Working Papers
Is Special Education a Pathway to Supplemental Security Income for Children?
[Working paper]
Abstract: 

Between 1989 and 2013 the number of children receiving Supplemental Security Income (SSI) grew from 0.26 million to 1.3 million. Multiple explanations have been put forward including an increase in the number of children diagnosed with mental impairments and changes in SSI rules that made it easier for children receiving special education services to apply for SSI. I document a direct pathway from receipt of special education to SSI using a two sample fuzzy regression discontinuity design. First, using administrative records from North Carolina, I corroborate earlier findings that children born the month before the kindergarten entry eligibility cutoff date are more likely to receive special education services relative to children born the month after the school cutoff date. Next, using National Health Interview Survey respondents linked to Social Security Administration records, I document that the children born just before the cutoff date are 0.78 percentage points (or 30%) more likely to apply for and 0.55 percentage points (or 59%) more likely to receive an award for SSI between the ages of 5 and 12 relative to children born just after the school cutoff date. I find no increase in awards among groups unlikely to be affected by the relationship between school-starting-age and special education; these include children with physical impairments or those too young for school enrollment. Two sample fuzzy RD estimates indicate that a 1 percentage point increase in the fraction of children receiving special education services induces a 0.16 percentage point (or 10%) increase in the fraction of children with an SSI award. Back of the envelope calculations suggest that approximately 18% of the growth in the SSI caseload can be attributed to rising rates of special education and spillovers between these two programs.

Relative Age-in-Grade and Child Disability: Evidence from North Carolina 
Using administrative student level data and regression discontinuity methods, I corroborate earlier research indicating that in the early grades (grades 3-8) the youngest children in the classroom are more likely to be classified for other health impairments (including attention deficit hyperactivity disorder (ADHD)), speech impairments, and emotional/behavioral disorders. I also corroborate findings that the youngest children in the classroom score below their older classmates on end of grade exams in both math and reading in grade 3. While these effects diminish with each subsequent grade, younger students are still performing 0.06 standard deviations below their older classmates in math and 0.08 standard deviations below their older classmates in reading at the end of grade 8. I augment the existing literature by showing that the youngest children are more likely to be diagnosed for specific learning disabilities in mathematics and reading, in all grades, and that these relative-age-in-grade effects for special education placement for other health impairments and specific learning disabilities persist through 12th grade. In addition, contrary to Dhuey and Lipscomb (2010) I find similar effect sizes for both boys and girls given that the base diagnosis rate for girls is smaller than the base diagnosis rate for boys.
Interactions Among Social Safety Net Programs: Evidence from the Earned Income Tax Credit and Supplemental Security Income for Children 
In 2017, approximately 1.2 million children between the ages of 0 and 18 received Supplemental Security Income (SSI) payments totaling just over $9.6 billion. Few studies have examined how aspects of the current social safety net interact with the SSI program. Using microdata on children aged 15-18 in the American Community Survey (ACS) I generate a measure of exposure to the Earned Income Tax Credit (EITC), which reflects the cumulative maximum credit available to that individual based on their year of birth, state of residence, and number of qualifying children within the household. Using EITC exposure as the measure of treatment, I implement a difference-in-differences strategy to identify the effect of the EITC on the child SSI program. My strategy leverages three sources of variation: federal expansions in the EITC for various family structures in 1990, 1994, and 2009; introduction of state EITCs; and changes in family structures
across ages.
Beyond Reading, Writing, and Arithmetic: The Role of Teachers and Schools in Reporting Child Maltreatment (with Sam Bondurant and Maria Fitzpatrick)
Estimates show that 4 in 10 children experience some form of maltreatment by the time they are teenagers.  However, we know little about the role of reporters.  For example, although educators report many instances of child maltreatment, it could be the case that they are identifying and reporting maltreatment that would be reported by others.  We study the role of teachers and other educators in reporting child maltreatment by causally identifying the effect of exposure to school on child maltreatment reporting. Unique administrative data on nearly all reported cases of child maltreatment across the U.S. over a 14 year period allows us to use two different regression discontinuity methods, one based on school entry laws and the other based on school calendars. Both methods show an increase in reports by educators due to time in school that is not accompanied by a decrease in reports by others.  Overall, reports increase by 5 to 10 percent.  The additional reports due to time in school (and the resulting time with educators) are no less likely to be substantiated than other reports, suggesting that educators are no more likely to over-report than others. Our results indicate that educators play an important role in the early detection of child maltreatment.
Increasing the Share of Female Faculty within Humanities Departments: Does the Gender of University Leaders Matter? Cornell Higher Education Research Institute Working Paper 170 (with Todd Jones, Sarah Prenovitz, and Ronald Ehrenberg) [Working Paper]
Do Administrators' Disciplinary Backgrounds Influence Humanities Departments' Staffing Patterns? Cornell Higher Education Research Institute Working Paper 168 (with Todd Jones, Sarah Prenovtiz, and Ronald Ehrenberg) [Working Paper]
Conference Proceedings
"Bidding For Classes: Course Allocation Under the Colorado College Auction System" (with Kristina Lybecker and Daniel Johnson), 2014, Conference Proceedings of the Academy of Economics and Finance; also published as a Colorado College Working Paper 2013-2 [Working Paper]