Low-stakes standardized tests in British Columbia, Canada: system accountability and/or individual feedback? (with Arthur Sweetman), Education Economics, https://doi.org/10.1080/09645292.2022.2091113
A jurisdiction-wide zero-stakes Foundational Skills Assessment administered in grade 4 in British Columbia, Canada, used a three-point scale to publicly disseminate aggregate school/district-level results, and a five-point scale to convey results to students/parents. For a variety of long-term outcomes, a regression discontinuity analysis shows a positive, but modest in magnitude, ‘system accountability’ effect for girls who fall just below the lowest threshold on the three-point scale for reading. For numeracy the findings are less precise but suggest a positive ‘feedback effect’ for girls who fall just below the highest five-point scale cut-off. No statistically significant effects are observed for boys.
More than a Basic Income: The Role of Training and Employment Support Programs (with Arthur Sweetman), Chapter 17. In D. Green, J. R. Kesselman, D. Perring, L. M. Tedds (Eds.), Basic Income and a Just Society: Policy Choices for Canadas Social Safety Net. McGill-Queens University Press, link
(Un)intended Consequences of Permanent School Closures: The Effects on Students, Parents, and Communities in British Columbia (JMP)
TBA
Return to an extra year of high school: evidence from Newfoundland and Labrador (with Arthur Sweetman)
In 1983-84, the province of Newfoundland and Labrador increased the number of years of high school from 11 to 12, introducing new (pre-)vocational courses but not substantially changing the core academic curriculum. We leverage this large-scale exogenous variation to estimate long-run returns to schooling in the presence of general equilibrium effects. First, using the 2001 Canadian Census, we demonstrate that this change increased the number of completed years of secondary school by around 0.5-0.6 for both sexes. Employing an instrumental variable (IV) approach, we estimate that when the treated are in their early thirties, the return to an additional year of schooling is around 10-11% of employment income. This effect translates to increased wealth as measured by average dwelling value. These effects persist into the affected group’s early fifties. We suggest two channels that partially explain the observed return. First, an additional year of schooling substantially raises migration out of the province, especially for males. Second, it changes the occupational structure with females switching to higher-paying occupations. However, we do not find evidence that the reform improved post-secondary completion rates. Moreover, the reform improved earnings without fundamentally reshaping the income distribution, with gains concentrated at or just above the median. We argue that the introduction of new (pre-)vocational courses plays a key role in the observed responses. Finally, we discuss the implications of the unique economic context of Newfoundland on the interpretation of the results.
Association between type of long-term care accommodations and mortality (with Andrew Costa, Arthur Sweetman, Ahmad Rahim, Jeff Poss, Darly Dash, Ahmad von Schlegell, Rhonda Collins)
It is well established that blind curriculum-based system-wide exam marks and teacher-assigned term marks are not perfectly correlated. However, little is known about their relationships with long-term labour market outcomes. We shed light on these issues using students’ grade 12 English marks in British Columbia, Canada, linked to tax data 13-15 years post-completion. From 1991/92 to 2019/20, the jurisdiction’s high school completion rate increased while average teacher marks rose much faster than exam marks. These two marks represent overlapping but different information sets. First, we find that the term-minus-exam mark gap is positively associated with earnings for students from English-speaking households, but negatively associated for English-as-a-Second-Language (ESL) students. Second, we document heterogeneity along two dimensions: (i) in schools with stricter grading—where teacher marks closely match or fall below exam scores—the teacher grade appears to be relatively more important for predicting labour market success, whereas in grade-inflating schools exam scores matter more; (ii) for English-speaking students, the gap’s positive association with earnings is strongest in occupations with high social-skill requirements, as classified by O*NET. Finally, we find important non-linearities in how the association between the term-exam mark gap and earnings varies across the distribution of final marks. Together, these results underscore the complementary value of teacher grades and standardized exams and suggest that eliminating the latter could erode important earnings signals, especially for language-minority students.
Association of nursing home room type and crowding with COVID‑19 infections and mortality in Ontario, Canada (with Andrew Costa, Arthur Sweetman, Ahmad Rahim, Jeff Poss, Darly Dash, Ahmad von Schlegell, Rhonda Collins)
Assessment in High School: The Predictive Power of Teacher and Exam Marks (with Arthur Sweetman)
Long-term consequences of dropping out of school in British Columbia: the case of falling below the passing threshold