2026/07/14 by Tensay Hadush Meles, Gunnar Köhlin
paper · doi:10.1111/obes.70110
ABSTRACT Leveraging over 40 million residential electricity billing records from more than one million meters in Ethiopia, we test whether consumers respond to marginal prices at kink points under an increasing block tariff (IBT). We exploit a unique institutional setting in which two distinct billing systems (manual postpaid and automated prepaid) operate under a common IBT within a state‐owned utility, allowing us to test whether observed bunching reflects behavioural responses or measurement error. In the raw postpaid data, we observe bunching at kink points, but this bunching disappears after normalizing for billing‐cycle length and aggregating readings to smooth billing anomalies, suggesting the spikes reflect the billing process rather than underlying behavioural responses to marginal prices. In contrast, prepaid data show no bunching at kink points but exhibit spikes at round‐number recharge amounts, suggesting rule‐of‐thumb purchasing behaviour. These findings highlight that, in a setting with low electricity prices, no evidence of marginal price responses, and widespread shared connections, the IBT is unlikely to achieve its intended goals of subsidy targeting, conservation, and cost recovery.