Paper for the Ministry of Economic Development Seminar Series: 25 February, 2004.
Keywords: Growth & Innovation
Notes: This is a long paper, and is in two parts. There is a short version (of about a third the length). Despite its length, many of the arguments have had to be abbreviated. Some guidance is given in the text of web references where there is greater detail. More material will be found in the Index of New Zealand’s Economic Performance. The foundation source is the book “In Stormy Seas”“In Stormy Seas”
1. The Political Economy of New Zealand’s Economic Development
2. Changing Sectors
3. The Course of GDP
4. The Long Run: 1861-2003
5. The Post-war Era
6. The 1966 External Shock and After
7. Explanations for the Slow New Zealand per capita GDP Growth
8. Non-Explanations for the Slow New Zealand per capita GDP Growth
9. Some Errors of Method
10. What Happened After 1984? Why the Great Post-War Stagnation?
11. The Importance of Thinking Sectorally
12. The Next Political Economy?
13. Conclusion
Appendix: Rankings and Relativities
Appendix II: Recent Developments in the Terms of Trade
This year, 2004, is the thirtieth anniversary of when I first identified an anomaly in the behaviour of the New Zealand economy, which led to the research program in economic growth and development which I have worked on ever since. to I am grateful for the invitation from the Ministry of Economic Development for the opportunity to present an overview of that program, although inevitably space means that much of the detail is omitted. It can be found in my In Stormy Seas, and other research papers.
This work has been carried out in a context of what between them, George Santayana and Karl Marx famously said. ‘Those who do not learn the lessons of history, are doomed to repeat them: the first time is tragedy, the second time it is farce.’ For much of the New Zealand economic debate is has been woefully a-historical, with little reference to our economic history.
Mindful that the invitation came from a ministry for development, and not just for growth, I will begin with a political economy account of the past, which emphasise that economic change is not just about increases in material output, but a variety of other changes including the mix of sectoral outputs, the products consumed, the production technologies used, the way the economy and society is organised, the way people live. After the political economic account I will describe the main outlines of aggregate economic output through time. Then, focussing on recent times, I shall look at New Zealand’s aggregate performance compared to other countries, and finish with a quick summary of my own explanatory account of what happened and some critical remarks of the inadequacies of some of the other accounts, together with an indication of the policy implications.
1. The Political Economy of New Zealand’s Economic Development
(For an earlier but more elaborate version of this section see “Towards a Political Economy of New Zealand: The Tectonics of History”.)
Political economy can be described through the metaphor of movements of the earth’s crust. The geologists’ tectonic plates are great slabs of rock which shift about – pushing, crushing, and overriding one another. In a similar manner the economist’s tectonic plates are systems of organisation, which are in conflict and over time change as new ideas and circumstances create new ways of organising the economy, while old organisations disappear subducted by the overriding new. Just as in geology the clash of the plates generates earth movements which modify the land on which we live, the conflict between the political economy plates also leads to political and social change. The earthquakes we record, in geology – or in politics and sociology – are the visible outcomes of the long term movements of the plates.
The first such plate in New Zealand – the beginnings of an economy – began about 750 years ago when the first Polynesians reached these shores. They came from a very different tropical environment, to one rich in protein food sources from birds and the sea. Unfamiliar with the new environment and with inappropriate organisational forms, they exploited the available resources in unsustainable ways. Birds became extinct (most spectacularly the Moa), seal colonies in the north became exhausted, and many fish and shell fish declined in numbers.
The term for an unsustainable political economy based upon exhausting the resources is ‘quarry’. In the depleted environment, any surviving communities have to develop a new political economy – a new tectonic plate has to arise.. We know little about the transition but the outcome was a more sustainable economy and society which relied on kumara and fern roots. That involved the development of new technologies, kumara pit storage over winter and more efficient gardening rather than gathering. It also seems to have resulted in the development of stable property rights, enforced by military control centred on the pa. This led to a new political economy – the ‘Classic Maori’ – because the technology changed the way society was organised.
About 500 years ago Maori longevity was similar to that in Western Europe. Given food clothing and shelter was the substantial core activity of such economies, we might cautiously conclude that even though the Maori had no metals, their standard of living was then probably similar to the Europeans, although with less inequality. Moreover the population was slowly expanding. There is much we do not know about the Classic Maori political economy – the quantitative material is very deficient – but it had one fundamental difference from all those that came after, and the Polynesian quarry before. It was a closed economy without interaction with the rest of the world.
This changed just over 200 years ago with first explorers and then the sealers and whalers. Just as those early Polynesians did not understand the environment they had come to, neither did the early Europeans. They quarried natural resources too: whales, seal, timber, kauri gum, gold, other minerals, even soil was washed to the sea. We might also think of the European quarry treating the Classic Maori in just as an exploitive way, and war and land speculation are not sustainable either. So the first European political economy in New Zealand was what the French described as a ‘colony of exploitation’ rather than a ‘colony of permanence’. It is a world in which the visitor comes, exploits, and moves on, leaving behind debris and ruin.
Histories of New Zealand tend to ignore the quarry phase, even though it continued in some regions – notably the West Coast – until recently, and the Taranaki is a new quarry province based on the hydrocarbon reserves. Conventional history’s vision is a permanent settlement from the beginning. In practice the early settlements were primarily suppliers to quarries, and without them would have been even more impoverished. While the settlers demanded produce from the rest of the world, they had little to offer other than the depleting natural resources. The exception was wool. It is not inconceivable that New Zealand could have ended up as the Falkland Islands of the South Pacific.
But from 1882 new technologies transformed New Zealand: refrigeration, the steamer and telegraph came from offshore, while the grasslands revolution was largely indigenous. Over the next eighty years the political economy based on producing grass, processing it into wool, meat, and dairy products, and selling them overseas in return for the desired imports. There is much to be told of this story, especially in the way the new economy impacted on social and political organisation. But there is space for only one example. Women are useful in the quarry for their sexual services. They are necessary in the sustainable settlement because it needs children to survive. So we see a rise of the importance of women in the social and political life of late nineteenth century.
By now the Maori political economies and the quarries had both adopted much of the technology and even the organisation of the European tectonic plates, but had been marginalised, not least because they had lost so much land, although their vital role in provedoring the quarry should not be forgotten. Now the pastoral economy dominated New Zealand from the 1880s to the 1960s.
However, sometime in the inter-war period there evolved a section of the economy which was based on import substituting industrialisation. I am not sure whether this was an entirely new political economy – certainly there were intense political clashes between the two – or whether it evolved out of the prosperity of the pastoral economy. By the 1990s it had largely ended, although it was to leave a successor in a export oriented industrial base.
The pastoral dominance had ended too. In 1966 the premium prices that farmers got for wool collapsed, never to return (except temporarily in the 1972-3 commodity boom), while meat and dairy prices were under pressure. The response was diversification – into horticulture, timber, fish, some minerals, tourism, and a little general manufacturing mainly to Australia.
Again the new political economy, which was based on the sustainable exploitation of primary resources, led to changes in the way New Zealand was governed and how New Zealanders lived. Again the story could be illustrated in many ways, but time allows only the example of the more market element of the 1984 economic reforms because the greater diversity of the export sector meant decentralisation in the economic mechanism became necessary.
We are now so close to our own times that it is difficult to see the great movement of the tectonic plates of the political economy. At the end I shall suggest that there may be a new plate arising: that appears to be the intention of the Government’s Growth and Innovation Strategy.
The political economy of tectonic plates is a qualitative story, which reminds us that development is not simply about a single aggregate output. There are a few quantitative indicators which support this aggregate story. Some come up later – when we review the great diversification of the 1970s – but a couple of longer term ones can be inserted here, although they dont cover the entirety on New Zealand’s economic history.
Industry Composition
Table 1: Industry (percentage) Shares in Nominal GDP
| YEM | 20 | 30 | 39 | 53 | 60 | 70 | 80 | 90 | 99 |
| AGR | 29.8 | 26.2 | 23.2 | 22.1 | 18.0 | 11.7 | 10.1 | 6.1 | 5.2 |
| OPI | 2.9 | 3.9 | 4.3 | 4.3 | 5.1 | 7.1 | 6.8 | ||
| MAN | 21.6 | 23.7 | 21.7 | 21.1 | 21.8 | 22.5 | 23.3 | 19.2 | 16.6 |
| CON | 4.0 | 6.6 | 8.0 | 7.1 | 7.2 | 5.7 | 4.6 | 4.2 | 3.9 |
| WRT | 15.2 | 16.4 | 18.7 | 20.7 | 20.0 | 17.7 | 18.3 | ||
| T&C | 5.8 | 8.5 | 7.4 | 8.0 | 7.9 | 7.6 | 7.1 | ||
| FBS | 7.7 | 7.3 | 8.2 | 9.1 | 9.6 | 14.2 | 16.3 | ||
| OS | 16.0 | 13.6 | 14.4 | 18.0 | 19.4 | 23.4 | 25.7 |
Notes:
The data is from a variety of sources, and involves some issues of changed definitions over time.
MYE = March year ended
AGR = Agriculture
OPI = Other primary sectors (including electricity, water and gas)
MAN = Manufacturing
CON = Construction
WRT = Wholesale and retail trade, restaurants and hotels
T&C = Transport and communications
FBS = Financial and business services
OS = Other services
Sources: Table 9.1, page 140, In Stormy Seas
Table 1 shows the sectoral composition (by value) of GDP for about as far back as we can go.
There have been major changes to the structure of GDP, particularly a substantial reduction of the share of agriculture in GDP over the 80 years (and which today is exceeeded by the share of other primary industries), a diminution of the manufacturing sector for about 20 years, with the service sector expanding but not uniformly. There are complex stories hidden within these sectors. For instance, the increasing share of the finance and business sector in the economy partly reflects outsourcing, but it also is in part of its poor productivity record so its prices rise faster than average. Conversely, the IT part of transport and communication has expanded rapidly but with reductions in prices so the sector is relatively smaller in nominal terms. In summary, development involves changes in the composition of GDP, which are lost in the focus on the growth of the aggregate.
The lesson of history is that development involves changes in the composition of GDP, which are lost in the focus on the growth of the aggregate.
Deflators
Changing industry composition, means changing relative prices. This is nicely illustrated by comparing the Consumer Price Index (CPI) with the GDP price deflator (GDEF), an exercise I first did to evaluate the usefulness of the CPI as a proxy for GDEF. (I concluded it was a poor one.)
While there are technical difficulties in using the official GDEF or CPI indexes over long periods, Chart 1 shows their ratio over the entire period back to 1954/55 when the GDEF is available. There is hardly any trend, so the long run growth of the CPI and GDEF were broadly the same. But there are major short term swings around this flat trend.
Consider the upswing in the early 1970s, when the ratio rose 10 percent in three years. Had the Statistics New Zealand used the CPI to deflate nominal GDP, it would have miscalculated volume GDP growth by over 3 percent p.a. over the short period. This is the most spectacular example, but in over half of the years for which we have data, the divergence in the change between the CPI and GDEF was 1 percentage point, in over a quarter it is more than 2 percentage points. (The standard deviation of the divergence is 2.0 percentage points.) Using consumer prices (the CPI )as an indicator of production prices (the GDEF) gives a misleading account of the business cycle.
While at this stage one could raise some macro-economic policy issues, particularly the relevance of the CPI in the Policy Targets Agreement, there is also an important growth versus development dimension here. The divergence occurs because the CPI and the GDEF are covering quite different groups of products. The CPI covers what consumers spend, including that produced in New Zealand, and imports of consumer goods. GDEF covers only what is produced in New Zealand, including for export and what goes into investment as well as consumption. But it excludes imports. GDEF is about the prices which New Zealand producers influence, which are not the same as the prices which New Zealand consumers face. The terms of trade and nominal exchange rate changes, as well as productivity differences between sectors reflecting in price differences contribute to the divergence.
The difference is a salutary reminder that aggregates operate on the basis that there is only a single product. Two key prices diverging so markedly arises because an economy is about many products. An economist concerned with the growth of the economy cannot just look at aggregate GDP. Sectors are important; prices are important; and profitability and other factor prices are important.
I turn now from a political economy account of New Zealand to focusing on the aggregate economic growth as measures by GDP. I am not going to defend GDP as a measure of welfare – I have probably written more than any other New Zealand economist of the difficulties it involves. See Index on Wellbeing and Materialism for some of this writing. The focus of it here is that it is a measure of the aggregate output of the monetary economy.
There are no direct estimates of GDP before the late 1910s. There are synthetic estimates which use a money multiplier on the stock of money to estimate GDP. The first was by Gary Hawke, but his figures did not meet the common sense test of conforming to what else we know about the economy. In particular there was a long depression in the 1880s, which is not apparent in his figures.
1859-1939: The Rankin Series
Keith Rankin pointed out that the money multipliers should be adjusted for the business cycle. I have serious doubts about the synthetic method, but as Bob Solow once quoted an inveterate gambler by ‘I know it is crooked, but it is the only casino in town.’ Chart 1 shows per capita GNP at 1910/11 prices, which largely ignores the Maori economy. (It is smoothed by a two year moving average.)
This series meets the common sense test, although Rankin may have used a bit of common sense choosing his multipliers. We see a sharp rise in the early 1860s with the gold finds and than the running down as the easy gold was exhausted. The boom in the 1870s reflects the Vogel borrowing, which came to an end in 1878 with the collapse of the Bank of Glasgow, in London, followed by the long depression of the 1880s. (To return to a theme, initially the northern half of the North Island boomed while the South stagnated – a divergence lost in the aggregation.) There is little overall per capita economic growth in the period. What seems to have happened is that population flowed in, (there was a big increase in this period) attracted by the relativity good prospects New Zealand then offered, while any improvements in productivity were offset by the depletion of the natural resources.
Sometime in the mid 1890s, despite the running down of the quarry, the economy began to expand quickly as the new pastoral economy accelerated and export prices rose relative to import prices (in part as a result from lower shipping costs). Rankin thinks GNP per capita may have increased by over 40 percent in a dozen years, a per capita growth rate of 3 percent per annum.
The expansion came to an end in the late 1900s – perhaps signalled by the 1908 Blackball strike when mine-owners began to reduce working conditions as their markets stagnated. The stagnation seems to have lasted two decades to 1929, followed by the trough of the Great Depression, and then we get a sharp upswing in the late 1930s.
Over the eighty years of Rankin’s series, volume GNP per person rose almost 90 percent, or .8 percent p.a. However we must be cautious. The synthetic method can easily tilt the trend slightly, so the average is subject to a large margin of error. In any case given that lifestyles and consumption patterns were so different between 1859 and 1939 one wonders whether any comparison is meaningful.
1917/8-1938/39: The Lineham Series
Brent Lineham directly estimated GDP from 1917/18 to 1938/39 by aggregating the value added of each sector. The result is nominal a GDP series, which has a sounder basis than the Rankin series, albeit it covers only a quarter of his period. Rankin notes that his and Lineham series diverge, which is a bit disappointing, but that emphasises the dangers of the synthetic series.
To derive a volume GDP series from the Lineham series. I constructed a GDEF up to 1954/55 by weighting the available price indexes. (This was almost twenty-five years ago, and with hindsight I realise that the fixed weights over a forty year period may have been unwise, and I have even thought how to deal with this in the most important area of the external economy.)
The picture is broadly that of the Rankin series, albeit with more precision. There was some expansion following the First World War. as soldiers came back to work, but the 1920s were a period of near stagnation, followed by a slump in the early 1930s. and a strong upswing thereafter. We shall see more of this upswing in the next series. Note how the projection of the trend of the 1920s suggests the economy recovered from the Great Depression by as early as 1936, although an alternative interpretation is that the whole of the 1920s were also a period of ‘depression’, and the high growth we see after 1933 reflects the recovery from the 1920s as well as the early 1930s. In some of my writing I have argued the ‘Interwar Depression’ thesis, distinguishing that experience from the ‘Great Depression’ which is confined to the early 1930s.
Surprisingly, the fall in per capita output during the great depression seems only to be about 13 percent over two years (to about the level the economy was at the end of the war), and there is hardly any fall in the earlier, shorter, but perhaps as harsh one in the beginning of the 1920s. Rankin thinks that Lineham may have got the downturn of the early 1930s wrong, because he is combining data from different year endings, but there is also the problem of interpreting the concepts undelying the series.
The Lineham series is showing is the pattern of the volume of output, not the pattern of real income, the ability to purchase goods and services. The terms of trade – the price of exports relative to the price of imports – fell substantially during the Great Depression. A fall in the terms of trade would have reduced spending power even had there been no reduction in output, because the exports would have bought less imports. That reduction in income also reduces spending and the purchase of domestically produced goods and services, and that is what we see in the Lineham series.
This distinction between output and spending power is measured by today’s National Accounts statisticians in their measures of real GDP and RGDI (Real Gross Disposable Income). It is a distinction often crucial for understanding the New Zealand economy, because it experienced bigger swings in its terms of trade than any other OECD country I have looked at.
Cautiously, in order to give an idea of the effect, I adjusted the Lineham series for the changes in value of exports from the terms of trade to give a rough RGDI series. That suggests there was an 18 percent contraction in RGDI in the two years Additionally there was a contraction of credit and the ability to borrow, together with the general economic dislocation that price and volume changes generate, plus major reduction in the market value of wealth.
The (Almost) Official Series: 1932-1960
There is sufficient bits of official series to construct a nominal GDP series from 1931/32, at the bottom of the Great Depression. Again we have to deflate it by the GDP deflator I constructed. Chart 4 shows the result to 1960, which gives almost a decade overlap with the Lineham series at the beginning, and a decade lap with the next (official) series at the end.
Again we see a high growth rate in the 1930s and through to the early 1940s, long after the recovery from the Great Depression. The peak is in 1943/44, following a decade of 6.5 percent per capita annual growth, comparable to that of the East Asian economies in the 1980s and 1990s, or the Irish in the 1990s. However that includes the extra effort of a war economy and it probably makes more sense to think of the per capita growth rate from the mid 1930s to the late 1940s of between 5 and 6 percent p.a. There is post-war stagnation, as presumably things got back to normal (soldiers got back and replaced women who had been in the paid workforce), a recession in 1947/48 and 1948/49 which may have cost the Labour Government the 1949 election. Even so, in its 14 year reign, GDP per capita rose 54 percent. (For comparison, the Rankin series suggests the previous increase of that size took over 85 years, and the next series suggests it took 30 years after 1950 for per capita production to rise 54 percent.) It was a period too, when New Zealand seems to have grown faster than Britain and Australia and probably the US. Comparisons with war ravaged countries are hardly appropriate, but the reverse will happen after the war, when the ravaged grow faster than those that were not invaded.
Thus the late 1930s and 1940s were the best sustained economic growth rate in New Zealand’s history. It was not due to the recovery from The Great Depression, which was over by 1936. There has not been a lot of work on why there was the success. One factor must have been the application of underutilised capacity that existed in the 1920s, but external conditions were favourable, there was major technological change in the pastoral sector from grass growth, and as discussed earlier, perhaps import substituting industrialisation was important. Interestingly, the high degree of government intervention during the period does not seem to have handicapped growth.
The Official Series (1949/50-2002/3)
We are now at the stage where we can use official series, although we are splicing together a series of differently constructed measures. To give but one indication, after 1991 the population estimate allowed for the census undercount, so I have had to increase the population before 1991 by that measure. The way volume GDP has been measured over the years has varied also. (On the principle of getting as long a series as possible, I have added some Treasury published – in the 1956 Economic Survey – estimates of the GDP volumes before 1954/5, although Statistics New Zealand has never owned them.)
Chart 5 covers only 50 years in contrast to the Rankin series’ 80 year coverage. It is harder to interpret, because it shows a much stronger trend. GNP per capita growth averaged .85 percent per capita, in the 1859 to 1939 period, whereas the GDP per capita growth in the 1950 to 2003 period averaged 1.55 percent p.a. (The difference between GNP and GDP growth may not matter much in this context.) Moreover the earlier period shows considerable stagnation with a couple of growth booms, whereas more recently we have experienced increases in most years with six, usually short, periods of setbacks. I use these breaks to tell the story of post-war growth.
Except for the Korean War Wool boom, the immediate post-war era seems largely to have been a period of stagnation. However from the mid 1950s New Zealand went into a period of strong GDP growth of just over 2.2 percent p.a. in per capita term.
This comes to an end in 1966 when the prices of wool fell. There is an immediate downturn with per capita growth reduced to 1.4 percent p.a.. This is a controversial period, because it is the context for the debate over the policies of the mid 1980s, and it is complicated by the 1971/1972 international commodity price boom, and a clear measurement error between 1976/7 and 1977/8 (for which I have adjusted). I’ll come back to the period shortly, but all the evidence points to the slowdown being from New Zealand adapting to the lower price of wool (it fell relative to import prices by 40 percent), remembering that not only did wool make up almost a third of exports in the early 1960s.
The resulting external diversification had largely worked its way through by the mid 1970s, and the economy went onto a higher per capita growth path of about 1.4 percent, until 1985, a growth rate not too different from the rest of the OECD. It’s an erratic path – befitting the governance of Muldoon.
As the graph shows, the seven good years were followed by seven lean years of stagnation to 1992 under the regimes which we know as Rogernomics and Ruthanasia. The downturn at its end was probably due to the fiscal measures of late 1990 and 1991 which contracted the economy.
The new upswing begins in 1992/3. Just how rapid it has been depends how much one adjusts for the Ruthanasia recession, but I reckon trend per capita growth rate has been about 2.2 percent p.a. There is a view, which I would not rule out, that the growth rate has been faster in recent years, although that is a bit dependent upon how one treats the Asian slump of 1998.
The data on which this chart is based is available at “A Long Run GDP Series”
I have cobbled together the various GDP series, to give a 142 year run from March year 1861 to 2003, always using the better quality data. Chart 6, which uses a logarithmic or ratio scale, shows the stagnations in GDP per capita in the nineteenth century, and from the around 1908 to 1935, in the late 1940s to the early 1950s, and in the late 1980s and early 1990s. Noting a logarithmic scale graph, steeper means faster, we observe that there were rapid expansions in the 1890s and early 1900s, and the rapid growth from 1935 to 1945, plus a steady growth, with the odd hiccough from the 1950s to the early 1980s. In summary the last hundred years have seen an average growth of per capita GDP of about 1.6 percent p.a., a doubling of output per person every 44 years.
Chart 6 also shows a trend line based upon a fourth order polynomial. It recognises the nineteenth century stagnation, but sees a strong upward trend in the twentieth. However notice that the trend bends down late in the twentieth century. (For the record, the point of inflexion is 1938.) It may reflect the stagnation of the following years, an interpretation supported by that GDP levels have been above trend in the past few years. Alternately it may indicate a slowing of the long run growth rate for New Zealand.
We obtain an insight into what happened by from Chart 7 of NZ GDP from 1954/5. (Note this is not a per capita measure. For more about the population see my In Stormy Seas.)
Chart 7 shows the path of New Zealand volume GDP from March year 1955, where it is indexed to 1000. Over this New Zealand GDP path is superimposed three OECD GDP paths. The first, on the left of the chart, is set so that OECD GDP at the same 1000 in the March 1955 year. (I mention this is for the entire 29 OECD countries, and so it is a little – but not significantly different – from my earlier work, which used the fewer countries which were OECD members at the time.) The middle path has the OECD GDP set at 820 in the March 1955 year, that is 18 percent lower than the first OECD path. The third path, on the right, has the OECD GDP set at 730 in the March 1955 year, or 11 percent lower than the middle path.
So the slowing down we saw in that long term trend was not continuous, but due to a couple of periods when shocks – which I discuss below – lowered the level of GDP relative to the OECD, rather like dropping a step or two on the ladder. Indeed in two thirds of the years – perhaps more – the New Zealand economy grew at much the same rate as the rest of the OECD.
Chart 7 suggests five stages in the development of the New Zealand post-war economy relative to the OECD, although the endpoints may not be precisely those chosen here.
1954/5 to 1966/7: Upswing
In the 1954/5 to 1966/7 period, New Zealand GDP grew at about the same rate as the OECD, perhaps fractionally less. (This may be due to measurement error.)
1966/7 to 1977/8: Stepdown
Then, in 1966 New Zealand suffered a shock which put it on a slower growth path for about ten years. The next section shows the earthquake was the collapse in the wool price at the end of 1966.
1977/8 to 1984/5: Upswing
In the following seven years, the economy broadly followed the OECD growth path again, but at a relative level that was 18 percent lower than the path of the 1950s and early 1960s. The shock of 1966 reduced New Zealand’s output by 18 percent in the long run relative to the rest of the OECD.
1984/5 to 1993/4: Stepdown
Then from 1985 New Zealand underwent another period of stagnation, through to 1993, losing 11 percent by relative to the rest of the OECD. I shall return to why this happen.
1993/4 – ? :Upswing
Since 1994 the economy has been growing at broadly the same rate as the rest of the OECD, with fluctuations around the trend (e.g. the dip from the Asian Crisis in 1998). It may be the economy has been above trend in recent years, although it may reflect different cycles between New Zealand and the rest of the OECD, as occurred in the opposite direction in 1998 or in the early 1990s.
The new growth path is 11 percent below the path of the late 1970s and early 1980s. It is fatuous to say, as no less than authority the OECD did recently, that the New Zealand reforms are paying off. It is true that we appear to have returned to a growth rate comparable with the rest of the OECD – perhaps marginally higher – but the reforms will not have ‘paid off’, until New Zealand is above the 1977/8-1984/5 track, and has made up for the deficit between.
6. The 1966 External Shock and After
Chart 8 shows Terms of Trade (the price of exports relative to the price of imports) since 1927 (when the official data series first became available). There is a relatively low level before 1950, a high period (which in In Stormy Seas is called ‘the plateau’ up to 1966 and then (except for the 1971-1972 commodity boom) a low period from 1967 to 1985, followed by some recovery though not to the plateau levels. An alternative interpretation, discussed in In Stormy Seas is from 1950 there was a trending down of the terms of trade. (I shall return to what happened after 1984 in a section below.)
In December 1966 the export price of wool fell about 40 percent. Except for a brief flurry during the 1971-1972 world commodity boom, it never recovered relative to import prices. Indeed the wool price terms of trade were to sink to levels comparable to the Great Depression, although fortunately those for dairy and meat did not, although they have trended down in the post-war era.
In 1966 wool made up over 30 percent of export revenue, so total export revenue suffered by about 12 percent. With sheep-meats making another fifth of export revenue, New Zealand’s single largest export sector, providing over half of exports, experienced a severe blow in its profitability, while capital and skills which had been sunk into the sector became devalued and even valueless.
The immediate effect was for the economy to contract with unemployment beginning to rise in 1968. There was in 1967 – as there was had been in the 1932 – a devaluation to share the burden of the commodity price downturn across the entire economy, rather than concentrating it in a leading sector. However this time the terms of trade downturn was permanent – whereas there had been some recovery by 1940. Fortunately New Zealand was better prepared this time. Instead of clinging to the weakened sector, as happened in the 1930s, the New Zealand economy in the 1970s went through an export diversification – into horticulture, forestry, fishing, mining, general manufactures, and tourism. The diversification was spectacular – one of the most impressive economic performance triumphs of post-war New Zealand. John Gould has shown New Zealand shifted from being an extremist economy among the OECD in 1965 measure in terms of export concentration by destination and product, to a middling one in the 1981. No other OECD economy compared.
Even so this devaluation of capital and skills in the sheep industry together with the learning-by-doing diversification, slowed down the economy. The analytic issue is how long would such a transition take. Adjustment was confused by the 1971-1972 world commodity boom, so the external transition was largely completed by 1977. When New Zealand recommenced upon its growth path it was at a level some 18 percent below the previous one.
(The account can be modified without serious loss, by discerning a half step between OECD=1000 and the OECD=820 trend lines. It could be argued that the economy had adjusted by 1971 and was further devastated by the 1974 terms of trade downswing. This is still an external shock theory, but in this version there are two shocks. Its weakness is how to account for an upward terms of trade shock in 1971. The 1972-1973 peak seems to me to have been ephemeral – like the 1949-1950 one. Moreover, focussing solely on the 1974 terms of trade collapse misses the post-1966 slowdown. On a historical note, I mention I first identified the 1966 climacteric in 1974, before one could attribute any slowdown to the events of that year.)
7. Explanations for the Slow New Zealand per capita GDP Growth
My methodological position is that one looks at all the explanations to a problem and assesses each’s significance. While this may not be a particular profound – although it is anchored in a Popperian view of science – it is relatively unusual in New Zealand economics, where the explanations typically lock onto some hypothesis, and completely ignore any others (and any unpalatable facts). Among the explanations I have investigated and given some credence too are:
Post-war Catchup
The countries which were devastated by the war grew faster than those were not in the 1950s. It hints that the human capital, the technology and the social organisation which remained after the wear are all more important than the physical capital which had been destroyed.
Measurement Errors
My investigation showed a small but systematic downward bias in our estimates of volume GDP growth, arising from the difficulties of measuring volume changes in the service sector. Statistics New Zealand thinks it has now largely eliminated this bias, but it will reduce volume growth up to the 1980s. We dont know how widespread this problem has been in other OECD economies, although some compensated for it.
Population Growth
It is well established that population growth slows down per capita GDP growth, probably because it reduces capital deepening. New Zealand’s population grew faster than the OECD’s in much of the post-war era, and this would have slightly depressed its GDP growth.
The Convergence Effect
It appears that high income OECD economies grow more slowly than low income ones, probably because it is easier to import new technologies than create them. That effect would have slowed down New Zealand’s relative economic in the early part of the post-war era, but it is now a bonus for the economy, providing it has good policies for technology importation.
Terms of Trade
Declining terms of trade, that is lower relative returns for exports, act as a brake on the economy by slowing the supply of imports. A terms of trade shock can be very destructive, but there was a general pastoral terms of trade decline throughout the post-war era, partly because of the rise of substitutes – synthetics for wool, white meats for red meats, margarine for butter – but also because of increased Northern Hemisphere protectionism of domestic pastoral product markets, and dumping of their subsidised surpluses into third markets.
All these effects seem to have slowed per capita New Zealand GDP growth in the post-war era to some extent. But none – except the terms of trade – explain the transition from the pre 1966 track to the post 1977 one, nor the magnitude of the difference between two resulting paths.
8. Non-Explanations for the Slow New Zealand per capita GDP Growth
There are some popular explanations which hardly conform to any known scientific methodology. Basically they say that a phenomenon X exists, and therefore that explains phenomenon Y. There is little attempt to provide a causal path, to measure the impact, or to compare the explanation – such as it is – with other explanations. Basically the accounts conform more to the methodologies of pre-scientific superstition, although out of politeness we might call it ‘ideology’.
Excessive Intervention
It has been popular to argue in the 1980s that the New Zealand economic mechanism had been too dependent upon centralist interventions, which slowed down the economic growth rate. The policy prescription was that a major economic liberalisation, shifting the mechanisms to more-market, would accelerate economic growth. The evidence of the 1990s is that it did not. But here we evaluate the theory from an early 1980s perspective.
There was no attempt to demonstrate the connectedness of the proposition, nor to measure it. In particular, was New Zealand was more intervened than the countries with which any comparison was (implicitly) being made? Additionally the account is ahistorical: it is not obvious interventions intensified in 1966, while the period of fastest growth – from 1935 to 1945 – was a time when the economic mechanism was highly interventionist, much more so than it was in the 1970s.
The Popperian methodology demands we try to strengthen such a flimsy argument. A better theory might argue that the interventionism up to 1966 was basically benign: it supported economic growth and may even have accelerated it (noting there had been a process of steady post war liberalisation: by the 1960s the economy was not as closely intervened as it had been during the war). This benevolence of intervention probably applied after 1966 as illustrated by the great external diversification of the 1970s,. However there were two changes which required an acceleration of that liberalisation. The first was that the centralisation of the intervention in the internal economy had to be replaced by a more decentralised one because the diversification made quality central decision-making increasingly difficult. The diversification had created a new political economy requiring a new economic mechanism. Additional to these changes in production was greater social diversity, arising from affluence and perhaps some other social changes (such as increasing tolerance towards diversity), whose different needs could not be met by a centralised economic mechanism. To finish the story off, the Muldoon era from 1975 slowed down the rate of liberalisation so was a backlog in 1984.
Like the cruder theory, this suffers from an inability to measure the consequences of various levels of intervention. There are two identifiable quantifiable caveats to this,. First, a better signalling price system may reduce wasted investment. Second, periods of growth are associated with long business cycles. It seems likely that a more flexible the economy is able to prolong the peak of a cycle by, say, another quarter, hence promoting long run growth. (Note, the impact of different interventions may be more on the distribution and composition of output – its quality – than the aggregate level of output.).
Size of the Economy
By OECD standards New Zealand is a small economy. Such smallness has been equated with slower growth. But the same problems apply here as apply to the market mechanism thesis: there is a lack of connectedness in the proposition, there is no measurement, it is ahistorical because it gives no explanation as to what happened around 1966 (New Zealand did not suddenly get smaller), and it suffers from the defect that New Zealand was smaller in the past and yet grew rapidly at a relatively high standard of living.
The ‘advantages of size thesis’ has been recently challenged by Alberto Alesina and Enrico Spoloare, who point out that smaller OECD economies have better performance records than larger ones. They think there may be economies of scale which favour large economies, but these can be more than offset by the advantages a smaller nation has in governance over a less diverse group of people. They also argue that smaller economies can obtain many of the advantages of size by international trade.
Distance
It is unquestionable that distance has greatly affected New Zealand’s development . But that does not mean that distance from major markets has slowed down the growth of the New Zealand economy. The previous non sequiturs apply: there has been no attempt to demonstrate the connectedness of the proposition, nor to measure it, it is ahistorical, because it gives no explanation as to what happened around 1966 (the rest of the world suddenly became more distant), and it suffers from the defect that New Zealand was more distant in the past and yet grew rapidly.
(Consider the continuing and remarkable reductions in the cost of distance in the post-war era. If distance has been was an inhibition, one would have thought that they ought to have speeded up New Zealand’s economic growth rate. Moreover it seems likely that the reductions promise opportunities for new industries. The argument that distance is the cause of poor growth is not only unscientific but it reeks of policy defeatism.)
Before turning to analysing the second step-down, I want to list some faulty methods which sometime occur in the New Zealand economic growth debate.
Correlation is Not Causation
The tendency to connect unrelated facts which appear about the same time, without any analytic account of how they are connected, or empirical verification has already been mentioned. It is typically associated with the ignoring of facts which contradict the connection and alternative theories which might prove more robust.
Choose the Period Carefully
Statisticians must continually worry whether the conclusions are robust to the period chosen. This is particularly applies to the business cycle, since a trend can be changed by selecting the bottom of one cycle to the top of another. Another problem is the choice of a longer period. Beginning the analysis of post-war growth from 1970 misses the 1966 step-down.
Tautologies are Not Explanations
Much of the debate uses a mathematical formula as if it is a behavioural formula. For example there is the standard definition of Total Factor Productivity (TFP):
ΔTFP/TFP ≡ ΔY/Y – αΔL/L – (1-α)ΔK/K,
with a congruence sign (≡), rather than an equality sign (=), to remind that this is a definition.
In essence the equation says that TFP is the bit of growth of aggregate that cannot be explained by increases in labour and capital. As early as 1962, Tommy Balogh and Paul Streeten said TFP was the ‘coefficient of ignorance’, the part of growth that we cannot attribute to any measurable factors. Almost 50 years later we still have little empirical evidence as what determines TFP. We simply assign various things we suspect are relevant – such as technology, human capital, organisation. But advocating increasing TFP, without any behavioural theory, is merely arguing we should increase the coefficient of ignorance. (Many of the advocates are well placed to contribute to any increase.)
More popularly, the statistical summary ‘productivity’ is treated in a similar way. It does not actually tell us anything, and saying that the problem for New Zealand is we need greater productivity is a tautology even if it sounds impressive.
As a further example of an overused tautology, consider
Y ≡ (Y/A)(A/B)(B/C)….(W/X)X
One may be able to add behavioural content to the ratios in the brackets, but too often they remain ratios. Y is used in the example, to remind that the left hand variable is often aggregate GDP. Because it is difficult to provide accounts of aggregate variables, there is a tendency to lapse into tautologies when analysing them.
Relativities Not Rankings (See also appendix.) There is a return at its end to come back to here.
Thus far this paper has used the actual levels of New Zealand GDP, and implicitly – in Chart 7 – its relativity with the rest of the OECD. Much of the New Zealand discussion has been in terms of its ranking measured by GDP per capita among OECD countries.
(There are very real problems in using the PPP adjusted GDP figures. They are not very accurate, they dont project well back through time, and they suffer from an anomaly – which arises from inconsistencies in pricing between the exports and imports versus domestic production and consumption. They may not be measures of production at all, but more analogous to Real Gross Domestic Income. We proceed with caution.)
Anyone with a little mathematical skill would eschew working with rankings over relativities, since an ordinal measure is much harder to work with than a cardinal one. Moreover the cardinals are more information rich than ordinals, since the rankings can be derived from relativities but not vice versa.
However, whatever the mathematical distaste for using an inferior measure, rankings have misled researchers. Chart 8 shows both OECD relativities and rankings. Not only does the ranking pattern not closely follow the relativity, but for the first 15 years New Zealand hardly changed its ranking, although its relativity fell dramatically. The same applies to the last twenty years, when only Ireland passed New Zealand. Even so, New Zealand’s GDP per capita fell from the about the OECD average to just above 83 percent.
(The last time New Zealand was at the OECD average was in 1985, following a slight recovery in the previous half decade, but unrecorded in the rankings. Those who demand that we should aim to return to the top half would do well to remember that, for often they are associated with advocating the policies that dominated the post 1984 environment.)
A regrettable result from the focus on rankings has been the focus on the 1970s when New Zealand dropped nine placings, and ignore the problems of the post 1984 period. The earlier period is easily explained able in terms of the 1966 terms of trade crash. The later period is more complicated to explain.
Got to The Development of the New Zealand Economy: Part II
*************
APPENDIX I: Rankings and Relativities
The figures here based on Maddison with some interpolations for countries he does not report, using the official New Zealand series. The mean is based on 28 countries.
1950 (148 percent of OECD)
The following (OECD) economies (in probable rank order) already had a higher GDP per capita than New Zealand in the early 1950s.
United States
Switzerland
Luxembourg(?)
Canada,
So New Zealand was ranked fifth (although because the NZ series has not been adjusted for the secular measurement error, this ranking places New Zealand above Australia, rather than marginally below as occurs in In Stormy Seas p.27). New Zealand was then about 148 percent of the OECD average.
1951 to 1960 (148 to 131 percent of OECD)
No additional OECD country’s GDP per capita was above New Zealand by 1961. So New Zealand was still fifth, even though it had been growing more slowly that the OECD average and its relativity had fallen to about 131 percent of the average.
1961-1970 (131 to 111 percent of OECD)
Over the 1960s the following six OECD countries’ GDP per capita became higher than New Zealand’’s between 1961 and 1970, additional to the earlier four.
Denmark
Sweden
Australia
Netherlands
France
Iceland (?)
So now New Zealand was now eleventh, and its GDP per capita was about 111 percent of the OECD average.
1971-1980 (111 to 96 percent of OECD)
In the 1970s, the following eight OECD countries’ GDP per capita became higher than New Zealand’s between 1975 and 1980, additional to the earlier eleven.
Belgium
Germany (West & East combined)
Norway
Austria
United Kingdom
Japan
Italy
Finland
New Zealand was now nineteenth, and its GDP per capita was about 96 percent of the OECD average (although it would return to just above 100 percent (the mean) by 1984.
1997: TWENTIETH (86 percent of OECD)
In 1997 Ireland’s per capita GDP passed New Zealand’s. So New Zealand became twentieth, when its GDP per capita was about 86 percent of the OECD average. New Zealand was still ahead of
Spain (just)
Korea
Portugal
Greece
Poland
Czech Republic
Slovak Republic
Hungary
Turkey
Mexico
Return to Section of Rankings and Relativities