Vertical occupational mobility and its measurement

This paper describes a number of alternative approaches to devising a vertical occupational scale and compares the outcomes of different scales on calculations of occupational mobility. The paper describes the conceptual issues relevant to calculating occupational mobility and documents the measurement error embedded in the choice of measure, as applied to different data sets. The ranking schemes used include SOC (9) major codes ranked by mean occupational hourly earnings, Hope-Goldthorpe collapsed 36-point scores, a 15-category SOC ranking based on educational qualifications, and a 77 category ranking based on 2-digit SOC90 occupations, wage rates, educational qualifications, training and job tenure. These ranking schemes are applied to data from the 1958 NCDS cohort between the ages of 23 to 33 and 33 to 42, and to 1.25 year transitions in the Quarterly Labour Force Survey panel data. The calculations carried out show that variations in the extent of vertical occupational mobility, both upward and downward, had systematic elements. The extent of mobility was found to vary by the composition of the individuals´ data particularly in terms of lifecourse stages and gender, the number of categories in the ranking scheme, attrition in the data and flows out of employment over the mobility period, and changes in labour market conditions over time. However, the sizes of these effects were very variable.