ABSTRACT
In recent years, long-term evaluation of educational reforms has re-occurred as a hot topic in Swedish politics, not the least due to suspicions that educational reforms of the 1990s have, in fact, had long-term negative impacts on study results and equity among schools and regions. Also, the 1960s and 1970s saw evaluations and political debates on the long-term impacts of educational reforms, but since then the political landscape has changed, for instance by an increased political unwillingness to wait for evaluation results before implementing another reform. By comparing recent to earlier long-term school reform evaluations in Sweden, this chapter investigates what can be expected from long-term evaluation under the present circumstances. It elaborates on what kind of information best serves policy-making given the nature of this sector, discussing combinations of short- and long-term evaluation, forward mapping and backward mapping, and different causal theories and approaches that capture the sizes of achieved impacts (counterfactual difference) as well as the how and why of impacts (INUS-conditions and generative mechanisms). An ideal combination of approaches is probably difficult to maintain under high-speed and ideologically driven policy-making, although careful planning and thorough data collection is needed.
