Many studies have documented the relationship of educational attainment to social and economic success but, until IALS, few studies had sufficient data to allow the empirical exploration of how this relationship depended, in turn, on more fundamental processes such as actual demonstrated skill. IALS has revealed that literacy and education are not synonymous, and that social and economic success depends, in part, on tested skill. The study has also revealed interesting variation in these relationships both within occupations and between countries, facts that fit with common wisdom about the impact of economic and social organisation on markets for skill. The ALL study allows for an extension of this basic analysis to additional skill domains, and where IALS data are available, an understanding of how these relationships are evolving as the knowledge economy develops.

Fourth, the ALL study intended to:

• Identify sub-populations whose performance places them at risk.

Much of the rhetoric employed in the North American debate about skills has been focussed on their impact on the so-called "high performance" workplace. Despite this fact, much of the attention of governments continues to be directed towards those groups of people whose skill levels place them at risk of being socially and economically marginalised. IALS has revealed that, in some economies at least, individuals with poor skills experience significant wage and employment penalties. The design of policies and programmes to attenuate the worst of these impacts and to provide remedial education depend entirely on understanding the number, geographic distribution, and characteristics of the people so affected. The ALL study attempts to profile those whose performance places them at risk.

The ALL also hoped to meet a number of longer-term objectives, including:

• To shed light on the causes and consequences of the observed skill distributions

Longitudinal data is needed in order to truly understand the causes and consequences of any human phenomena, as it is only longitudinal data that allows one to disentangle cohort, life-cycle, and period effects. Thus the ALL, conceived to provide a cross-sectional "snapshot" of the distribution of skill in several domains, cannot be expected to advance our understanding in this regard. When analysed in conjunction with key co-variates provided by the background questionnaire, however, the ALL can be expected to yield tantalising evidence about the relative impact that the various factors might have on the observed distributions of skill. This will be particularly so in those countries where IALS data has already been collected. In this case, changes in the distribution of prose and document literacy skill between the two observations can be related, in a synthetic-longitudinal analysis, to changes in the underlying co-variates. Given the nascent state of skill measurement at the population level, the high cost of longitudinal surveys, the length of time it takes a longitudinal study to yield data, and the horrendous cost of measuring the wrong things on such a study, the ALL can provide critical data to inform the next generation of true longitudinal studies, including the design of Canada's Youth in Transition Longitudinal Survey program.

• To contribute to the literature on the basis of human cognition