It is difficult to assess the time it takes for inputs into education or education policy reforms to affect learning outcomes. Attributing changes in results to system-wide reforms can be complex when there are many different programmes and elements affecting outcomes. Data are not always available on learning outcomes over time and may be complicated by changes in testing. Learning outcomes also vary between regions within countries.

Evidence found includes:

• Improved learning outcomes were recorded between 2001 and 2007 following School Improvement Plans in Namibia in 2000.  However, improvements were not found in all regions.
• Test results in El Salvador don’t appear to have changed following a series of policy and system changes since 1992. However, tests have been revised making it difficult to compare results over time.
• Large-scale reform in Argentina occurred between 1992 and 1994 when dependent secondary schools were transferred from national to provincial governments. Test outcomes between 1994 and 1998 were on average 1.2 standard deviations higher.

Evaluations of smaller-scale education projects have recorded changes in learning outcomes. Results vary between no change in learning outcomes, improved learning over a few months and improved outcomes over a number of years. There were no studies that documented a point in time from the initiation of a programme that changes in outcome were achieved. Evaluations tend to occur either a year or a number of years after a specific input or change.

The aim of Education for All has sometimes adversely affected student outcomes. A significant amount of education reforms have tended to focus on improving access and availability rather than quality.