Waiting-time targets in healthcare markets: How long are we waiting?

Waiting-time targets are frequently used by policy makers in the healthcare sector to monitor provider's performance. Such targets are based on the distribution of the patients on the list. We compare and link such distribution with the distribution of waiting time of the patients treated, as opposed to on the list, which is arguably a better measure of welfare or total disutility from waiting (although it can only be calculated retrospectively). We show that the latter can be estimated from the former, and viceversa. We also show that, depending the hazard function, one distribution may be more or less favourable than the other. However, empirically we find that the proportion of patients waiting on the list more than x months is a downward estimate of the proportion of patients treated waiting more than x months, therefore biasing downwards the total disutility from waiting.