The second determinant considered was education. Various pathways through which education could influence entry into marriage were reviewed. The findings suggest that though areas with higher levels of schooling also have a higher marriage age for women, this relationship is not causal. It appears that increases in the marriage age are independent, for the most part, from the increases in schooling. The results from the panel models indicate that the marriage age for women between 1981 and 1991 would have increased even if schooling levels remained constant and did not change. Further analysis discount the possibility of reverse causation between schooling and the marriage age (i.e. increasing marriage age as the cause of increasing schooling levels). It is suggested that marriage norms have changed, and it has become more acceptable and normative for women to marry late.