There are large differences in intergenerational income mobility across US states and local labour markets. The probability that a child born into a family in the bottom quintile of the national income distribution will reach the top quintile during adulthood is 14.3% in Utah, but only 7.3% in Tennessee (Chetty et al. 2014).
But we do not know much about which factors make a place successful at generating high income and intergenerational mobility. High-mobility places tend to have:
- low income and racial segregation,
- low inequality,
- high social capital, and
- better schools (as proxied by test scores; see Chetty et al. 2018).
These patterns suggest that institutions and public policies have a role in promoting mobility. This cannot, though, be interpreted as a causal relationship.
The first step to mitigating these differences and improving mobility is to nderstand the role of public policies. Recently I examined the causal role of school finance equalisation – a reduction in the differences in public school revenues and expenditures across school districts in a state – on the intergenerational mobility of students exposed to different types of funding plans while in school (Biasi 2019).
US schools were historically funded mostly from local levies such as property taxes, and so wealthier districts (with a larger tax base) were able to spend more per pupil than poorer districts. A funding formula expresses each district's revenues as a combination of state funds and local levies, and it allocates state aid to each district. In an attempt to equalise expenditure and guarantee equal opportunities to all children, over the past 40 years states have reformed their school finance schemes through changes to these funding formulas.
School finance equalisation reforms have varied across states and over time, although sharing a common objective, As a result, reforms in the same state, implemented under the same name, and with the same objective have had different effects on both the level and the distribution of school expenditure across districts.
Isolating exogenous changes in school districts’ revenues driven by school finance reforms
Using variation in the distribution of per-pupil revenues generated by 13 school finance reforms, passed in 20 states, between 1986 and 2004, I study the causal effects of equalisation on intergenerational mobility of children born between 1980 and 1986, who were exposed to these reforms while in school.
One cannot simply use post-reform expenditures and revenues as an exogenous variable to explain mobility. First, variables entering the funding formula (such as house prices and income) might change over time, and therefore affect district revenues, while also having a direct effect on mobility. Second, changes to the formula alter the relationship between the 'price' of school spending to taxpayers and the amount of public good they receive in return. This might induce households to 'vote with their feet' and move across districts, in a way that affects house prices and, in turn, district revenues. Changes in house prices after a school finance reform could therefore cause post-reform revenues to be endogenous to an extent that varies across states, depending on their funding plan.
I address this issue using a simulated-instrument approach that exploits plausibly random changes in the funding formula, idiosyncratic to each state.
- I codify the formulas in place in each state and year using information from administrative and legislative sources.
- I collect data on all the district-level variables entering each formula (these data are available for a sample of 20 states covering 62% of student enrolment).
- I then simulate post-reform revenues for each school district using the post-reform formula, keeping the district's characteristics (such as property values, enrolment, income) fixed at their pre-reform values.
This approach allows me to separate the changes in the distribution of school revenues driven by exogenous changes in the funding formula from the changes driven by endogenous household sorting. It allows for differences in the extent of this endogeneity across states, driven by the fact that different states carried out very different reforms which could have affected revenues and household sorting in heterogeneous ways (Hoxby 2001).
A driver of intergenerational mobility
I measure equalisation in school revenues as the correlation between per-capita income and per-pupil revenues across districts in each state and year. Two-stages least squares estimates of the effects of equalisation indicate that a one standard deviation reduction in this correlation leads to a 5.6 percentile increase in the expected income rank for children with parental income in the 10th percentile, a 5.2 percentile increase for children from the 25th percentile, and a 3.5 percentile increase for children from the 90th percentile. These estimates correspond to a 16.2%, 14.9%, and 9.5% increase in income, respectively.
My results also indicate that the average reform would increase mobility of children from families on the 25th percentile by 3.3 percentiles, and close approximately 10% of the gap between the lowest-mobility and the highest-mobility commuting zone.
Figure 1 Changes in intergenerational income mobility in states with successful reforms, unsuccessful reforms, and no reform, 1986–2004
Source: Biasi (2019).
Notes: The figure shows the trend in intergenerational mobility (measured as the expected income rank of children with parents on the 25th percentile and relative to 1980) across cohorts, separately for states with a successful school finance reform between (defined as producing either a negative correlation between income and revenues, or a decline of at least 50% after the reform), states with an unsuccessful reform (defined as producing either a positive correlation, or a decline smaller than 50% after the reform), and states with no reform.
Importantly, 2SLS estimates are approximately 50% larger than OLS. This highlights the importance of addressing endogeneity in post-reform revenue while accounting for heterogeneity in the effects of different school finance reforms on revenues and household responses. Grade-specific effects show that equalisation is most effective if experienced during high school, immediately before the transition to college. This hints at the importance of college for intergenerational mobility.
The effects of equalisation in school revenues might vary, depending on the degree of income inequality and segregation within each commuting zone. When cross-district income inequality is high, the same reduction in the income-revenue correlation might translate into a much larger increase in revenues in lower-income districts relative to higher-income ones. Similarly, when segregation is high, a reduction in the correlation is more likely to translate into an increase in revenues for lower-income children. 2SLS estimates confirm that equalisation has the largest effects on commuting zones with higher income inequality and higher segregation.
Through which channels does school finance equalisation affect intergenerational mobility? Specifically, equalising revenues and expenditures across districts reduces the gap in basic school inputs (such as the number of teachers) and in intermediate educational outcomes (such as the probability of attending college by age 19) between richer and poorer districts.
An engine of mobility
Reducing the differences in school revenues and expenditure between more- and less- wealthy districts in each state has a causal, positive effect on mobility. School finance equalisation can be an important way to improve intergenerational mobility, especially for children from families below the income median. This also highlights the importance of accounting for differences across states in the effects of each reform on revenues and in household responses to each reform. The direct use of funding formulas is a viable approach to obtain more reliable estimates, and this is an approach that can be used in other studies as well.
Biasi, B (2019), “School Finance Equalization Increases Intergenerational Mobility”, NBER working paper 25600.
Chetty, R, N Hendren, P Kline, and E Saez (2014), "Where Is The Land of Opportunity? The Geography of Intergenerational Mobility In The United States", Quarterly Journal of Economics 129(4): 1553–1623.
Chetty, R, and N Hendren (2018), "The Impacts of Neighborhoods on Intergenerational Mobility (I): Childhood Exposure Effects", Quarterly Journal of Economics 133(3): 1107-1162.
Hoxby, C M (2001), "All School Finance Equalizations Are Not Created Equal", Quarterly Journal of Economics 116(4): 1189–1231.