3.3 The hours and productivity of the new employees
To calculate the extra GDP generated by the new employees, it is necessary to make assumptions about the number of hours these new employees work, and their output per hour (ie their productivity). The simplest assumption, and one that is often used, is that new employees have the same average hours and productivity as existing workers.
It would be preferable to avoid this assumption. Survey evidence shows that people who are not currently employed differ systematically from people who are currently employed, in ways that affect hours and productivity. An analysis of data from New Zealand’s Household Economic Survey shows, for instance, that people who are not currently employed are more likely to have young children, and less likely to have an advanced education, than people who are employed (Kalb and Scutella 2003: Table 1).
The only information on the new employees that is contained in our scenarios is the employees’ age and sex. However, even this limited information allows us to make some allowance for differences in hours worked and productivity between new and existing employees.
Table 5 shows estimates of hours worked and wages for the different age groups in New Zealand in the period 1991-2001. These estimates were calculated by Guyonne Kalb and Rosanna Scuttella, based on data from Household Economic Surveys (Kalb and Scutella 2003). As well as calculating hours and wages for people actually in employment, Kalb and Scuttella used data on the background characteristics of people not in employment to predict the wages that they would have earned if they had, in fact, been employed. The background characteristics included such things as sex, age, number of children, marital status, education, and ethnicity.
| Age group | Hours worked by people in employment | Hourly wages for people in employment | Imputed hourly wages for people not in employment | |||
|---|---|---|---|---|---|---|
| Males | Females | Males | Females | Males | Females | |
| 15 to 19 | 28.4 | 21.2 | $9.35 | $9.79 | $7.86 | $8.76 |
| 20 to 24 | 41.3 | 34.6 | $13.43 | $12.31 | $9.51 | $10.64 |
| 25 to 34 | 46.2 | 35.6 | $17.27 | $15.69 | $11.87 | $12.05 |
| 35 to 44 | 47.5 | 33.0 | $20.00 | $16.46 | $13.17 | $12.38 |
| 45 to 54 | 47.7 | 35.6 | $20.55 | $15.01 | $13.08 | $11.76 |
| 55 to 59 | 46.0 | 33.5 | $18.61 | $15.55 | $12.80 | $11.28 |
| 60 to 64 | 42.5 | 30.4 | $16.91 | $16.69 | $12.06 | $10.79 |
| 65+ | 32.6 | 23.7 | $16.91* | $16.69* | $12.06* | $10.79* |
*In the absence of data on the wages of people aged 65 and over, we assume that they are equal to the wages of people aged 60-64.
Source – Unpublished tabulations provided by Guyonne Kalb, based on the study described in Kalb and Scutella (2003)
We adopt the standard assumption that output is proportional to wages.[3] Given data on employment numbers, population, and GDP in 2001, it is then possible to calculate GDP per hour worked for people in employment, and imputed GDP per hour worked for people not in employment. Appendix 2 describes the calculations in detail, and Table 6 shows the results.
| Age group | Existing employees | New employees | ||
|---|---|---|---|---|
| Males | Females | Males | Females | |
| 15 to 19 | $17.68 | $18.51 | $14.86 | $16.56 |
| 20 to 24 | $25.39 | $23.27 | $17.98 | $20.12 |
| 25 to 34 | $32.65 | $29.66 | $22.44 | $22.78 |
| 35 to 44 | $37.81 | $31.12 | $24.90 | $23.41 |
| 45 to 54 | $38.85 | $28.38 | $24.73 | $22.23 |
| 55 to 59 | $35.19 | $29.40 | $24.20 | $21.33 |
| 60 to 64 | $31.97 | $31.56 | $22.80 | $20.40 |
| 65+ | $31.97 | $31.56 | $22.80 | $20.40 |
3.4 The extra GDP generated by the new employees
Table 7 shows the extra GDP generated by the new employees under the two scenarios. The results for each age group and sex are calculated by multiplying together the number of new employees of that age and sex, as shown in Table 4; the number of hours worked per week by people of that age and sex, as shown in Table 5; the number of weeks in a year (ie 52), and the imputed GDP per hour of new employees of that age and sex, as shown in Table 6.
| Age group | “Young women” scenario | “Overall” scenario | ||
|---|---|---|---|---|
| Males $m | Females $m | Males $m | Females $m | |
| 15 to 19 | 0 | 0 | 134 | 107 |
| 20 to 24 | 0 | 0 | 237 | 213 |
| 25 to 34 | 0 | 1,215 | 669 | 567 |
| 35 to 44 | 0 | 0 | 863 | 599 |
| 45 to 54 | 0 | 0 | 752 | 514 |
| 55 to 59 | 0 | 0 | 263 | 172 |
| 60 to 64 | 0 | 0 | 192 | 131 |
| 65+ | 0 | 0 | 381 | 306 |
| Total | 1,215 | 6,101 | ||
Under the “young women” scenario, the new employees generate an additional $1,215 million of GDP; under the “overall” scenario, they generate an additional $6,101 million. Actual GDP in 2001 was $120,509 million (Statistics New Zealand series SNCQ.S1NB15). This means that GDP is 1.0% higher than it would otherwise have been in the “young women” scenario and 5.1% higher in the “overall” scenario.
It should be noted that the changes offer a one-off increase in economic activity, rather than an increase in the growth rate. In addition, while increased participation leads to a higher level of economic activity, it might not necessarily increase economic welfare, since participation has opportunity costs.
Each new worker needs a concomitant quantity of capital in order to maintain a constant level of productivity. If the current stock of capital was underutilised, capital utilisation could be increased. Alternatively, if the current stock of capital was fully utilised, new investment would be needed in order to maintain the capital: labour ratio. To maintain a constant ratio of capital to hours worked, the capital stock would need to rise by the same proportion as the increase in participation.
Notes
- [3]Whilst it seems likely that wage differentials between age-sex groups reflect more than differentials in marginal products alone, wages still seem to provide the best indicator of relative marginal productivities. Hellerstein, Neumark and Troske, (1999) conclude for the United States that the higher pay of prime-aged and to a lesser extent older workers is accompanied by higher estimates of marginal products whilst the lower wages of women were not accompanied by lower marginal products. However, Crépon, Deniau and Pérez-Duarte (2002) find essentially the opposite for France.
