This is a response to “Money Matters After All?” by Eric Hanushek, published July 17, 2015 on the Ed Next blog, which was a response to “Boosting Educational Attainment and Adult Earnings,” by C. Kirabo Jackson, Rucker C. Johnson and Claudia Persico, published in the Fall 2015 issue of Education Next. Eric Hanushek has responded to this piece in a blog entry published on July 20, 2015 on the Ed Next blog.
We would like to thank Eric Hanushek for his comments and interest in our work. We appreciate the opportunity to offer a brief response. Hanushek provides an accurate description of our study and is correct that the methodological details matter. His critique, however, is not an objection to any of our methodological choices; he instead disputes our results. He states “while these [questions about measurement and how spending reactions to court decision is measured…] are important methodological issues, it is more useful to focus on the substance of their findings.” We take this as clear evidence that Hanushek finds our methodology sound. When the methods are sound, the results must be taken seriously. We appreciate that Hanushek has done so in this case. His single, important critique of our key results is the “time trend” argument. Following the summary of our findings below, we present the “time trend” argument and highlight its flaws. We then discuss how we overcome the problems of the previous studies on which Hanushek bases his opinions. Finally, we discuss how our results differ from previous literature because (a) existing studies suffered from biases, and (b) the spending increases analyzed in our analysis were spent on more productive inputs than the spending increases examined in other studies.
Overview of our findings:
In most states, prior to the 1970s, most resources spent on K–12 schooling were raised at the local level, through local property taxes (Howell and Miller 1997; Hoxby 1996). Because the local property tax base is typically higher in areas with higher home values, and there are persistently high levels of residential segregation by socioeconomic status, heavy reliance on local financing contributed to affluent districts’ ability to spend more per student. In response to large within-state differences in per-pupil spending across wealthy/high-income and poor districts, state supreme courts overturned school finance systems in 28 states between 1971 and 2010, and many states implemented legislative reforms that spawned important changes in public education funding. The goal of these school finance reforms (SFRs) was to increase spending levels in low-spending districts, and in many cases to reduce the differences in per-pupil school-spending levels across districts. By design, some districts experienced increases in per-pupil spending while others may have experienced decreases (Murray, Evans, and Schwab 1998; Card and Payne 2002; Hoxby 2001). Our key finding is that increased per-pupil spending, induced by court-ordered SFRs, increased high school graduation rates, educational attainment, earnings, and family incomes for children who attended school after these reforms were implemented in affected districts. We find larger effects for low-income children, such that these reforms narrowed adult socioeconomic attainment differences between those raised in low- vs. high-income families.
What we do not find:
There are two misunderstandings about our findings that critics appear to make. As such, we feel it is helpful to outline what we do not conclude from our study.
1. We do not find that merely increasing spending will improve student outcomes irrespective of how it is spent. Though Hanushek’s critique may lead readers to think otherwise, at no point in our paper do we make claims suggesting that “policy makers…only have to concern themselves with how much money was provided to schools and not with how money was used.” We are very careful to highlight that how money is spent matters. We find that increased spending that leads to reductions in class sizes, increased teacher salaries and more instructional school days in a year improved outcomes. As such, one of our key conclusions is that, while how much money one spends does clearly matter, how it is spent is very important. The final lines of our full paper read, “Money alone may not be sufficient, but our findings indicate that provision of adequate funding may be a necessary condition. Importantly, we find that how the money is spent may be important. As such, to be most effective it is likely that spending increases should be coupled with systems that help ensure spending is allocated toward the most productive uses.”
2. We do not find that increasing spending by 22.7 percent will eliminate all differences in outcomes by socioeconomic status. This is a common misunderstanding of our findings that is also made by Hanushek. We find that a 22.7 percent spending increase is large enough to eliminate the average outcome differences between the poor (those with family incomes below twice the poverty line) and the non-poor (those with family incomes above twice the poverty line). Because there are large differences by socioeconomic status among those in each income group (e.g., the wealthy tend to have better outcomes than the average non-poor person, and the very poor tend to have worse outcomes than those just above the poverty line) eliminating the average difference in outcomes across the two broad groups does not eliminate all differences by socioeconomic status within each group. Simply put, just because a 22.7 percent spending increase is large enough to eliminate the average outcome differences between the poor and non-poor it does not mean that a 22.7 percent spending increase is large enough to eliminate the difference in outcome between the very poor and the very wealthy or differences across other measures of socioeconomic status. Also, we do not speculate that this spending increase will eliminate differences in outcomes by other categories such as race and gender. To illustrate this logic, consider the following simple mathematical example.
Illustrative example: There are 4 people in a society of different income levels. Persons are ranked by income level so that Person 1 is the richest and person 4 is the poorest. Richer individuals tend to have better outcomes such that Person 1 has 20 years of education, Person 2 has 18 years, Person 3 has 18 years and person 4 has 16 years of education. The average educational attainment for the two richest persons is 19 years and the average educational attainment for the two poorest persons is 17. The average gap between the high income group and the low income groups is 2 years. However, the gap between the richest and poorest person is 4 years. If one could increase the level of education for both lower income persons (persons 3 and 4) by 2 years, the average gap across the two groups would be eliminated. However, the richest person would still have 2 more years of education that the poorest person. This simple example illustrates that eliminating the average difference across the two groups will only remove all differences by socioeconomic status if there are no differences in outcomes by socioeconomic status within the broad income groups. Given that there are large difference in outcomes by socioeconomic status within broad income groups in the United States, this condition clearly does not hold in reality.
The Problem with Hanushek’s “Time Trend” Critique:
Now that the reader should have a clear sense of our paper and its implications, we now describe the Hanushek “time trend” argument. Hanushek points out that school spending in the United States has increased substantially between 1970 and present day. As such, he argues that, if our results are correct and school spending really does improve student outcomes (with larger effects for low-income children), outcomes should have improved over time and achievement gaps by income should have been eliminated over this time period. He then argues that any improvements between 1970 and today have been small so that it is unlikely that our conclusion that school spending improves student outcomes is correct.
While this “time trend” argument is intuitive, it is flawed for two reasons. The first reason is that it relies on the same flawed understanding of our results outlined above (i.e., that eliminating differences across two broad income groups implies eliminating all differences by socioeconomic status). The second problem with this “time trend” argument is that it is a facile argument based on fuzzy (albeit intuitive) logic. We highlight the problems of his logic below.
To see the problems of Hanushek’s logic, consider the following true statistics: between 1960 and 2000 the rate of cigarette smoking for females decreased by more than 30 percent while the rate of deaths by lung cancer increased by more than 50 percent over the same time period. An analysis of these time trends might lead one to infer that smoking reduces lung cancer. However, most informed readers can point out numerous flaws in looking at this time trend evidence and concluding that “if smoking causes lung cancer, then there should have been a large corresponding reduction in cancer rates so that there can be no link between smoking and lung cancer.” However, this is exactly the facile logic invoked by Hanushek regarding the effect of school spending on student achievement.
While there are several problems with this simplistic argument, to avoid going too deeply into the weeds we focus on the most important flaw in this “time trend” argument. Simply put, the “time series” argument will hold only if nothing else has changed between 1970 and present day. It is important to bear in mind that these spending increases occurred against the backdrop of countervailing influences, such as the rise in single-parent families, more highly concentrated poverty, deterioration of neighborhood conditions for low-income families, the exodus of the middle class to the suburbs, mass incarceration, the crack epidemic, changes in migration patterns, and others. Consider just one countervailing factor: the significant rise in segregation by income between neighborhoods over the past four decades. This increased residential segregation was driven mostly by families with school-age children (Owens 2015), a simple reflection that quality of local schooling options is a key driver of segregation. This significant increase in residential sorting by income among families with school-age children would have likely led to far greater disparities in school resources by community socioeconomic status had SFRs not been an effective leveling tool.
In short, 1970 and 2010 is not an “apples-to-apples” comparison, so there is no reason to expect that the correlation between aggregate spending and aggregate outcomes over such a long time span will yield anything resembling a “causal” relationship. In fact, the observation that using simple correlations over time is unlikely to yield the true “causal” relationship is exactly what motivated us to follow a different methodological approach. Our methodological approach allows for an “apples-to-apples” comparison and allows us to disentangle the effects of school spending from that of all these other countervailing forces. Though Hanushek has chosen not to discuss the methodological advances in our work, they are important, and methods matter.
How We Overcome These Problems to Facilitate “Apples-to-Apples” Comparisons:
We make several decisions in order to facilitate more of an apples-to-apples comparison. First, we use fine-grained data on individual students, rather than comparing the entire United States in 1970 to the entire United States in 2010. With these finer-grained data we are able to account for a variety of other factors that may have changed over time such as family structure, childhood poverty, and neighborhood factors. Using these finer grained data, our main approach is to compare the outcomes of individuals with similar background characteristics born in the same school district but who attended public schools during different years (when per-pupil spending levels may have been different) — i.e., an apples-to-apples comparison. However, this is not all that we do to ensure that our results yield real causal relationships.
In our paper, we point out that even if one can carefully account for several observable factors (as we do), correlating all actual changes in school spending with changes in student outcomes is unlikely to yield causal relationships. We point out that some spending changes are unrelated to other factors that may obscure the real effect on outcomes (i.e., clean spending changes), while other kinds of spending changes would clearly yield erroneous results (i.e., confounded spending changes). We point out that many of the spending changes analyzed in previous studies may have been of the confounded variety. To give an example of such confounded spending changes, consider the following example. The federal Elementary and Secondary Education Act allocates additional funding to school districts with a high percentage of low-income students, who are more likely to have poor educational outcomes for reasons unrelated to school spending. As such, school districts serving declining neighborhoods are also those that are most likely to receive additional per-pupil spending over time. Such compensatory policies generate a negative relationship between changes in school spending and student outcomes that obscure the true relationship between school spending and student outcomes. We avoid this kind of problem by focusing only on clean spending changes. Specifically, we focus on the relationship between external “shocks” to school spending and long-run adult outcomes. The “shocks” we use are the sudden unanticipated increases in school spending experienced by predominantly low-spending districts soon after passage of court-mandated SFR.
As discussed above, by design, very soon after a court-ordered SFR in a state, some districts experienced sudden unanticipated increases in per-pupil spending (i.e., shocks) while others may have experienced decreases. Our analytic approach compares the outcomes of individuals who attended school before these spending shocks to those of similar individuals from the school district after these spending shocks. The validity of our design relies on the idea that districts that experienced sudden increases in school spending right after the passage of a court-ordered SFR were not already improving in other ways in exactly those same years. For this reason, we spend much time in our work showing that the timing of these spending shocks has nothing to do with underlying neighborhood changes or changes in family characteristics, so that changes in outcomes due to these shocks are likely to reflect a causal relationship. We encourage interested readers to consult the full paper for further detail.
Reconciling our results with the Older Literature:
Even though we outline the faulty assumptions in Hanushek’s “time trend” argument, in the interest of good social science it is helpful for us to try to reconcile our findings with the simple time-series evidence. As we explain above, our results do not imply that a 22.7 percent increase will eliminate all differences by parental socioeconomic status. However, they do suggest the much more realistic prediction that one might observe some convergence across groups over time as school spending has increased. Indeed this has been the case. For example, Krueger (1998) uses data from the NAEP and documents test score increases over time, with large improvements for disadvantaged children from poor urban areas; the Current Population Survey shows declining dropout rates since 1975 for those from the lowest income quartile (Digest of Education Statistics, NCES 2012). Murnane (2013) finds that high school completion rates have been increasing since 1970 with larger increases for black and Hispanic students; Baum, Ma and Pavea (2013) find that postsecondary enrollment rates have been increasing since the 1980s, particularly for those from poor families. Contrary to Hanushek’s assertions, outcomes have improved. Importantly, these improvements are consistent with increase in school spending playing a key role.
Finally, Hanushek proposes three reasons why our estimates (if true) may not track the national time trends very well. His ideas are not novel — we considered, tested, and addressed them ourselves in the paper and herein. First, he says there may be diminishing marginal returns to schools spending. Indeed we find that this is the case in our study. Areas with the lowest initial spending levels were also those for which increased spending had the most pronounced positive effect. The second reason he cites is that spending induced by the courts might have large effects while spending not related to judicial rulings have small effects. Indeed we find evidence of this also. Specifically, spending increases associated with court-mandated reform are much more strongly related to improvement in measured school inputs (e.g., student-to-teacher ratios, length of the school year) than ordinary spending increases. There are a few explanations for this that we explore in our study. Finally, he proposes that our estimates are wrong. We propose an alternative: the time series evidence Hanushek relies on does not reflect a causal relationship. Indeed in our larger study, we show that simple correlations are obscured by a variety of other factors that also influence student outcomes. We also present numerous pieces of analysis in our larger study that support a causal interpretation of our results.
To be clear, we do not think that our study is the final word on the question of whether increasing school spending will improve student outcomes in all contexts. As Hanushek himself concedes “none of this discussion suggests that money never matters. Or that money cannot matter.” Here we will make a similar concession; none of what we show suggests that money always matters. We show that money did matter and that it mattered quite a lot. What our study does is dispels the notion that school spending does not matter, so that one must look only at how it is spent. We find that money does matter and how it is spent matters. Contrary to Hanushek’s claims, our findings do not let policymakers off the hook. Our findings suggest that it is extremely important that money is allocated effectively and also that it is allocated equitably so that all schools have the resources necessary to help all children succeed.
— Rucker C. Johnson, C. Kirabo Jackson and Claudia Persico
Rucker C. Johnson is associate professor of public policy at University of California, Berkeley. C. Kirabo Jackson is associate professor of human development and social policy at Northwestern University. Claudia Persico is a doctoral candidate in human development and social policy at Northwestern University. This article was originally posted on EducationNext.
NOTE: The lung cancer rates for males has been on the decline since 2000 and has been relatively stable for females between 2000 and 2009.
Per-pupil spending can vary drastically between school districts, with affluent suburban districts often outspending their neighbors by significant margins. Such disparate school spending is frequently identified as a primary culprit in our nation’s wide achievement gaps between students of different socioeconomic and racial backgrounds. The argument makes intrinsic sense to many: if one school district spends significantly more educating its students, then of course those students will perform better academically. Existing research on the topic, however, paints a muddier picture.
In 1966, James Coleman conducted one of the largest education studies in history to analyze aspects of educational equality in the United States, including the relationship between school spending and student outcomes. Coleman found that variation in school resources (as measured by per-pupil spending and student-to-teacher ratios) wasunrelated to variation in student achievement on standardized tests. In the decades following the release of the Coleman Report, the effect of school spending on student academic performance was studied extensively, and Coleman’s conclusion was widely upheld.
Given that substantial funding is needed to hire teachers and staff, purchase instructional materials, and maintain facilities, the lack of a positive relationship between school spending and student outcomes is surprising. Two key limitations of previous studies, however, make it difficult to draw firm conclusions from their results—limitations that we address in this study.
The first limitation is that test scores are imperfect measures of learning and may be only weakly linked to important long-term outcomes such as adult earnings. Yes, many interventions that boost test scores, such as being assigned to an effective teacher, have been shown to generate substantial gains in later earnings (see “Great Teaching,” research, Summer 2012). But several recent studies have also shown that effects on adult outcomes may go undetected by test scores. We address this limitation by focusing on the effect of school spending on such long-run outcomes as educational attainment and earnings rather than on test scores.
The second limitation of previous work is that most national studies simply examine correlations between observed changes in school spending and changes in student outcomes. This is problematic because many changes in how schools are funded are designed to provide additional resources to districts at risk of low performance. For example, the federal Elementary and Secondary Education Act allocates additional funding to school districts with a high percentage of low-income students, who are more likely to have poor educational outcomes for reasons unrelated to school quality. Such compensatory policies generate a negative relationship between changes in school spending and student outcomes that would bias analyses of the effects of school spending based on correlations alone.
We overcome this second limitation by focusing on the effects of exogenous shocks to school spending, that is, shocks that should be unrelated to family and neighborhood characteristics or the characteristics of any particular district or school. The exogenous shocks we use are the passage of court-mandated school-finance reforms (SFRs). In order to remove the confounding influence of unobserved factors that have an impact on both school spending and student outcomes, we calculate how much spending in a given school district would have been predicted to change due solely to the passage of an SFR, and use that prediction, rather than the spending change the district actually experienced, as our key variable. We then see if, within districts predicted to experience larger reform-induced spending increases, “exposed” cohorts (children young enough to have been in school when or after the reforms were passed) have better outcomes than “unexposed” cohorts (children who were too old at the time of passage to be affected by the reforms).
Our findings provide compelling evidence that money does matter, and that additional school resources can meaningfully improve long-run outcomes for students. Specifically, we find that increased spending induced by SFRs positively affects educational attainment and economic outcomes for low-income children. While we find only small effects for children from nonpoor families, for low-income children, a 10 percent increase in per-pupil spending each year for all 12 years of public school is associated with roughly 0.5 additional years of completed education, 9.6 percent higher wages, and a 6.1-percentage-point reduction in the annual incidence of adult poverty.
To document the causal relationship between school spending and long-run outcomes, we isolate variation in spending that occurred in response to the passage of court-mandated SFRs. What do these finance reforms look like, and how do they affect school districts?
In most states, prior to the 1970s, the majority of resources spent on K–12 schooling was raised at the local level, through local property taxes. Because the local property tax base is typically higher in areas with higher home values, and there are persistently high levels of residential segregation by socioeconomic status, heavy reliance on local financing enabled affluent districts to spend more per student. In response to lawsuits that identified large within-state differences in per-pupil spending across wealthy and poor districts, state supreme courts overturned school-finance systems in 28 states between 1971 and 2010, and many state legislatures implemented reforms that led to major changes in school funding. SFRs that began in the early 1970s and accelerated in the 1980s caused some of the most dramatic changes in the structure of K–12 education spending in U.S. history.
Most SFRs changed spending formulas to reduce differences in per-pupil spending across districts within a state. To document the equalizing effect of these reforms, Figure 1 compares the changes in spending in previously low-spending and high-spending districts during the 10 years leading up to a court-mandated SFR and the two decades that followed. We classify districts as low- or high-spending based on whether their average per-pupil spending levels were in the bottom or top 25 percent of districts in their state as of 1972, before any such reforms were implemented.
We see that court-mandated reforms were in fact successful at reducing spending gaps between previously low- and high-spending districts. In states that passed SFRs, low-spending districts initially experienced greater increases in per-pupil spending than similar districts in nonreform states, while high-spending districts experienced decreases. This general pattern was sustained over time.
Having established that court-mandated reforms, on average, affected school spending differently in different kinds of districts, we use more detailed information about the specific reforms enacted in each state to “predict” reform-induced spending changes for each district nationwide. That is, we ignore what actually occurred in a given district and instead calculate what would have been expected to occur based on the experiences of all other districts with similar characteristics experiencing the same kind of reform. We can therefore be confident that these predicted spending changes are unrelated to any unobserved changes in that particular district that may have influenced both school spending and adult outcomes.
The basic idea behind this approach is as follows: if certain kinds of reforms have systematic and predictable effects on certain kinds of school districts, then one can predict district-level changes in school spending based only on factors that are unrelated to potentially confounding changes in unobserved determinants of school spending and student outcomes (e.g., local commitment to education or the state of the local economy). With this clean, predicted variation in spending, one can then test whether in those districts that are predicted (based on pre-reform characteristics) to experience larger reform-induced spending increases, cohorts exposed to the reform have better outcomes than unexposed cohorts. By correlating outcomes with only the reform-induced variation in school spending (rather than all variation in spending), one removes the confounding effect of unobserved factors that might influence both school spending and student outcomes.
Of course, this strategy is only viable to the extent that one’s predictions of spending increases are reasonably accurate. Fortunately, we are able to examine actual spending in each district to confirm that, after reforms, districts with larger predicted spending increases experienced larger actual spending increases. Figure 2a shows that exposed cohorts in reform districts predicted to experience larger per-pupil school spending increases did exactly that, while exposed cohorts in reform districts predicted to experience smaller spending increases saw little change in school spending. Importantly, as our results show, predicted increases in per-pupil spending induced by SFRs are correlated not only with actual spending increases, but with improved outcomes for students as well.
Impact on Educational Attainment
Because test scores are not necessarily the best measure of learning or of likely economic success, we examine instead the relationships between SFR-induced spending increases and several long-term outcomes: educational attainment, high school completion, adult wages, adult family income, and the incidence of adult poverty. Our data on these outcomes come from the Panel Study of Income Dynamics (PSID), a survey that has tracked a nationally representative sample of families and their offspring since 1968. In particular, we use information on the roughly 15,000 PSID sample members born between 1955 and 1985, who have been followed into adulthood through 2011.
We find that predicted school spending increases are associated with higher levels of educational attainment. Figure 2b illustrates the effects of reform-induced changes in per-pupil spending on years of schooling completed. One can see clear patterns of improvement for exposed cohorts in districts with larger predicted spending increases. Cohorts with more years of exposure to higher predicted spending increases have higher completed years of schooling than cohorts from the same district who were unexposed or had fewer years of exposure. Also, the increases associated with exposure are larger in districts with larger predicted increases in spending (the line for districts with high predicted increases is consistently above that of districts with low predicted increases for the exposed cohorts). The patterns in timing and in intensity strongly indicate that policy-induced increases in school spending were in fact responsible for the observed increases in educational attainment. Taking into account the relationship between predicted and actual spending increases, we find that increasing per-pupil spending by 10 percent in all 12 school-age years increases educational attainment by 0.3 years on average among all children.
Because prior research has shown that children from low-income families may be more sensitive to changes in school quality than children from more-advantaged backgrounds, we also separately examine the effects of spending on low-income and nonpoor children. We define children as being low-income if their family’s annual income fell below two times the federal poverty line at any point during childhood.
For children from low-income families, increasing per-pupil spending by 10 percent in all 12 school-age years increases educational attainment by 0.5 years. In contrast, for nonpoor children, a 10 percent increase in per-pupil spending throughout the school-age years increases educational attainment by less than 0.1 years, and this estimate is not statistically significant.
To put these results in perspective, the education gap between children from low-income and nonpoor families is one full year. Thus, the estimated effect of a 22 percent increase in per-pupil spending throughout all 12 school-age years for low-income children is large enough to eliminate the education gap between children from low-income and nonpoor families. In relation to current spending levels (the average for 2012 was $12,600 per pupil), this would correspond to increasing per-pupil spending permanently by roughly $2,863 per student.
Predicted spending increases are also associated with greater probabilities of high school graduation, with larger effects for low-income students than for their nonpoor peers. Specifically, increasing per-pupil spending by 10 percent in all 12 school-age years increases the probability of high school graduation by 7 percentage points for all students, by roughly 10 percentage points for low-income children, and by 2.5 percentage points for nonpoor children. Figure 3 highlights the difference in effect size for these two childhood family-income groups and illustrates the closing of the high-school-graduation-rate gap between low-income and nonpoor children as a result of reform-induced spending increases.
In short, increases in school spending caused by SFRs lead to substantial improvements in the educational attainment of affected children, with much larger impacts for children from low-income families.
Impact on Adult Economic Outcomes
Our analyses also reveal sizable effects of increased school spending on low-income children’s labor market outcomes and their economic status as adults. For children from low-income families, increasing per-pupil spending by10 percent in all 12 school-age years boosts adult hourly wages by $2.07 in 2000 dollars, or 13 percent (see Figure 4). In contrast, the estimated effect of spending increases on wages for children from nonpoor families is small and statistically insignificant.
Increased per-pupil spending also has a positive effect on exposed students’ family income in adulthood. For children from low-income families, increasing per-pupil spending by 10 percent in all 12 school-age years increases family income by 17.1 percent. For children from nonpoor families, the estimated effect is small and not statistically significant. Effects on family income may reflect a) increases in one’s own income,
b) increases in other income due to increases in the likelihood of being married, or c) increases in the income of one’s family members (which is likely if children tend to marry individuals who were also affected by spending increases). Consistent with the effects on family income, which reflect, in part, any family composition effects, we find that, among low-income children, a 10 percent spending increase is associated with a 10-percentage-point increase in the likelihood of being married and never divorced. Spending increases have no effect on the probability of ever being married, however, suggesting that the higher marriage rates reflect higher levels of marital stability.
Our final measure of overall economic well-being is the annual incidence of adult poverty. Because this is an undesirable outcome, lower numbers are better. Our analysis finds that for children from low-income families, increasing per-pupil spending by 10 percent in all 12 school-age years reduces the annual incidence of poverty in adulthood by 6.1 percentage points. The effect for children from nonpoor families is once again small and statistically insignificant.
In summary, for children from low-income families, predicted increases in school spending are associated with increases in adult economic attainment in line with their educational improvements, and likely reflect improvements in both the quantity and quality of education received. Taken together, these analyses show that increased school spending caused by SFRs had important positive effects on adult wages, family income, and poverty status.
As mentioned previously, a large literature inspired by the Coleman Report has compared outcomes of individuals exposed to different levels of school spending without accounting for the possibility that changes in spending may have resulted from factors that also directly affect the outcomes of interest. One of the benefits of our approach is that we exploit only plausibly exogenous variation in school spending that is driven by court-mandated reforms.
We confirm that our approach generates significantly different results than those that use observed increases in school spending, by comparing our results to those we would have obtained had we used actual rather than predicted increases as our measure of changes in district spending. For all outcomes, the results based simply on observed increases in school spending are orders of magnitude smaller than our estimates based on predicted SFR-induced spending increases, and most are statistically insignificant.
This stark contrast provides an explanation for why our estimates differ from those of other influential studies in the literature, including the Coleman Report itself. We suspect prior studies that relied on variation in actual spending and found only modest effects of school spending may have been influenced by unresolved biases.
Another possible explanation for our findings of large school-spending effects is that how the money is spent matters a lot and that districts use the resources that come from unexpected increases in school spending more productively than they use other resources. Given that money per se will not necessarily improve student outcomes (for example, using the funds to pay for lavish faculty retreats or to shore up employee pension funds will likely not have a large positive effect on student outcomes), understanding how the increased funding was spent is key to understanding why we find large spending effects where others do not.
To shed light on the causal pathways through which education spending affects adult outcomes, we examine the effects of court-mandated spending increases on spending for school support services, physical capital, and instruction. We also estimate effects on student-to-teacher ratios, student-to-guidance-counselor ratios, teacher salaries, and the length of the school year.
We find that when a district increases per-pupil school spending by $100 due to reforms, spending on instruction increases by about $70, spending on support services increases by roughly $40, spending on capital increases by about $10, while there are reductions in other kinds of school spending, on average. While instructional spending makes up about 60 percent and support services make up about 30 percent of all total school spending, the two categories account for about 70 percent and 40 percent of the marginal increase, respectively. This suggests that exogenous increases in school spending are more likely than other forms of school spending to go to instruction and support services. The increases for instruction and for support services (which include expenditures to hire more teachers and/or increase teacher salaries along with funds to hire more guidance counselors and social workers) may help explain the large, positive effects for students from low-income families.
We also examine the effects of court-mandated spending increases on three commonly used proxies for school quality: the length of the school year, teacher salaries, and student-teacher ratios. We find that a 10 percent increase in school spending is associated with about 1.4 more school days, a 4 percent increase in base teacher salaries, and a 5.7 percent reduction in student-teacher ratios. Because class-size reduction has been shown to have larger effects for children from disadvantaged backgrounds, this provides another possible explanation for our overall results.
While there may be other mechanisms through which increased school spending improves student outcomes, these results suggest that the positive effects are driven, at least in part, by some combination of reductions in class size, having more adults per student in schools, increases in instructional time, and increases in teacher salaries that may help to attract and retain a more highly qualified teaching workforce.
Previous national studies have examined the relationship between school resources and student outcomes and found little association for students born after 1950. Those studies, however, suffer from major design limitations. We address those limitations and demonstrate that, in fact, when examined in the right way, it becomes clear that increased school spending is linked to improved outcomes for students, and for low-income students in particular. Investigating the causal effect of school spending increases generated by the passage of SFRs, we conclude that increasing per-pupil spending yields large improvements in educational attainment, wages, and family income, and reductions in the annual incidence of adult poverty for children from low-income families. For children from nonpoor families, we find smaller effects of increased school spending on subsequent educational attainment and family income in adulthood.
Taken together, these results highlight how improved access to school resources can profoundly shape the life outcomes of economically disadvantaged children and thereby reduce the intergenerational transmission of poverty. Money alone may not lift educational outcomes to desired levels, but our findings confirm that the provision of adequate funding may be critical. Importantly, we also find that how the money is spent matters. Therefore, to be most effective, spending increases should be coupled with systems that help ensure spending is allocated toward the most productive uses.
This article is based on the full paper “The Effects of School Spending on Educational and Economic Outcomes: Evidence from School Finance Reforms,” The Quarterly Journal of Economics (forthcoming).
Rucker C. Johnson is associate professor of public policy at University of California, Berkeley. C. Kirabo Jackson is associate professor of human development and social policy at Northwestern University. Claudia Persico is a doctoral candidate in human development and social policy at Northwestern University.
This article was originally posted on EducationNext. Read Eric A. Hanushek's response to this article titled, “Money Matters After All?” here, and the three author's counter-response in an article titled, “Money Does Matter After All” here.
The last seven years have been disastrous for many workers, particularly for lower-wage workers with little education or formal training, but also for some college-educated and higher-skilled workers. One explanation is that lackluster wage growth and, until recently, high unemployment reflect cyclical conditions—a combination of a lack of demand in the U.S. economy and greater sensitivity of workers on the bottom-rungs of the job ladder to changes in the business cycle. A second explanation attributes stagnant wages and employment losses to structural changes in the labor market, including long-term industrial and demographic shifts and policy changes that reduce the incentive to work. This explanation interprets recent trends as the “new normal” and suggests that the U.S. economy will never return to pre-recession labor market conditions unless policies are changed dramatically.
My research, based on a review of extensive data on labor market outcomes since the end of the Great Recession of 2007-2009, finds no basis for concluding that the recent trend of stagnant wages and low employment is the “new normal.” Rather, the data point to continued business cycle weakness as the most important determinant of workers’ outcomes over the past several years. It is only in the past few months that we have started to see data consistent with growing labor market tightness, and even this trend is too new to be confident. The continued stagnation of wages through the end of 2014 implies that, at a minimum, a fair amount of slack remained in the labor market as of that late date. In turn, policies that would promote faster recoveries and encourage aggregate demand during and after recessions remain key policy tools.
Why is this relevant for policymakers?
Labor force participation rates are still down sharply since the onset of the Great Recession, but the unemployment rate, which spiked from 5 percent to 9.5 percent during the recession, has almost returned to its pre-recession level. If the low participation rate reflects structural economic changes then the current labor market is the “new normal” and there is not much that policymakers can do to improve short-term performance. If instead the problems are due to cyclical economic weakness, generating continued labor market slack that is hidden by the low unemployment rate, then there is much more scope for fiscal and monetary policy to improve labor market conditions. Clearly, cyclical and structural explanations imply vastly different policy responses.
A number of structural shifts have been suggested as explanations for the “new normal,” among them a reduction in workers’ willingness to take jobs (perhaps driven by changes in the incentives created by government transfer programs such as extended unemployment insurance), an aging population that creates shortages of younger workers, and rapid shifts in employers’ needs toward newer types of skills that are in short supply in the labor force. My examination of recent data finds little basis for any of these hypothesized changes. Rather, the evidence—most notably stagnant wages among those who are employed—suggests that lackluster employment growth from 2009 through at least the end of 2014 reflected a continued shortage of demand for virtually all types of workers. It is only in the most recent data—which may well be a temporary blip—that we start to see wage growth consistent with a tightening labor market. It is far too soon to conclude that structural changes will prevent a full recovery to pre-recession labor force participation rates. In the meantime, it will be important to have accommodative fiscal and monetary policies, lest we strangle the belated, still nascent recovery in its infancy. What little wage growth we have seen to date suggests little reason to worry that increases in demand for labor above the current level will trigger meaningful wage inflation.
What do the data say?
The unemployment rate has been below 6 percent since September 2014, lower than many estimates of the level consistent with “full employment.” (Even in a full-employment labor market, we would expect some unemployment as workers transition from one job to another.) Yet the employment-to-population ratio—the share of working-age adults who hold jobs—has been much slower to recover after the Great Recession, and remains lower than was seen at any point between 1984 and 2009. The difference between these measures of labor market slack reflects a sharp decline since 2007 in the share of the population that is participating in the labor market. These declines have continued throughout the recovery, and show no sign of being reversed.
Diagnoses of the situation have thus depended on which data series one chooses to emphasize. The unemployment rate data suggested a robust recovery from early 2011 onward. By 2014, the economy appeared to have little room left to improve, leading some to conclude that the still low employment rate and weak wage growth must have been the “new normal.” But the employment rate series suggested that there remained substantial slack left in the labor market throughout the period as four percent of the population who had been employed before 2007 but were not being pulled back into the labor market. Neither data series in isolation could reveal the true state of the labor market.
To distinguish between these “glass-half-full” and “glass-half-empty” views, I look to evidence regarding employment and wage growth by industry and demography, seeking indications of imbalances between labor supply and demand. If the labor market in 2013-14 was as tight as the unemployment rate alone indicated then we should have seen wage increases as employers bid against each other for workers who were in increasingly short supply. By contrast, if wage growth remained anemic throughout the period, and if employment shortfalls were spread evenly across high- and low-skill demographic groups, then that would be an indication that the unemployment rate was misleading and that the labor market remained quite slack.
Findings by industry
One potential source of structural problems is an imbalance between employers’ needs and the skills being offered by job seekers. Rapid technological changes can lead to increases in the demand for workers with specialized skills, yet slack might still remain in other parts of the labor market. There is clear evidence of this sort of imbalance in the mining and logging sector, which has grown substantially since before the recent recession and where there are clear signs that employers are having trouble finding workers to fill open jobs. But outside of this sector, there is little sign that demand growth has been disproportionately concentrated in sectors such as information and technology that typically require specialized skills.
Rather, job openings have grown most in sectors such as transportation, lodging and food services, and arts and recreation. These data generally appear consistent with the view that the increase in job openings reflects reduced recruiting efforts, lower starting wages, or higher minimum qualifications rather than shortages of qualified workers. (See Figure 1.)
It also is possible that demand for labor within certain industries created shortages of some particular types of workers that are masked by weakness in other subsectors. This explanation is perhaps most plausible for the finance and information sectors, where one can easily imagine shortages of workers with industry-specific skills. The information sector, where technological changes requiring new skills are most likely to be an important component of labor demand, and thus where structural labor supply shortages are most plausible, has had only a modest increase in job openings, and total employment remains below its 2007 level.
Findings by demography
Another source of evidence about mismatches between workers skills’ and firms’ needs lies in the demographic distribution of unemployment. In the recent recession, unemployment rose much more for non-college workers than for those who had attended college, and at each education level more for men than for women. The latter likely reflects the disproportionate declines in construction and manufacturing, which are cyclically sensitive industries that were very hard hit in this cycle. The former could be consistent with a shift in favor of higher-skill workers.
But data from the subsequent economic recovery contradict this explanation. The unemployment rate fell faster in the recovery for less-skilled workers than for college-educated workers, and particularly fast for non-college men. There is no indication that the unemployment rate for college-educated workers has reached any sort of a floor since it remains—even in the most recent data—notably higher than in 2007. (See Figure 2.)
Findings regarding wages
Ultimately, the most decisive way to diagnose the adequacy of labor demand is by examining wages: If employers are having trouble finding suitable workers then they will compete against each other for those workers who are available, bidding up wages. Across-the-board labor shortages would mean increases in wages across the economy while shortages for workers with specialized skills would mean raises in particular sectors.
If the economy were pushing against overall limits then we would expect to see rising wages. But the data through 2014 showed no signs of upward pressure on wages. Average real wages (adjusted for inflation) were stagnant since 2009, with increases below 1 percent per year even in 2014. Workers at the very top of the wage distribution saw larger increases, but even these totaled only 2 to 3 percent between 2008 and 2014, and they were concentrated among the top 20 percent of workers. Below the 80th percentile, real wages fell by about 3 percent at the median. It is only in the most recent data (since the beginning of 2015) where there is any sign of real wage growth, at roughly a 3 percent annual rate. If this is sustained, and especially if it accelerates in the coming months, then it might indicate that the labor market has finally begun to tighten. But a few months of data are too little to support this conclusion, particularly when real wage growth has been boosted by low inflation attributable to declines in energy prices. (See Figure 3.)
Over the longer period, there is no sign of meaningfully larger wage increases in sectors with rising job openings, as would be expected if these sectors faced persistent labor shortages. Across industries, only the mining and finance sectors appear to have posted meaningful wage increases, and even these have averaged less than 1 percent per year real wage growth. Once again, the patterns in the data are fully consistent with continued demand weakness, and not at all consistent with growing shortages of workers in growing sectors. (See Table 1.)
In the years since the Great Recession, the unemployment rate has gradually crept downward while other indicators of the health of the labor market have been stagnant. Lackluster wage growth and high unemployment rates among lower-skilled workers appear to be attributable to a continued shortage of demand in the U.S. economy, combined with greater sensitivity to cyclical conditions of workers on the bottom-rungs of the job ladder. That means the high nonemployment rate among lower-skilled workers is not the “new normal” but rather could be substantially resolved by more robust economic growth and better fiscal and monetary demand-management policies.
Further, my research suggests that increased aggregate demand for workers, at the current level, will not create inflation. At best, we are only in the past few months seeing meaningful tightness. More likely, this is an artifact of declining oil prices, which means the current labor market still has substantial slack. Under the latter interpretation, additional labor demand would improve employment outcomes, with particular benefits for low-skilled workers and other disadvantaged groups who suffer disproportionately from cyclical downturns.
My results also counsel against many of the recommendations made by proponents of the view that the economy has settled into a “new normal.” In particular, there are two ill-advised responses to current conditions in the labor market predicated on a misdiagnosis of the economy as having escaped the cyclical downturn. First, tax cuts or reductions in unemployment insurance or means-tested government transfer programs aimed at increasing labor supply will do more to reduce wages than to increase employment.
Second, education and training programs aimed at increasing the skill of low-wage workers are unlikely to do much to help the labor market when there are demand shortages at every rung of the job ladder. So education and training programs are unlikely to help in the short term. That said, these programs alongside increased income support for low earners still make sense as a response to long-term trends—even if they cannot be expected to contribute meaningfully in the short run.
Taking ill-advised policy steps, such as failing to implement needed fiscal and monetary policies to boost demand for labor, or, worse, implementing policies aimed at tamping down an overheating economy, could extend periods of underemployment, damaging workers’ productivity for many years to come. Every month that the economy continues to underperform is making us poorer for decades into the future. Over-cautious policy could cause substantial damage. It is also crucial to put policies in place now to prepare for the next downturn, to avoid such a sustained, weak recovery.
Claims that the economy is nearing its growth potential, and that ongoing low employment rates are the unavoidable consequence of structural changes in the labor market, are at odds with the evidence. Neither comparisons across industries or education groups, nor analyses of wage growth offer any evidence of tight labor markets pushing up against their limits. Unemployment rates remain higher than in 2007 for all ages, education levels, genders, and industries. Sectors that have been more cyclically sensitive in the past saw larger increases in unemployment in the Great Recession, but there is remarkably little difference beyond this observation in the current data. And wages have continued to stagnate for the vast majority of workers, at least until the very most recent data. All of these patterns are consistent with an ongoing shortfall in aggregate labor demand, and less so with a gradual adjustment to technological or demand-driven shocks that created demand for new types of skills that cannot be satisfied by the current workforce.
—Jesse Rothstein is an associate professor of public policy and economics and the director of the Institute for Research on Labor and Employment at the University of California, Berkeley. This featured research was originally posted on the Washington Center for Equitable Growth.