3.3  Technical skills 

More skilled workers can improve the output of the firm in various ways, such as undertaking tasks more quickly and with fewer mistakes, performing more complex or responsible tasks, and implementing new technologies. Much of the evidence on the value to firms of skills is from the estimated relationship between an individual’s education and their earnings. However, other cognitive capacities and behavioural traits also seem to be influencing earnings outcomes, such as motivation, trustworthiness, and adaptability. Work experience is also rewarded. Thus, firms seek skills and abilities which are only in part formed within the formal education/training system. Nevertheless, the evidence reported here and later points to a positive relationship between higher skill levels and both earnings and firm productivity.

The starting point for analysing the contribution of skills is to use level of education as a proxy for skill. The strength of this approach is that it allows a focus on the knowledge accumulation process over the working life, including the level of education, work experience and on-the-job training. It also provides a direct link back to major policy instruments – the funding of education and industry training. However, this approach has several limitations. A major concern is over the strength of the signalling effect, where credentials reflect underlying innate ability rather than productivity improvements from further education. Without controls for ability effects, the estimates of the benefits of education will be biased upwards. Wages may also not be a satisfactory proxy for productivity. Reasons for this include wages following training not reflecting productivity gains, wages reflecting labour demand/supply interactions, the payment of efficiency wages and the exclusion of knowledge spillovers.[10]

New Zealand evidence on the returns from education and qualifications have been provided by Sylvia Dixon (1996) and Sholeh Maani (1999). Sylvia Dixon uses Household Economic Survey data for two yearly intervals from 1984 to 1994 to explore changes in the distribution of earnings over time. She finds, when pooling this data, that workers with no qualifications have earnings 16% below those who have school qualifications, while having university qualifications yielded 31% higher earnings. The introduction of industry and occupational dummies reduced the earnings advantage from having a university qualification to 19%, suggesting that some of the variation in earnings is due to job characteristics or unmeasured variables correlated with industry or occupational groups. Sholeh Maani used 1996 Census data to analyse private and public returns to higher education, and compares these with returns in earlier census years. Relative to no school qualifications, the private rates of return from attaining School Certificate or higher qualifications are significant, and rose from 1981 to 1996 with few exceptions. It should be noted that these models do not control for innate ability.

Earning equations can also separate out the effects of work experience. The OECD (1998, p60) provides cross-country estimates of literacy, educational attainment and experience effects. These estimates show that the earnings effects of experience are significant, and that the effects of additional years since formal education have effects of a similar order to years of schooling. Sylvia Dixon (2000) has assessed the effect on returns of reduced work experience for women in contributing to the gender wage gap. It is not clear what mix of skills is being rewarded beyond initial qualifications, such as on-the-job upgrading of technical skills, or acquiring firm specific information, or developing a wider range of capabilities. The gains from work experience may be affected by firm size, with smaller firm having more limited internal labour markets. Similarly, firm turnover could accelerate the obsolescence of firm specific skills, and may encourage the development of generalist skills to manage the risks involved (see Haltiwanger (2002)).

More disaggregated estimations are able to show variations in earnings along a number of dimensions, such as returns in relation to subjects taken (see (New Zealand Treasury 2001a)). Hanushek and Kimko (2000) report that average students scores on maths and science are strongly related to growth. Gemmell (1997) reports that the only group of graduates with positive growth effects are scientists and engineers. Paul Ryan (2001) summarises the empirical evidence on the earnings and employment effects of general education compared with vocational training. Blundell (1999, p7) reports significant returns to vocational training consistently found, especially for higher vocational qualifications. Hyslop, Mare and Timmins (2003) provide further New Zealand evidence on links between changes in qualifications and the patterns of employment and income growth across industries, using Census data from 1986 to 2001.

Bowles et al (2001) note that much of the variation in earnings is unexplained by an individual’s years of schooling, labour market experience, and parental characteristics. Human capital is multi-dimensional, and other behavioural traits appear to be influencing employment and earnings outcomes. They note in passing that employer surveys have ranked characteristics like attitude, motivation and communication more highly than technical skills. More senior management may need to be able to adjust to shocks and opportunities to capture rents, which may be assisted by a degree of self-directedness). Since employment contracts are typically incomplete, and monitoring is costly, employers may seek behaviours that reflect trustworthiness, motivation and low discount rates that reduce the problems of managing such contracts. Seemingly irrelevant personal characteristics (like appearance and tidiness of home) may be valued because they are correlated with behaviours sought by employers such as conformity with appropriate behaviour and self-control. Family backgrounds and the school environment can also be important, as they help to develop capacities like punctuality, dependability, industriousness that both teachers and employers reward.

A related argument has been made that skills affect the responsiveness to technical change, for instance in aiding organisational change. For instance, Aghion and Howitt (2002) have argued that technological change is not skill-biased in the sense that it raises the demand for technical production skills, but rather that it raises the rewards to adaptability. During the introduction of new technologies, unobserved skills like higher ability, reliability, “trainability” appear to be in increased demand. They point to the rise in the within education and age group wage inequality occurring before the rise in between group inequality. Their argument is that this reflects the progressive diffusion of general purpose technologies within industries. Caroli (2001) has argued in a somewhat similar manner that a skilled workforce is crucial for the adoption of high tech equipment and production of high quality goods. She argues, in the context of technological and organisational change, that more general education and training is required that increase adaptability, problem solving learning and inter-personal skills at intermediate levels in the occupational distribution. The implication is that technical change requires a different set of skills, and not just higher skills.

Evidence of the importance of intermediate level skills comes from the detailed case studies undertaken by the UK’s National Institute for Economic and Social Research (NIESR). These mainly compared matched manufacturing plants and service sector firms in the UK and European countries. Keep and Mayhew (2002) summarise this evidence as pointing to large differences in shop floor, supervisory and managerial skill levels that largely account for differences in firm productivity levels and product market strategies (with UK firms producing low value added standardised products). Higher European rates of innovation and technology adoption were attributed to the higher level of intermediate skills held by supervisory staff. The later UK-Germany-US comparisons painted a somewhat different picture. These have qualified the view that superior vocational skills are the main difference in productivity levels, with the US putting more stress on state investment in higher education compared with European employer-funded vocational training. The larger US domestic markets has also allowed economies of scale in manufacturing. Broadberry (2003) and Broadberry and Ghosal (2003) also document the relative productivity performance of Britain, the US and Germany, and assess the role of human capital and organisation in the provision of market services. Blundell et al (1999b) see the NIESR evidence as “suggesting strong links between employment of graduates, including professional scientists and engineers, and the adoption and use of high-level technologies in the firm, and between the extent of investment in worker training and the speed and successful adaptation of new technology.”