Nudging teachers to underserved schools at zero cost

Teachers are not distributed evenly across schools. More qualified teachers tend to be concentrated in more advantaged schools, while low-income schools are more likely to face shortages of qualified staff. This matters because teachers are among the most important school inputs, and their impact is often greatest for low-performing and disadvantaged students (Araujo et al. 2016, Bertoni et al. 2020). As a result, this unequal distribution widens achievement gaps and limits equality of opportunity.

One key driver of teacher sorting is that candidates tend to disproportionately apply to more desirable, socioeconomically advantaged schools. In some systems, this dynamic is reinforced by hiring rules that prioritise candidates with higher scores or credentials, who are, in turn, more likely to secure positions in these more competitive vacancies.

Policymakers typically address this problem by offering financial incentives to attract teachers to hard-to-staff schools. However, salary increases are costly and their effects tend to be modest (Clotfelter et al. 2008, Falch 2011, Glazerman et al. 2012, Springer et al. 2016, Rosa 2017, Bueno and Sass 2018, Feng and Sass 2018, Elacqua et al. 2022). To complement these efforts, we propose a low-cost behavioural intervention aimed at reducing the sorting of candidates across teaching vacancies (Ajzenman et al. forthcoming). Our intervention was implemented nationwide in Ecuador as part of their teacher recruitment process.

Teacher selection and market congestion in Ecuador

Since 2013, Ecuador's Ministry of Education has selected and assigned teacher candidates to school vacancies through a centralised process known as Quiero Ser Maestro (QSM). Our intervention was conducted in 2019, during the sixth edition (QSM6), which comprised three phases:

1. Eligibility phase: candidates must pass standardised subject and psychometric tests to qualify for the process.
2. Merits and public examination phase: candidates receive a composite score based on exam results and verified credentials (e.g. education and experience).
3. Application phase: Eligible candidates apply to and rank up to five school vacancies on an online platform, after which assignments are made through a centralised matching algorithm (similar to a deferred acceptance algorithm).

Although Quiero Ser Maestro has improved transparency, Ecuador's teacher selection process still generates inefficiencies and inequities. While some schools receive more applications than available vacancies, others struggle to attract applicants. As a result, a large share of teaching positions remains unfilled at the end of the process, and many candidates are unable to secure a job offer.

The intervention: Listing hard-to-staff schools first

In the experiment, candidates were randomly assigned to one of two groups, stratified by district of residence:

• Control group: Vacancies were displayed in alphabetical order.
• Treatment group: Hard-to-staff school vacancies were displayed first.

The design exploits a well-documented behavioural phenomenon: people tend to give disproportionate weight to options that appear first in a list, especially when facing many alternatives. Rather than changing incentives or restricting choice, the intervention simply reordered the vacancy list to put hard-to-staff schools at the top, leveraging this order effect to nudge teachers toward schools that typically struggle to attract applicants.

The Ministry of Education classified schools as ‘hard-to-staff’ based on a high proportion of unfilled vacancies in prior selection processes, a high share of teachers on temporary contracts, poor infrastructure, and low-performing teachers and students. On the platform, these schools were marked with an icon and a label indicating that in such institutions, teachers could have a high social impact.

Crucially, the information shown to teachers did not change across arms; the only difference was the order in which options appeared.

Order effects in a high-stakes context

We found strong order effects: candidates in the treatment group were 4.3 percentage points (pp) more likely to rank a hard-to-staff school as their first choice (control mean: 29.8%), 2.5 pp more likely to include at least one hard-to-staff school among their first two choices (control mean: 46.3%), and 0.9 pp more likely to include a hard-to-staff school in their choice set (control mean: 32.1%).

We also found impacts on the final allocation of candidates. The probability of being assigned to a teaching position in a hard-to-staff school was 1.9 pp higher in the treatment group (control mean: 19.9%). In short, treatment candidates were more likely to apply to hard-to-staff schools, rank them as their top choice, and be assigned to a vacancy in one of these schools.

Given that approximately 21% of the cohort was assigned to a hard-to-staff school, the intervention implies that the number of new public-school teachers allocated to hard-to-staff schools could increase by as much as 6.7% per year at zero cost.

What explains these order effects?

We explored different mechanisms that may be behind these effects. First, the results are not driven by inattentive or lower-performing teachers. We find no evidence that cognitive ability or limited attention explains the effects. Second, the effects are not stronger among teachers who report higher altruism, suggesting that simply priming a ‘social impact’ identity is not the main mechanism.

Instead, the results seem to be driven by choice overload. The order effects are larger when teachers face more vacancies. When confronted with many options, candidates may rely on simpler decision rules, giving greater weight to options that appear first.

Interestingly, although hard-to-staff schools are on average located farther from candidates’ homes, the intervention did not increase commuting distances. This suggests that teachers were not ignoring distance altogether. Rather, they selected schools that were satisfactory on key dimensions, such as commuting time, even if they were not fully optimising across all alternatives. 

Finally, we find no difference in turnover across arms two years after implementation. Importantly, teachers assigned to hard-to-staff schools through the intervention stayed at the same rates as controls, suggesting the placements were stable. Consistent with this, a government satisfaction survey conducted at the same time shows similar levels of job satisfaction across groups.

Policy implications

Teacher sorting is a major concern for policymakers. Because teachers have short- and long-term effects on students' educational outcomes – especially among the most vulnerable (Aaronson et al. 2007, Araujo et al. 2016) – teacher shortages and a preponderance of temporary and non-certified teachers in disadvantaged schools can exacerbate socioeconomic inequalities. Moreover, the concentration of applications in more advantaged schools is inefficient and reduces candidates' chances of securing a job.

The intervention succeeded on dimensions crucial for equity. Teachers in the treatment group were not only more likely to apply to hard-to-staff schools, serving poorer and lower-performing students, but also more likely to remain in them. Moreover, their substitution of oversubscribed vacancies for undersubscribed ones suggests the intervention, if scaled, would likely reduce the number of unfilled positions. Addressing teacher shortages is challenging, and while traditional policies such as financial incentives play an important role, they may not fully close existing gaps given their costs. In this context, this zero-cost intervention offers a valuable complement to ongoing efforts to reduce teacher sorting and improve allocation.

References

Aaronson, D, L Barrow, and W Sander (2007), "Teachers and student achievement in the Chicago public high schools," Journal of Labor Economics, 25(1): 95–135.

Ajzenman, N, G Elacqua, L Marotta, and A S Olsen (forthcoming), "Order effects and teacher labor supply: A nation-wide RCT in Ecuador," American Economic Journal: Applied Economics.

Araujo, M C, P Carneiro, Y Cruz-Aguayo, and N Schady (2016), "Teacher quality and learning outcomes in kindergarten," Quarterly Journal of Economics, 131(3): 1415–1453.

Bertoni, E, G Elacqua, L Marotta, M Martínez, C Méndez, V Montalva, A S Westh Olsen, S Soares, and H Santos (2020), "El problema de la escasez de docentes en Latinoamérica y las políticas para enfrentarlo," Unpublished manuscript.

Bueno, C, and T R Sass (2018), "The effects of differential pay on teacher recruitment and retention," Unpublished manuscript.

Clotfelter, C, E Glennie, H Ladd, and J Vigdor (2008), "Would higher salaries keep teachers in high poverty schools? Evidence from a policy intervention in North Carolina," Journal of Public Economics, 92(5–6): 1352–1370.

Elacqua, G, D Hincapie, I Hincapié, and V Montalva (2022), "Can financial incentives help disadvantaged schools to attract and retain high-performing teachers? Evidence from Chile," Journal of Policy Analysis and Management, 41(2): 603–631.

Falch, T (2011), "Teacher mobility responses to wage changes: Evidence from a quasi-natural experiment," American Economic Review, 101(3): 460–465.

Feng, L, and T R Sass (2018), "The impact of incentives to recruit and retain teachers in 'hard-to-staff' subjects," Journal of Policy Analysis and Management, 37(1): 112–135.

Glazerman, S, A Protik, B Teh, J Bruch, and N Seftor (2012), "Moving high-performing teachers: Implementation of transfer incentives in seven districts," Unpublished manuscript.

Rosa, L (2017), "Teacher preferences in developing countries," Unpublished manuscript.

Springer, M G, W A Swain, and L A Rodriguez (2016), "Effective teacher retention bonuses: Evidence from Tennessee," Educational Evaluation and Policy Analysis, 38(2): 199–221.