In the late 1920s schooling in New South Wales (NSW) was organised as a liberal meritocratic system. It was liberal because it was focused on individuals and presented schooling as a means of enhancing individual development. It was meritocratic because merit, individual's pre-social talent or intelligence, rather than their birth, class or wealth, was taken to be a crucial determinant of educational opportunity and success. Gender was accepted as a justification for differentiated schooling. The liberal meritocratic schooling which had been established by the late 1920s was a universal system of educational provision in which all students were required to stay at school until they were 14 years old and in which public and private educational provision was regulated by the state. Within this universal system, differentiated educational routes for academic and non-academic students were constructed through schools of different types offering different academic and prevocational curricula. These routes led to different destinations in the paid and unpaid workforce. Students, in both public and private schools, who passed the necessary examinations and who could be maintained at school were able to climb the educational ladder to the university which gave access to the professions and better public service employment. Those students who did not pass the necessary exams, who chose pre-vocational rather than academic courses or who lacked the support to stay on eventually left school from whatever level in the system they had reached. Many left directly from primary school in this ability graded system. By the late 1930s there was still a universal system of schooling which required student attendance until the age of 14 and which was differentiated for academic and non-academic students. The academic and non-academic educational routes still led to different destinations in the labour market and the state still orchestrated educational provision. But in a number of respects the pattern of educational provision had shifted in important ways. The state was now widely seen to be responsible not only for primary but also for secondary education. While formal compulsory attendance requirements had not changed, there had been a de facto increase in the school leaving age. The structure of schooling had been simplified. Instead of a dualist structure of academic and non-academic schools which streamed young people to different labour market destinations, there was a trend toward common schools within which the streaming of academic and non-academic students was effected through curriculum provision and assessment. Age grading became more significant. Students moved between primary and secondary schools and between grade levels on the basis of age. Selection on ability only remained important in relation to the allocation of students to different academic and non-academic courses. Even here the importance of IQ was de-emphasised in favour of teacher assessment, counselling, guidance and records. Selection and assessment became more explicitly culturally based. The increased range of curriculum offerings within schools was further extended by an increased use of technology in teaching (films, radio and libraries). The effect of this diversity in curriculum, pedagogy and assessment was to enhance student subject choices. In ten years, over a period of deep depression, the pattern of educational provision in NSW showed some striking discontinuities, but also significant continuities, with the past. How can this pattern of continuity and discontinuity be explained? How can it be conceptualised? Does the shift in educational provision indicate change? Or is it a continuity? If it is change, is it a policy change, for my research does not indicate to what extent the shifts in provision affected lives in schools and classrooms? Is it an educational change, even though the state and the parameters it established for the provision of schooling seem to be the most significant dimensions of change? Is a study of such change case study in policy sociology or a case study in the sociology of education? What, in other words, is the best problematic for the analysis of education and change? As these questions suggest this paper tackles the history and theory of educational change. Its primary purpose is to begin to explain why the shifting pattern of educational provision through the 1930s occurred. In presenting this historical analysis, I offer some comments on the conceptualisation of change in education which I want to extend in more detail in another paper. My aim in discussing these conceptual issues is to address the debate about an appropriate problematic for the analysis of educational change which has been drawn to the fore in recent research in policy sociology (Ozga, 1990; Ball, 1990; Bowe, Ball with Gold, 1992; Dale, 1992a; Lingard, 1993). The current paper is organised in three sections. In the first part of the paper I outline different approaches to theorising educational change and indicate the parameters of my preferred conceptual framework. In the body of the paper, I summarise the findings of a more detailed analysis of education in New South Wales in the 1930s depression (Seddon, 1993) with particular attention to the processes which gave rise to the trends in educational provision outlined above. Finally, I draw conclusions concerning a problematic for research on educational change from the case study.