed] [Google Scholar]
Fletcher JM, & Kim J. (2019). The effect of sibship size on non-cognitive Skills: Evidence from natural experiments. Labour Economics, 56, 36–43. 10.1016/J.LABECO.2018.11.004 [DOI] [Google Scholar]
Fletcher J, & Marksteiner R. (2017). Causal Spousal Health Spillover Effects and Implications for Program Evaluation. American Economic Journal: Economic Policy, 9(4), 144–166. 10.1257/POL.20150573 [DOI] [PMC free article] [PubMed] [Google Scholar]
Fletcher J, & Noghanibehambari H. (2023). The effects of education on mortality: Evidence using college expansions. Health Economics. 10.1002/HEC.4787 [DOI] [PMC free article] [PubMed] [Google Scholar]
Fletcher J, & Noghanibehambari H. (2024). The siren song of cicadas: Early-life pesticide exposure and later-life male mortality. Journal of Environmental Economics and Management, 123, 102903. 10.1016/J.JEEM.2023.102903 [DOI] [PMC free article] [PubMed] [Google Scholar]
Fletcher J, Schwarz H, Engelman M, Johnson N, Hakes J, & Palloni A. (2022). Understanding Geographic Disparities in Mortality. Demography. [DOI] [PMC free article] [PubMed] [Google Scholar]
Frenette M. (2011). Why do larger families reduce parental investments in child quality, but not child quality per se? Review of Economics of the Household, 9(4), 523–537. 10.1007/S11150-010-9115-0/TABLES/5 [DOI] [Google Scholar]
Friedman EM, & Mare RD. (2014). The Schooling of Offspring and the Survival of Parents. Demography, 51(4), 1271–1293. 10.1007/S13524-014-0303-Z [DOI] [PubMed] [Google Scholar]
Galama T, Lleras-Muney A, & van Kippersluis H. (2018). The Effect of Education on Health and Mortality: A Review of Experimental and Quasi-Experimental Evidence. Oxford Research Encyclopedia of Economics and Finance. 10.1093/ACREFORE/9780190625979.013.7 [DOI] [Google Scholar]
Geiger RL. (2000). The American college in the nineteenth century. Vanderbilt University Press. [Google Scholar]
Goldin C, & Katz LF. (2011). Putting the “Co” in Education: Timing, Reasons, and Consequences of College Coeducation from 1835 to the Present. Journal of Human Capital, 5(4), 377–417. 10.1086/663277 [DOI] [Google Scholar]
Goldstein JR, Alexander M, Breen C, Miranda González A, Menares F, Osborne M, Snyder M, & Yildirim U. (2021). Censoc Project. In CenSoc Mortality File: Version 2.0. Berkeley: University of California. https://censoc.berkeley.edu/data/ [Google Scholar]
Guo R, Lin L, Yi J, & Zhang J. (2020). The cross-spousal effect of education on health. Journal of Development Economics, 146, 102493. 10.1016/J.JDEVECO.2020.102493 [DOI] [Google Scholar]
Hajat A, Kaufman JS, Rose KM, Siddiqi A, & Thomas JC. (2011). Long-term effects of wealth on mortality and self-rated health status. American Journal of Epidemiology, 173(2), 192–200. 10.1093/aje/kwq348 [DOI] [PMC free article] [PubMed] [Google Scholar]
Halpern-Manners A, Helgertz J, Warren JR, & Roberts E. (2020). The Effects of Education on Mortality: Evidence From Linked U.S. Census and Administrative Mortality Data. Demography, 57(4), 1513–1541. 10.1007/S13524-020-00892-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
Hamad R, Elser H, Tran DC, Rehkopf DH, & Goodman SN. (2018). How and why studies disagree about the effects of education on health: A systematic review and meta-analysis of studies of compulsory schooling laws. Social Science & Medicine, 212, 168–178. 10.1016/J.SOCSCIMED.2018.07.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
Hayward MD, & Gorman BK. (2004). The long arm of childhood: The influence of early-life social conditions on men’s mortality. Demography 2004 41:1, 41(1), 87–107. 10.1353/DEM.2004.0005 [DOI] [PubMed] [Google Scholar]
Hayward MD, Hummer RA, & Sasson I. (2015). Trends and group differences in the association between educational attainment and U.S. adult mortality: Implications for understanding education’s causal influence. Social Science & Medicine, 127, 8–18. 10.1016/J.SOCSCIMED.2014.11.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
Heckman JJ. (1979). Sample selection bias as a specification error. Econometrica, 47(1), 153–161. 10.2307/1912352 [DOI] [Google Scholar]
Hendren N, Sprung-keyser B, Dockes C, Eppsteiner H, Fernandes A, Hoyle J, Maghzian O, Musen K, Thor N, Finkel-stein A, Friedman J, Goodman-Bacon A, Grogger J, Hoynes H, Eric Humphries J, Katz L, Miller S, Soltas E, Summers L, … Wherry L. (2020). A Unified Welfare Analysis of Government Policies. The Quarterly Journal of Economics, 135(3), 1209–1318. 10.1093/QJE/QJAA006 [DOI] [Google Scholar]
Huebener M. (2019). Life expectancy and parental education. Social Science & Medicine, 232, 351–365. 10.1016/J.SOCSCIMED.2019.04.034 [DOI] [PubMed] [Google Scholar]
Huebener M. (2020). Parental education and children’s health throughout life. The Economics of Education: A Comprehensive Overview, 91–102. 10.1016/B978-0-12-815391-8.00007-0 [DOI] [Google Scholar]
Huebener M. (2025). The Effects of Education on Health. Journal of Human Resources, 60(3), 743–779. 10.3368/JHR.0219-10060R2 [DOI] [Google Scholar]
Jaffe DH, Eisenbach Z, Neumark YD, & Manor O. (2006). Effects of husbands’ and wives’ education on each other’s mortality. Social Science & Medicine, 62(8), 2014–2023. 10.1016/J.SOCSCIMED.2005.08.030 [DOI] [PubMed] [Google Scholar]
Johnson R, & Jackson K. (2019). Reducing Inequality through Dynamic Complementarity: Evidence from Head Start and Public School Spending. American Economic Journal: Economic Policy, 11(4), 310–349. 10.1257/POL.20180510 [DOI] [Google Scholar]
Kniesner TJ, & Viscusi WK. (2019). The Value of a Statistical Life. Oxford Research Encyclopedia of Economics and Finance. 10.1093/ACREFORE/9780190625979.013.138 [DOI] [Google Scholar]
Kravdal Ø. (2017). Large and Growing Social Inequality in Mortality in Norway: The Combined Importance of Marital Status and Own and Spouse’s Education. Population and Development Review, 43(4), 645–665. 10.1111/PADR.12096 [DOI] [Google Scholar]
Lawlor DA, Sterne JAC, Tynelius P, Davey Smith G, Rasmussen F, Smith GD, & Rasmussen F. (2006). Association of childhood socioeconomic position with cause-specific mortality in a prospective record linkage study of 1,839,384 individuals. American Journal of Epidemiology, 164(9), 907–915. [DOI] [PubMed] [Google Scholar]
Lee C. (2018). Adult Children’s Education and Physiological Dysregulation Among Older Parents. The Journals of Gerontology: Series B, 73(6), 1143–1154. 10.1093/GERONB/GBX044 [DOI] [PMC free article] [PubMed] [Google Scholar]
Lewis SK, & Oppenheimer VK. (2000). Educational assortative mating across marriage markets: Non-Hispanic whites in the United States. Demography, 37(1), 29–40. 10.2307/2648094/METRICS [DOI] [PubMed] [Google Scholar]
Linde S, & Egede LE. (2024). Individual-Level Exposure to Residential Redlining in 1940 and Mortality Risk. JAMA Internal Medicine. 10.1001/JAMAINTERNMED.2024.4998 [DOI] [PMC free article] [PubMed] [Google Scholar]
Lindeboom M, Portrait F, & Van Den Berg GJ. (2010). Long-run effects on longevity of a nutritional shock early in life: The Dutch Potato famine of 1846–1847. Journal of Health Economics, 29(5), 617–629. 10.1016/J.JHEALECO.2010.06.001 [DOI] [PubMed] [Google Scholar]
Lleras-Muney A. (2005). The Relationship Between Education and Adult Mortality in the United States. The Review of Economic Studies, 72(1), 189–221. 10.1111/0034-6527.00329 [DOI] [Google Scholar]
Lleras-Muney A, & Moreau F. (2022). A Unified Model of Cohort Mortality. Demography, 59(6), 2109–2134. 10.1215/00703370-10286336 [DOI] [PubMed] [Google Scholar]
Lleras-Muney A, Price J, & Yue D. (2022). The association between educational attainment and longevity using individual-level data from the 1940 census. Journal of Health Economics, 84, 102649. 10.1016/J.JHEALECO.2022.102649 [DOI] [PubMed] [Google Scholar]
Lundborg P, Nordin M, & Rooth DO. (2018). The intergenerational transmission of human capital: the role of skills and health. Journal of Population Economics, 31(4), 1035–1065. 10.1007/s00148-018-0702-3 [DOI] [Google Scholar]
Marti-Soler H, Gonseth S, Gubelmann C, Stringhini S, Bovet P, Chen PC, Wojtyniak B, Paccaud F, Tsai DH, Zdrojewski T, & Marques-Vidal P. (2014). Seasonal Variation of Overall and Cardiovascular Mortality: A Study in 19 Countries from Different Geographic Locations. PLOS ONE, 9(11), e113500. 10.1371/JOURNAL.PONE.0113500 [DOI] [PMC free article] [PubMed] [Google Scholar]
Meara ER, Richards S, & Cutler DM. (2008). The Gap Gets Bigger: Changes In Mortality And Life Expectancy, By Education, 1981–2000. Health Affairs, 27(2), 350–360. 10.1377/HLTHAFF.27.2.350 [DOI] [PMC free article] [PubMed] [Google Scholar]
Monden CWS, Van Lenthe F, De Graaf ND, & Kraaykamp G. (2003). Partner’s and own education: does who you live with matter for self-assessed health, smoking and excessive alcohol consumption? Social Science & Medicine, 57(10), 1901–1912. 10.1016/S0277-9536(03)00055-8 [DOI] [PubMed] [Google Scholar]
Montez J, & Hayward MD. (2011). Early Life Conditions and Later Life Mortality. International Handbook of Adult Mortality, 187–206. 10.1007/978-90-481-9996-9_9 [DOI] [Google Scholar]
Montez J, & Hayward MD. (2014). Cumulative Childhood Adversity, Educational Attainment, and Active Life Expectancy Among U.S. Adults. Demography, 51(2), 413–435. 10.1007/S13524-013-0261-X [DOI] [PMC free article] [PubMed] [Google Scholar]
Montez JK, Hummer RA, & Hayward MD. (2012). Educational Attainment and Adult Mortality in the United States: A Systematic Analysis of Functional Form. Demography, 49(1), 315–336. 10.1007/S13524-011-0082-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
Montez JK, Zajacova A, Hayward MD, Woolf SH, Chapman D, & Beckfield J. (2019). Educational Disparities in Adult Mortality Across U.S. States: How Do They Differ, and Have They Changed Since the Mid-1980s? Demography, 56(2), 621–644. 10.1007/S13524-018-0750-Z [DOI] [PMC free article] [PubMed] [Google Scholar]
Myrskylä M. (2010). The Relative Effects of Shocks in Early- and Later-Life Conditions on Mortality. Population and Development Review, 36(4), 803–829. 10.1111/j.1728-4457.2010.00358.x [DOI] [PubMed] [Google Scholar]
Noghanibehambari H, & Fletcher J. (2023). Long-Term Health Benefits of Occupational Licensing: Evidence from Midwifery Laws. Journal of Health Economics, 92, 102807. 10.1016/J.JHEALECO.2023.102807 [DOI] [PMC free article] [PubMed] [Google Scholar]
Noghanibehambari H, & Fletcher J. (2024a). Unequal before death: The effect of paternal education on children’s old-age mortality in the United States. Population Studies. 10.1080/00324728.2023.2284766 [DOI] [PubMed] [Google Scholar]
Noghanibehambari H, & Fletcher JM. (2024b). Dust to Feed, Dust to Grey: The Effect of In-Utero Exposure to the Dust Bowl on Old-Age Longevity. Demography. 10.1215/00703370-11140760 [DOI] [PubMed] [Google Scholar]
Noghanibehambari H, & Tavassoli N. (2022). An ounce of prevention, a pound of cure: The effects of college expansions on crime. International Review of Law and Economics, 71, 106081. 10.1016/J.IRLE.2022.106081 [DOI] [Google Scholar]
Oreopoulos P. (2006). Estimating Average and Local Average Treatment Effects of Education when Compulsory Schooling Laws Really Matter. American Economic Review, 96(1), 152–175. 10.1257/000282806776157641 [DOI] [Google Scholar]
Price J, Buckles K, Van Leeuwen J, & Riley I. (2021). Combining family history and machine learning to link historical records: The Census Tree data set. Explorations in Economic History, 80, 101391. 10.1016/J.EEH.2021.101391 [DOI] [Google Scholar]
Rasciute S. (2023). Parental education and child health: The exploration of the cross-gender intergenerational transmission mechanism. Kyklos, 76(4), 642–658. 10.1111/KYKL.12344;WGROUP:STRING:PUBLICATION [DOI] [Google Scholar]
Rudolph F. (2021). The American college and university: A history. Plunkett Lake Press. [Google Scholar]
Ruggles S, Flood S, Goeken R, Grover J, & Meyer E. (2020). IPUMS USA: Version 10.0 [dataset]. Minneapolis, MN: IPUMS. 10.18128/D010.V10.0 [DOI] [Google Scholar]
Sasson I. (2016). Trends in Life Expectancy and Lifespan Variation by Educational Attainment: United States, 1990–2010. Demography, 53(2), 269–293. 10.1007/S13524-015-0453-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
Savelyev PA. (2020). Conscientiousness, Extraversion, College Education, and Longevity of High-Ability Individuals. Journal of Human Resources, 58(1), 0918-9720R2. 10.3368/JHR.58.1.0918-9720R2 [DOI] [Google Scholar]
Smith KR, Mineau GP, Garibotti G, & Kerber R. (2009). Effects of childhood and middle-adulthood family conditions on later-life mortality: Evidence from the Utah Population Database, 1850–2002. Social Science & Medicine, 68(9), 1649–1658. 10.1016/J.SOCSCIMED.2009.02.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
Sorber NM. (2018). Land-grant colleges and popular revolt: The origins of the Morrill Act and the reform of higher education. Cornell University Press. [Google Scholar]
Spoerri A, Schmidlin K, Richter M, Egger M, & Clough-Gorr KM. (2014). Individual and spousal education, mortality and life expectancy in Switzerland: a national cohort study. J Epidemiol Community Health, 68(9), 804–810. 10.1136/JECH-2013-203714 [DOI] [PubMed] [Google Scholar]
Sun L, & Abraham S. (2021). Estimating dynamic treatment effects in event studies with heterogeneous treatment effects. Journal of Econometrics, 225(2), 175–199. 10.1016/J.JECONOM.2020.09.006 [DOI] [Google Scholar]
Torssander J. (2013). From Child to Parent? The Significance of Children’s Education for Their Parents’ Longevity. Demography, 50(2), 637–659. 10.1007/S13524-012-0155-3 [DOI] [PubMed] [Google Scholar]
Turpin AL. (2016). A new moral vision: Gender, religion, and the changing purposes of American higher education, 1837–1917. Cornell University Press. [Google Scholar]
Vaiserman A. (2021). Season-of-birth phenomenon in health and longevity: epidemiologic evidence and mechanistic considerations. Journal of Developmental Origins of Health and Disease, 12(6), 849–858. 10.1017/S2040174420001221 [DOI] [PubMed] [Google Scholar]
Van Den Berg GJ, Doblhammer-Reiter G, Christensen K, den Berg GJ, Doblhammer-Reiter G, Christensen K, van den Berg GJ, Doblhammer-Reiter G, Christensen K, den Berg GJ, Doblhammer-Reiter G, & Christensen K. (2011). Being born under adverse economic conditions leads to a higher cardiovascular mortality rate later in life: Evidence based on individuals born at different stages of the business cycle. Demography, 48(2), 507–530. 10.1007/s13524-011-0021-8 [DOI] [PubMed] [Google Scholar]
Van Den Berg GJ, Lindeboom M, Portrait F, Van Den Berg GJ, Lindeboom M, Portrait F, den Berg GJ, Lindeboom M, & Portrait F. (2006). Economic Conditions Early in Life and Individual Mortality. American Economic Review, 96(1), 290–302. 10.1257/000282806776157740 [DOI] [PubMed] [Google Scholar]
Viscusi WK. (2018). Best Estimate Selection Bias in the Value of a Statistical Life. Journal of Benefit-Cost Analysis, 9(2), 205–246. 10.1017/BCA.2017.21 [DOI] [Google Scholar]
Vu H, Noghanibehambari H, Fletcher J, & Green T. (2023). Prenatal Exposure to Racial Violence and Later Life Mortality among Males: Evidence from Lynching. SSRN Electronic Journal. 10.2139/SSRN.4613372 [DOI] [Google Scholar]
Xue X, Cheng M, & Zhang W. (2021). Does education really improve health? A meta-analysis. Journal of Economic Surveys, 35(1), 71–105. 10.1111/JOES.12399 [DOI] [Google Scholar]
Yahirun JJ, Sheehan CM, & Mossakowski KN. (2020). Depression in Later Life: The Role of Adult Children’s College Education for Older Parents’ Mental Health in the United States. The Journals of Gerontology: Series B, 75(2), 389–402. 10.1093/GERONB/GBY135 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R1] Aaronson D, & Mazumder B. (2011). The impact of Rosenwald schools on black achievement. Journal of Political Economy, 119(5), 821–888. 10.1086/662962/SUPPL_FILE/2010282DATA.ZIP [DOI] [Google Scholar]

[R2] Aaronson D, Mazumder B, Sanders SG, & Taylor EJ. (2021). Estimating the effect of school quality on mortality in the presence of migration: Evidence from the jim crow south. Journal of Labor Economics, 39(2), 527–558. 10.1086/709783/SUPPL_FILE/17462DATA.ZIP [DOI] [Google Scholar]

[R3] Acemoglu D, & Angrist J. (2000). How Large Are Human-Capital Externalities? Evidence from Compulsory Schooling Laws. NBER Macroeconomics Annual, 15, 9–59. 10.1086/654403 [DOI] [Google Scholar]

[R4] Adermon A, Lindahl M, & Palme M. (2021). Dynastic Human Capital, Inequality, and Intergenerational Mobility. American Economic Review, 111(5), 1523–1548. 10.1257/AER.20190553 [DOI] [Google Scholar]

[R5] Aizer A, Eli S, Ferrie J, & Muney AL. (2016). The Long-Run Impact of Cash Transfers to Poor Families. American Economic Review, 106(4), 935–971. 10.1257/AER.20140529 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R6] Almond D, & Currie J. (2011). Killing Me Softly: The Fetal Origins Hypothesis. Journal of Economic Perspectives, 25(3), 153–172. 10.1257/JEP.25.3.153 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R7] Almond D, Currie J, & Duque V. (2018). Childhood circumstances and adult outcomes: Act II. Journal of Economic Literature, 56(4), 1360–1446. [Google Scholar]

[R8] Balaj M, York HW, Sripada K, Besnier E, Vonen HD, Aravkin A, Friedman J, Griswold M, Jensen MR, Mohammad T, Mullany EC, Solhaug S, Sorensen R, Stonkute D, Tallaksen A, Whisnant J, Zheng P, Gakidou E, & Eikemo TA. (2021). Parental education and inequalities in child mortality: a global systematic review and meta-analysis. The Lancet, 398(10300), 608–620. 10.1016/S0140-6736(21)00534-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R9] Baran C, Chyn E, & Stuart BA. (2024). The Great Migration and Educational Opportunity. American Economic Journal: Applied Economics, 16(3), 354–398. 10.1257/APP.20220079 [DOI] [Google Scholar]

[R10] Barban N, De Cao E, Oreffice S, & Quintana-Domeque C. (2021). The effect of education on spousal education: A genetic approach. Labour Economics, 71, 102023. 10.1016/J.LABECO.2021.102023 [DOI] [Google Scholar]

[R11] Barker DJP. (1992). Fetal growth and adult disease. British Journal of Obstetrics and Gynaecology, 99(4), 275–276. 10.1111/J.1471-0528.1992.TB13719.X [DOI] [PubMed] [Google Scholar]

[R12] Barker DJP. (1994). Mothers, babies, and disease in later life. BMJ publishing group London. [Google Scholar]

[R13] Barker DJP. (1997). Maternal nutrition, fetal nutrition, and disease in later life. Nutrition, 13(9), 807–813. 10.1016/S0899-9007(97)00193-7 [DOI] [PubMed] [Google Scholar]

[R14] Barr AB, Dennison CR, Lee KS, & Zhang Z. (2025). Adult children’s education, parental cognitive health, and the role of selection. Social Science & Medicine, 384, 118529. 10.1016/J.SOCSCIMED.2025.118529 [DOI] [PubMed] [Google Scholar]

[R15] Becker GS, & Lewis HG. (1973). On the Interaction between the Quantity and Quality of Children. Journal of Political Economy, 81(2, Part 2), S279–S288. 10.1086/260166 [DOI] [Google Scholar]

[R16] Black SE, Devereux PJ, & Salvanes KG. (2005). Why the Apple Doesn't Fall Far: Understanding Intergenerational Transmission of Human Capital. American Economic Review, 95(1), 437–449. 10.1257/0002828053828635 [DOI] [Google Scholar]

[R17] Blake J. (1981). Family size and the quality of children. Demography, 18(4), 421–442. 10.2307/2060941/METRICS [DOI] [PubMed] [Google Scholar]

[R18] Breen CF, & Osborne M. (2022). An Assessment of CenSoc Match Quality. 10.31235/OSF.IO/BJ5MD [DOI] [Google Scholar]

[R19] Brown DC, Hummer RA, & Hayward MD. (2014). The Importance of Spousal Education for the Self-Rated Health of Married Adults in the United States. Population Research and Policy Review, 33(1), 127–151. 10.1007/S11113-013-9305-6/TABLES/8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R20] Buckles K, Hagemann A, Malamud O, Morrill M, & Wozniak A. (2016). The effect of college education on mortality. Journal of Health Economics, 50, 99–114. 10.1016/J.JHEALECO.2016.08.002 [DOI] [PubMed] [Google Scholar]

[R21] Card D, Domnisoru C, Taylor L, Fahey A, Fromson-Ho A, Grogan J, Haegle I, Rose E, Saeme D, Wessel A, & John Heinz H. (2018). The Intergenerational Transmission of Human Capital: Evidence from the Golden Age of Upward Mobility. 10.3386/W25000 [DOI] [Google Scholar]

[R22] Carruthers CK, & Wanamaker MH. (2017). Returns to school resources in the Jim Crow South. Explorations in Economic History, 64, 104–110. [Google Scholar]

[R23] Chalfin A, & Deza M. (2018). The Effect of Parental Education on Children's Drug and Alcohol Use. AEA Papers and Proceedings, 108, 373–378. 10.1257/PANDP.20181105 [DOI] [Google Scholar]

[R24] Chou SY, Liu JT, Grossman M, & Joyce T. (2010). Parental education and child health: Evidence from a natural experiment in taiwan. American Economic Journal: Applied Economics, 2(1), 33–61. 10.1257/app.2.1.33 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R25] Conley D, & Glauber R. (2006). Parental Educational Investment and Children’s Academic Risk. Journal of Human Resources, XLI(4), 722–737. 10.3368/JHR.XLI.4.722 [DOI] [Google Scholar]

[R26] Cunha F, & Heckman J. (2007). The Technology of Skill Formation. American Economic Review, 97(2), 31–47. 10.1257/AER.97.2.31 [DOI] [Google Scholar]

[R27] Currie J. (2009). Healthy, wealthy, and wise: Socioeconomic status, poor health in childhood, and human capital development. Journal of Economic Literature, 47(1), 87–122. 10.1257/jel.47.1.87 [DOI] [Google Scholar]

[R28] Currie J, & Moretti E. (2003). Mother’s Education and the Intergenerational Transmission of Human Capital: Evidence from College Openings. The Quarterly Journal of Economics, 118(4), 1495–1532. 10.1162/003355303322552856 [DOI] [Google Scholar]

[R29] Cutler DM, & Lleras-Muney A. (2010). Understanding differences in health behaviors by education. Journal of Health Economics, 29(1), 1–28. 10.1016/j.jhealeco.2009.10.003 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R30] De Buhr E, & Tannen A. (2020). Parental health literacy and health knowledge, behaviours and outcomes in children: A cross-sectional survey. BMC Public Health, 20(1), 1–9. 10.1186/S12889-020-08881-5/TABLES/9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R31] Dennison CR, & Lee KS. (2021). Adult Children’s Educational Attainment and Parent Health in Mid- and Later-Life. The Journals of Gerontology: Series B, 76(9), 1857–1869. 10.1093/GERONB/GBAB109 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R32] Dioikitopoulos EV. (2014). Aging, growth and the allocation of public expenditures on health and education. Canadian Journal of Economics/Revue Canadienne d’économique, 47(4), 1173–1194. 10.1111/CAJE.12113 [DOI] [Google Scholar]

[R33] Duemer LS. (2007). The agricultural education origins of the Morrill Land Grant Act of 1862. American Educational History Journal, 34(1), 135–146. [Google Scholar]

[R34] Finkelstein A, Gentzkow M, & Williams H. (2021). Place-Based Drivers of Mortality: Evidence from Migration. American Economic Review, 111(8), 2697–2735. 10.1257/AER.20190825 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R35] Fletcher JM. (2009). All in the family: Mental health spillover effects between working spouses. B.E. Journal of Economic Analysis and Policy, 9(1). 10.2202/1935-1682.1967/XML [DOI] [PMC free article] [PubMed] [Google Scholar]

[R36] Fletcher JM. (2015). New evidence of the effects of education on health in the US: Compulsory schooling laws revisited. Social Science & Medicine, 127, 101–107. 10.1016/J.SOCSCIMED.2014.09.052 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R37] Fletcher JM. (2018). Spousal concordance in telomere length: New evidence from older adults in the US. PLOS ONE, 13(11), e0202388. 10.1371/JOURNAL.PONE.0202388 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R38] Fletcher JM, & Kim J. (2019). The effect of sibship size on non-cognitive Skills: Evidence from natural experiments. Labour Economics, 56, 36–43. 10.1016/J.LABECO.2018.11.004 [DOI] [Google Scholar]

[R39] Fletcher J, & Marksteiner R. (2017). Causal Spousal Health Spillover Effects and Implications for Program Evaluation. American Economic Journal: Economic Policy, 9(4), 144–166. 10.1257/POL.20150573 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R40] Fletcher J, & Noghanibehambari H. (2023). The effects of education on mortality: Evidence using college expansions. Health Economics. 10.1002/HEC.4787 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R41] Fletcher J, & Noghanibehambari H. (2024). The siren song of cicadas: Early-life pesticide exposure and later-life male mortality. Journal of Environmental Economics and Management, 123, 102903. 10.1016/J.JEEM.2023.102903 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R42] Fletcher J, Schwarz H, Engelman M, Johnson N, Hakes J, & Palloni A. (2022). Understanding Geographic Disparities in Mortality. Demography. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R43] Frenette M. (2011). Why do larger families reduce parental investments in child quality, but not child quality per se? Review of Economics of the Household, 9(4), 523–537. 10.1007/S11150-010-9115-0/TABLES/5 [DOI] [Google Scholar]

[R44] Friedman EM, & Mare RD. (2014). The Schooling of Offspring and the Survival of Parents. Demography, 51(4), 1271–1293. 10.1007/S13524-014-0303-Z [DOI] [PubMed] [Google Scholar]

[R45] Galama T, Lleras-Muney A, & van Kippersluis H. (2018). The Effect of Education on Health and Mortality: A Review of Experimental and Quasi-Experimental Evidence. Oxford Research Encyclopedia of Economics and Finance. 10.1093/ACREFORE/9780190625979.013.7 [DOI] [Google Scholar]

[R46] Geiger RL. (2000). The American college in the nineteenth century. Vanderbilt University Press. [Google Scholar]

[R47] Goldin C, & Katz LF. (2011). Putting the “Co” in Education: Timing, Reasons, and Consequences of College Coeducation from 1835 to the Present. Journal of Human Capital, 5(4), 377–417. 10.1086/663277 [DOI] [Google Scholar]

[R48] Goldstein JR, Alexander M, Breen C, Miranda González A, Menares F, Osborne M, Snyder M, & Yildirim U. (2021). Censoc Project. In CenSoc Mortality File: Version 2.0. Berkeley: University of California. https://censoc.berkeley.edu/data/ [Google Scholar]

[R49] Guo R, Lin L, Yi J, & Zhang J. (2020). The cross-spousal effect of education on health. Journal of Development Economics, 146, 102493. 10.1016/J.JDEVECO.2020.102493 [DOI] [Google Scholar]

[R50] Hajat A, Kaufman JS, Rose KM, Siddiqi A, & Thomas JC. (2011). Long-term effects of wealth on mortality and self-rated health status. American Journal of Epidemiology, 173(2), 192–200. 10.1093/aje/kwq348 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R51] Halpern-Manners A, Helgertz J, Warren JR, & Roberts E. (2020). The Effects of Education on Mortality: Evidence From Linked U.S. Census and Administrative Mortality Data. Demography, 57(4), 1513–1541. 10.1007/S13524-020-00892-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R52] Hamad R, Elser H, Tran DC, Rehkopf DH, & Goodman SN. (2018). How and why studies disagree about the effects of education on health: A systematic review and meta-analysis of studies of compulsory schooling laws. Social Science & Medicine, 212, 168–178. 10.1016/J.SOCSCIMED.2018.07.016 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R53] Hayward MD, & Gorman BK. (2004). The long arm of childhood: The influence of early-life social conditions on men’s mortality. Demography 2004 41:1, 41(1), 87–107. 10.1353/DEM.2004.0005 [DOI] [PubMed] [Google Scholar]

[R54] Hayward MD, Hummer RA, & Sasson I. (2015). Trends and group differences in the association between educational attainment and U.S. adult mortality: Implications for understanding education’s causal influence. Social Science & Medicine, 127, 8–18. 10.1016/J.SOCSCIMED.2014.11.024 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R55] Heckman JJ. (1979). Sample selection bias as a specification error. Econometrica, 47(1), 153–161. 10.2307/1912352 [DOI] [Google Scholar]

[R56] Hendren N, Sprung-keyser B, Dockes C, Eppsteiner H, Fernandes A, Hoyle J, Maghzian O, Musen K, Thor N, Finkel-stein A, Friedman J, Goodman-Bacon A, Grogger J, Hoynes H, Eric Humphries J, Katz L, Miller S, Soltas E, Summers L, … Wherry L. (2020). A Unified Welfare Analysis of Government Policies. The Quarterly Journal of Economics, 135(3), 1209–1318. 10.1093/QJE/QJAA006 [DOI] [Google Scholar]

[R57] Huebener M. (2019). Life expectancy and parental education. Social Science & Medicine, 232, 351–365. 10.1016/J.SOCSCIMED.2019.04.034 [DOI] [PubMed] [Google Scholar]

[R58] Huebener M. (2020). Parental education and children’s health throughout life. The Economics of Education: A Comprehensive Overview, 91–102. 10.1016/B978-0-12-815391-8.00007-0 [DOI] [Google Scholar]

[R59] Huebener M. (2025). The Effects of Education on Health. Journal of Human Resources, 60(3), 743–779. 10.3368/JHR.0219-10060R2 [DOI] [Google Scholar]

[R60] Jaffe DH, Eisenbach Z, Neumark YD, & Manor O. (2006). Effects of husbands’ and wives’ education on each other’s mortality. Social Science & Medicine, 62(8), 2014–2023. 10.1016/J.SOCSCIMED.2005.08.030 [DOI] [PubMed] [Google Scholar]

[R61] Johnson R, & Jackson K. (2019). Reducing Inequality through Dynamic Complementarity: Evidence from Head Start and Public School Spending. American Economic Journal: Economic Policy, 11(4), 310–349. 10.1257/POL.20180510 [DOI] [Google Scholar]

[R62] Kniesner TJ, & Viscusi WK. (2019). The Value of a Statistical Life. Oxford Research Encyclopedia of Economics and Finance. 10.1093/ACREFORE/9780190625979.013.138 [DOI] [Google Scholar]

[R63] Kravdal Ø. (2017). Large and Growing Social Inequality in Mortality in Norway: The Combined Importance of Marital Status and Own and Spouse’s Education. Population and Development Review, 43(4), 645–665. 10.1111/PADR.12096 [DOI] [Google Scholar]

[R64] Lawlor DA, Sterne JAC, Tynelius P, Davey Smith G, Rasmussen F, Smith GD, & Rasmussen F. (2006). Association of childhood socioeconomic position with cause-specific mortality in a prospective record linkage study of 1,839,384 individuals. American Journal of Epidemiology, 164(9), 907–915. [DOI] [PubMed] [Google Scholar]

[R65] Lee C. (2018). Adult Children’s Education and Physiological Dysregulation Among Older Parents. The Journals of Gerontology: Series B, 73(6), 1143–1154. 10.1093/GERONB/GBX044 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R66] Lewis SK, & Oppenheimer VK. (2000). Educational assortative mating across marriage markets: Non-Hispanic whites in the United States. Demography, 37(1), 29–40. 10.2307/2648094/METRICS [DOI] [PubMed] [Google Scholar]

[R67] Linde S, & Egede LE. (2024). Individual-Level Exposure to Residential Redlining in 1940 and Mortality Risk. JAMA Internal Medicine. 10.1001/JAMAINTERNMED.2024.4998 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R68] Lindeboom M, Portrait F, & Van Den Berg GJ. (2010). Long-run effects on longevity of a nutritional shock early in life: The Dutch Potato famine of 1846–1847. Journal of Health Economics, 29(5), 617–629. 10.1016/J.JHEALECO.2010.06.001 [DOI] [PubMed] [Google Scholar]

[R69] Lleras-Muney A. (2005). The Relationship Between Education and Adult Mortality in the United States. The Review of Economic Studies, 72(1), 189–221. 10.1111/0034-6527.00329 [DOI] [Google Scholar]

[R70] Lleras-Muney A, & Moreau F. (2022). A Unified Model of Cohort Mortality. Demography, 59(6), 2109–2134. 10.1215/00703370-10286336 [DOI] [PubMed] [Google Scholar]

[R71] Lleras-Muney A, Price J, & Yue D. (2022). The association between educational attainment and longevity using individual-level data from the 1940 census. Journal of Health Economics, 84, 102649. 10.1016/J.JHEALECO.2022.102649 [DOI] [PubMed] [Google Scholar]

[R72] Lundborg P, Nordin M, & Rooth DO. (2018). The intergenerational transmission of human capital: the role of skills and health. Journal of Population Economics, 31(4), 1035–1065. 10.1007/s00148-018-0702-3 [DOI] [Google Scholar]

[R73] Marti-Soler H, Gonseth S, Gubelmann C, Stringhini S, Bovet P, Chen PC, Wojtyniak B, Paccaud F, Tsai DH, Zdrojewski T, & Marques-Vidal P. (2014). Seasonal Variation of Overall and Cardiovascular Mortality: A Study in 19 Countries from Different Geographic Locations. PLOS ONE, 9(11), e113500. 10.1371/JOURNAL.PONE.0113500 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R74] Meara ER, Richards S, & Cutler DM. (2008). The Gap Gets Bigger: Changes In Mortality And Life Expectancy, By Education, 1981–2000. Health Affairs, 27(2), 350–360. 10.1377/HLTHAFF.27.2.350 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R75] Monden CWS, Van Lenthe F, De Graaf ND, & Kraaykamp G. (2003). Partner’s and own education: does who you live with matter for self-assessed health, smoking and excessive alcohol consumption? Social Science & Medicine, 57(10), 1901–1912. 10.1016/S0277-9536(03)00055-8 [DOI] [PubMed] [Google Scholar]

[R76] Montez J, & Hayward MD. (2011). Early Life Conditions and Later Life Mortality. International Handbook of Adult Mortality, 187–206. 10.1007/978-90-481-9996-9_9 [DOI] [Google Scholar]

[R77] Montez J, & Hayward MD. (2014). Cumulative Childhood Adversity, Educational Attainment, and Active Life Expectancy Among U.S. Adults. Demography, 51(2), 413–435. 10.1007/S13524-013-0261-X [DOI] [PMC free article] [PubMed] [Google Scholar]

[R78] Montez JK, Hummer RA, & Hayward MD. (2012). Educational Attainment and Adult Mortality in the United States: A Systematic Analysis of Functional Form. Demography, 49(1), 315–336. 10.1007/S13524-011-0082-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R79] Montez JK, Zajacova A, Hayward MD, Woolf SH, Chapman D, & Beckfield J. (2019). Educational Disparities in Adult Mortality Across U.S. States: How Do They Differ, and Have They Changed Since the Mid-1980s? Demography, 56(2), 621–644. 10.1007/S13524-018-0750-Z [DOI] [PMC free article] [PubMed] [Google Scholar]

[R80] Myrskylä M. (2010). The Relative Effects of Shocks in Early- and Later-Life Conditions on Mortality. Population and Development Review, 36(4), 803–829. 10.1111/j.1728-4457.2010.00358.x [DOI] [PubMed] [Google Scholar]

[R81] Noghanibehambari H, & Fletcher J. (2023). Long-Term Health Benefits of Occupational Licensing: Evidence from Midwifery Laws. Journal of Health Economics, 92, 102807. 10.1016/J.JHEALECO.2023.102807 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R82] Noghanibehambari H, & Fletcher J. (2024a). Unequal before death: The effect of paternal education on children’s old-age mortality in the United States. Population Studies. 10.1080/00324728.2023.2284766 [DOI] [PubMed] [Google Scholar]

[R83] Noghanibehambari H, & Fletcher JM. (2024b). Dust to Feed, Dust to Grey: The Effect of In-Utero Exposure to the Dust Bowl on Old-Age Longevity. Demography. 10.1215/00703370-11140760 [DOI] [PubMed] [Google Scholar]

[R84] Noghanibehambari H, & Tavassoli N. (2022). An ounce of prevention, a pound of cure: The effects of college expansions on crime. International Review of Law and Economics, 71, 106081. 10.1016/J.IRLE.2022.106081 [DOI] [Google Scholar]

[R85] Oreopoulos P. (2006). Estimating Average and Local Average Treatment Effects of Education when Compulsory Schooling Laws Really Matter. American Economic Review, 96(1), 152–175. 10.1257/000282806776157641 [DOI] [Google Scholar]

[R86] Price J, Buckles K, Van Leeuwen J, & Riley I. (2021). Combining family history and machine learning to link historical records: The Census Tree data set. Explorations in Economic History, 80, 101391. 10.1016/J.EEH.2021.101391 [DOI] [Google Scholar]

[R87] Rasciute S. (2023). Parental education and child health: The exploration of the cross-gender intergenerational transmission mechanism. Kyklos, 76(4), 642–658. 10.1111/KYKL.12344;WGROUP:STRING:PUBLICATION [DOI] [Google Scholar]

[R88] Rudolph F. (2021). The American college and university: A history. Plunkett Lake Press. [Google Scholar]

[R89] Ruggles S, Flood S, Goeken R, Grover J, & Meyer E. (2020). IPUMS USA: Version 10.0 [dataset]. Minneapolis, MN: IPUMS. 10.18128/D010.V10.0 [DOI] [Google Scholar]

[R90] Sasson I. (2016). Trends in Life Expectancy and Lifespan Variation by Educational Attainment: United States, 1990–2010. Demography, 53(2), 269–293. 10.1007/S13524-015-0453-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R91] Savelyev PA. (2020). Conscientiousness, Extraversion, College Education, and Longevity of High-Ability Individuals. Journal of Human Resources, 58(1), 0918-9720R2. 10.3368/JHR.58.1.0918-9720R2 [DOI] [Google Scholar]

[R92] Smith KR, Mineau GP, Garibotti G, & Kerber R. (2009). Effects of childhood and middle-adulthood family conditions on later-life mortality: Evidence from the Utah Population Database, 1850–2002. Social Science & Medicine, 68(9), 1649–1658. 10.1016/J.SOCSCIMED.2009.02.010 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R93] Sorber NM. (2018). Land-grant colleges and popular revolt: The origins of the Morrill Act and the reform of higher education. Cornell University Press. [Google Scholar]

[R94] Spoerri A, Schmidlin K, Richter M, Egger M, & Clough-Gorr KM. (2014). Individual and spousal education, mortality and life expectancy in Switzerland: a national cohort study. J Epidemiol Community Health, 68(9), 804–810. 10.1136/JECH-2013-203714 [DOI] [PubMed] [Google Scholar]

[R95] Sun L, & Abraham S. (2021). Estimating dynamic treatment effects in event studies with heterogeneous treatment effects. Journal of Econometrics, 225(2), 175–199. 10.1016/J.JECONOM.2020.09.006 [DOI] [Google Scholar]

[R96] Torssander J. (2013). From Child to Parent? The Significance of Children’s Education for Their Parents’ Longevity. Demography, 50(2), 637–659. 10.1007/S13524-012-0155-3 [DOI] [PubMed] [Google Scholar]

[R97] Turpin AL. (2016). A new moral vision: Gender, religion, and the changing purposes of American higher education, 1837–1917. Cornell University Press. [Google Scholar]

[R98] Vaiserman A. (2021). Season-of-birth phenomenon in health and longevity: epidemiologic evidence and mechanistic considerations. Journal of Developmental Origins of Health and Disease, 12(6), 849–858. 10.1017/S2040174420001221 [DOI] [PubMed] [Google Scholar]

[R99] Van Den Berg GJ, Doblhammer-Reiter G, Christensen K, den Berg GJ, Doblhammer-Reiter G, Christensen K, van den Berg GJ, Doblhammer-Reiter G, Christensen K, den Berg GJ, Doblhammer-Reiter G, & Christensen K. (2011). Being born under adverse economic conditions leads to a higher cardiovascular mortality rate later in life: Evidence based on individuals born at different stages of the business cycle. Demography, 48(2), 507–530. 10.1007/s13524-011-0021-8 [DOI] [PubMed] [Google Scholar]

[R100] Van Den Berg GJ, Lindeboom M, Portrait F, Van Den Berg GJ, Lindeboom M, Portrait F, den Berg GJ, Lindeboom M, & Portrait F. (2006). Economic Conditions Early in Life and Individual Mortality. American Economic Review, 96(1), 290–302. 10.1257/000282806776157740 [DOI] [PubMed] [Google Scholar]

[R101] Viscusi WK. (2018). Best Estimate Selection Bias in the Value of a Statistical Life. Journal of Benefit-Cost Analysis, 9(2), 205–246. 10.1017/BCA.2017.21 [DOI] [Google Scholar]

[R102] Vu H, Noghanibehambari H, Fletcher J, & Green T. (2023). Prenatal Exposure to Racial Violence and Later Life Mortality among Males: Evidence from Lynching. SSRN Electronic Journal. 10.2139/SSRN.4613372 [DOI] [Google Scholar]

[R103] Xue X, Cheng M, & Zhang W. (2021). Does education really improve health? A meta-analysis. Journal of Economic Surveys, 35(1), 71–105. 10.1111/JOES.12399 [DOI] [Google Scholar]

[R104] Yahirun JJ, Sheehan CM, & Mossakowski KN. (2020). Depression in Later Life: The Role of Adult Children’s College Education for Older Parents’ Mental Health in the United States. The Journals of Gerontology: Series B, 75(2), 389–402. 10.1093/GERONB/GBY135 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Parental Education and Children’s Old-Age Mortality: Evidence Using College Expansions

Hamid Noghanibehambari

Jason Fletcher

Abstract

1. Introduction

Figure 1 -.

Figure 2 –

2. Data and Sample Selection

Table 1 -.

Figure 3 -.

3. Econometric Method

4. Results

4.1. Main Results

Table 2 -.

4.2. Heterogeneity by Race/Ethnicity and Gender

Table 3 -.

4.3. Dynamic Complementarity with Educational Policies

4.4. Robustness Checks

Table 4 -.

Figure 5 -. The Effects on Indicators of Longevity beyond Various Thresholds.

4.5. Balancing Tests

Table 5 -.

4.6. Placebo Tests

Figure 4 -. The Effects of Parental College Exposure at Different Ages and Children’s Later Life Old Age Longevity.

4.7. Spillover Effects

5. Mechanism Channels

5.1. Parental/Children’s Educational Outcomes

Table 6 -.

5.2. Socioeconomic Conditions and Residential Mobility

5.3. Spousal Education

5.4. Parental Fertility

6. Discussion

7. Conclusion

Acknowledgments

Appendix A

Appendix Table A-1 -.

Appendix B

Appendix Table B-1 -.

Appendix C

Appendix Table C-1 -.

Appendix D

Appendix Table D-1 -.

Appendix E

Appendix Table E-1 -.

Appendix F

Appendix Table F-1 -.

Appendix G

Appendix Table G-1 -.

Appendix H

Appendix Table H-1 -.

Appendix I

Appendix Table I-1 -.

Appendix J

Appendix Table J-1 -.

Appendix K

Appendix Table K-1 -.

Footnotes

References

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases