Note: This paper has been replaced by a more recent version based on a more comprensive data base. Go here for the most recent version
Keywords: Growth & Innovation, Labour Studies
It is not always wise to promise an empirically based paper before the research has been done. When preparing New Zealand’s Post-War Growth Performance: Comparison with the OECD, I observed that the Maddison data base on which the OECD data derived also had some statistics of employment and hours worked, which allowed it to provide some estimates of productivity. New Zealand was not included, but since there was comparable data for New Zealand, I thought, it would be straight forward to include New Zealand in the data base. Hence the promise to produce this paper.
Having carried out the promise, I am somewhat more circumspect. The results reported below are not without merit, but I urge caution in drawing conclusions from them. I am neither sure how robust the Maddison data base is, nor how comparable is the added New Zealand data. For this reason the focus is on the 1998 year. Maddison provides post-war figures only for that year (the last in his series), 1990 (the base year for the data), 1973 and 1950. Estimates of New Zealand output per worker and their relation to other OECD countries for the four years are provided in the last section. The 1998 year data below uses the OECD estimate of New Zealand employment, and Statistics New Zealand estimate of hours worked per week.
The structure of this paper may be summarised in the following truisms
(Output per worker) = (Output per capita) divided by (Employment as a proportion of the population),
(Output per hour worked) = (Output per worker) divided by (Annual hours worked per worker),
where Output is measured as GDP in the same (1990) prices across different countries.
New Zealand in an OECD of 19: 1998
Maddison (2001) has only comprehensive employment OECD countries, but without the Czech Republic, Hungary, Iceland, Luxembourg, Mexico, Poland, Portugal, South Korea, and Turkey (as well as New Zealand). The per capita GDPs for the remaining 18 plus New Zealand for the 1998 year are in Table 1. (They are the same figures that are used in the Comparisons with the OECD paper).
Table 1: GDP per capita
The New Zealand GDP per capita is 68 percent of the average, lower than for the full 29 countries, because it is the poorer countries who have a less satisfactory statistical base.
The Worker Force in the Population
Different countries have different proportions of the population who are employed. The ratio is a complicated summary of the effects of the population age structure (a youthful population tends to have a lower ratio, as does one with a lot of retirees), the labour force participation rate for those in the working ages, and the unemployment rate. Table 2 reports the ratio of the employed to the population for the 19 countries.
Table 2: Ratio of Employed Population to Total Population: 1998
There is considerable variation of the proportion of the population who work. New Zealand is just below the median and 98 percent of the mean.
Ouput per Worker
Combining these two tables gives Table 3, GDP per worker, which is a measure of productivity.
Table 3: GDP per Employed Person: 1998
New Zealand is at the bottom of the rankings at 69 percent of the mean. Spain has leapt up to 12th, because of the low proportion of the population who are classified as employed.
Hours worked per Worker
Table 4 adds to Maddison’s data an estimate of hours worked per New Zealand worker. Hours work per week is derived from the quarterly employment survey. This is multiplied by 46 weeks of work per year, which is calculated by allowing for the 12 days of public holidays and the 18.6 days annual leave that applies to New Zealanders. We do not know is whether the Maddison data is calculated on a comparable basis.
Table 4: Annual Hours Worked per Worker:1998
New Zealand proves to be the second to highest of the OECD19 in terms of hours annual worked per worker, some 10.5 percent above the average. That is about 170 hours, the equivalent of almost 4.5 days a year.
Ouput per Hour Work
Combining the last two tables gives Table 5 where productivity is measured as output per labour hour.
Table 5: GDP per Hour Worked: 1998
New Zealand is at the bottom of the table at 63 percent of the average.
Whether this represents the actual situation I am unable to say. Given the difficulties of calculating employment and hours worked for New Zealand, one wonders how reliable the data for the other countries are.
Output per Worker Through Time
There is no standard employment series for New Zealand for the post-war era. I was able to cobble together one based on the Ppoulation Census, and interpolated using the employment Survey. Table 6 gives the ratio of New Zealand’s GDP per employed relative to the average of the OECD 19 countries, for Maddison’s four benchmark years.
Table 6: NZ GDP per Worker as Average of OECD 19
Basically the pattern is the same as for GDP with a secular decline throughout the post-war era. New Zealand is above average in 1950, ahead of many war ravaged countries, about average in the 1970s, and declines after. The decline in the 1990s (in contrast to a GDP relativity which was fairly constant) reflects the poor growth in labour productivity over the period offset by strong employment growth.
I have emphasised that I am not as confident of the robustness of the labour force data as the GDP data (and I have many concerns about that). If there is a conclusion to all this, it is that labour productivity patterns are much as what one might have expected from the relative GDP, and it is not obvious that labour market has been a particular problem to New Zealand’s poor post-war performance other than, of course, that arithmetically the poor growth performance is associated with a poor labour productivity performance.
1. B.H. Easton (2002) New Zealand’s Post-War Economic Growth Performance: Comparison with the OECD (Economic And Social Trust On New Zealand).
2. A. Maddison (2001) The World Economy: A Millennial Perspective (Development Centre Studies, OECD)
3. Maddison (2001) does provide some employment data, but not hours worked for Czech Republic, Hungary and Poland.
4. B.H. Easton (1998) The Impact Cost of Increasing Statutory Holiday Entitlements (discussion paper for NZ Engineering Printing and Manufacturing Union).
5. If the Czech Republic, Hungary and Poland are added the relativities with the OECD 22 are (1950:140.7; 1973: 101.2; 1990: 76.0; 1998: 71.8)