An information intervention and consent to data linkage : experimental evidence from teaching

Using new survey data of teachers in England we investigate the propensity for teachers to consent to data linkage, differences by observable characteristics and the effect of a randomly assigned information intervention. We find that consent rates are high (75 percent), possibly due to the relationship between participating schools and the research team, but observe differences by ethnicity, sex, and sector - teachers from a non-white background, male teachers and those who work in the independent sector are significantly less likely to consent. While we find that the provision of additional information does not increase consent to data linkage our heterogeneity analysis shows that the information treatment has a large, positive, effect on teachers who work in the independent sector - a subgroup of teachers who have a significantly lower rate of consent.