RDP 2000-08: Nominal Wage Rigidity in Australia 7. Are Broader Measures of Remuneration More Flexible?

While changes in wages have displayed features of downward rigidity, it is possible that firms are able to vary other forms of remuneration – which may be less important or visible to workers than base pay – to achieve desired adjustments in total labour costs. Indeed, there is some evidence to suggest that effects of nominal wage rigidity are at least partly overcome in this way (see, for example, Lebow et al (1999)).

A broad measure of remuneration is available over our full sample period and we call it total pay.^[29] It does, however, exclude bonuses, commissions and incentive schemes, as these have only been collected in recent years and not all respondents reveal full details. Furthermore, they can be difficult to value (especially when they include stock-based compensation).

As shown in Figure 4, for the entire sample, the distribution of total pay is wider than for wages. Of these observations, around 8 per cent are negative (more than twice the proportion for wages) and the spike at zero has been reduced to 10.5 per cent (from 15 per cent). However, while the distribution is slightly more dispersed than that for wages, it is still quite positively skewed. Furthermore, we find that the degree of skewness is also negatively correlated with inflation, suggesting that total pay is also rigid downwards, just to a lesser extent than wages.

Would a greater reduction in rigidity be obtained if our measure of remuneration included explicit performance-based earnings? We try and establish whether variations in explicit performance-based pay omitted from our data set would overturn our finding that broad measures of earnings also display some downward rigidity.

Given that performance payments are not available from MCED on a consistent basis through time, we focus on one recent period. For 1998–99 only, we identify annual changes in two broad measures of earnings: our existing measure and one that also includes actual annual bonuses, commissions and incentive schemes. The distribution of annual changes is slightly more dispersed for the measure that includes bonuses, but the difference is trivial (see Appendix B). Even though around 60 per cent of the employees in our sample receive some form of performance payment, the share of employees who receive highly variable performance payments is too small to generate a distribution that looks substantially different to that we have presented for total pay.^[30] Consequently, our result stays intact: broader measures of earnings display downward rigidity, but to a lesser extent than wages.

Should we be surprised that broader measures of earnings also display skewness away from pay cuts? A broader measure of earnings will only be more inherently flexible if the non-wage components represent ‘pay at risk’ that may either rise or fall in accordance with employees' performance. Bonuses are a good example of pay at risk. However, many reward systems are asymmetric; rewards are given for good performance but penalties are not applied for bad performance. Consequently, the level of earnings may increase following a reward for performance, but remain unchanged in the subsequent year when performance is not rewarded. These asymmetric types of reward systems contribute to downward nominal rigidity even for broad measures of earnings and would appear to play a role in explaining our results.

The finding that broad measures of earnings also display some downward rigidity is important. It implies that non-wage forms of remuneration, at least in their current form, do relatively little to counter the downward rigidity observed in wages. Furthermore, it is contrary to the findings of overseas studies which emphasise the reduction in rigidity imparted by non-wage forms of remuneration.^[31]

Footnotes

It includes all allowances, superannuation, loan benefits, company cars and the costs incurred by the employer through fringe benefits tax, as described in Appendix A. [29]

The Statistician also identifies ‘payment by measured result’, but this makes little contribution to either the level or growth in various measures of aggregate earnings, consistent with the results presented in Appendix B. [30]

It is, therefore, tempting to conclude that nominal wages in Australia may be more rigid downwards than those in other countries, particularly the US. However, such international comparisons are fraught with difficulties given the differences in the measurement of nominal wages. The apparent flexibility in wages or non-wage remuneration in other studies may stem from measurement problems, some of which are avoided by the special features of the MCED Survey. [31]