Revised version of a paper presented to the 1996 conference of the New Zealand Political Studies Association, Auckland.

Keywords Political Economy & History

Robert McDonald Chapman began his work on the New Zealand electoral system in an era where there were no opinion polls, no computers, and little readily available social data. It was a pioneering if, by today’s standards, primitive research program. which made him the father of psephology in New Zealand.(1) Yet his was not a mechanical manipulation of the available electoral data. Behind it all was an account of the political economy of New Zealand. It is this aspect that this paper recalls an updates.

Chapman’s key ideas are in the final chapter of the seminal study of the 1960 election which he co-authored with Keith Jackson and Austin Mitchell.(2) It begins:

“A basic social and sectional division in New Zealand separates city from country and this has revealed itself in our politics for as long as there have been parties. It is much more than a line separating two ways of making a living or two economic interest groups. The division passes between two approaches to social questions. It delineates different psychologies arising from two sorts of experience of life. It is so fundamental and arose so early – it is to be found in the 1840s – that it has long since found stable expression in the two-party structure itself, one party urban, one primarily rural, and within constituencies as a predictable and usually unequal balance of sections. Intercommunication between the two nations, rural and urban, is constant. Country families see their children go to the cities, parents retire to the towns and a high proportion of farmers, as recently as the 1930s, were city bred. Underneath the surface interchange, the geography of attitudes nevertheless remains fixed, while above it passes the political and economic weather, with an occasional cloudburst in one section or the other, but usually in fluctuations common to both because both sections are interdependent parts of the one economy.”(3)

The chapter divides the 80 electoral seats of 1960 into four groups: Maori (4), Main Centre (33), Farmer Seats (20), and 23 seats which were not entirely urban, but even so more than half lived in boroughs or town districts. This final group was then split into 10 provincial town seats, with the residual of 13 “mixed town and country” seats. The chapter looks at the all the categories, and is especially interested in the main city seats, plus a number of special topics (including non-voting, subsequently a favourite Chapman concern).

Mixed Town and Country Seats(4)

Chapman’s prose, which evokes a New Zealand that is now more a nostalgic memory than reality, draws attention to the economic structure of these mixed seats.

“A dozen years ago in analysing the elections between 1908 and 1935 I marked out a class of seats in which a distinct majority of farmers was admixed with a minority of townsmen. Changes of opinion in this class of seat won or lost the elections of the 1910s and 1920s and they proved more, not less, responsive than purely rural seats to whatever was affecting the farming community. Merchants in these service towns felt the changes of the market in exaggerated form as their customers increased or cut their orders according to the pattern of receipts in London. It was always possible for the farm family to live lean on the farm’s ancillary products, but a shopkeeper with full shelves and little trade in a stagnant district was driven to strike back at the situation by blaming the Government or demanding of candidates that they secure developmental money and works or fresh settlers to get the district moving. These demands were not different from some, at least, of pure farmer demands; they were simply made more sharply and more speedily by sufferers from the economy who lived in little towns. … The crux of the political question posed by the growing towns seems to lie in whether multiplying the self-employed and service employees, the shop assistants and sales clerks and small shop and storekeepers, will result on balance in an addition to the Labour vote so heavy as to outweigh the farmer, or else give rise to a vote sufficiently split by sympathy with the customer and identification with his values that National wins because nowadays the customer is always to the Right.” (5)

While he did not discount the relevance of income, Chapman was not trapped in an overpowering vision of voting outcome being determined by the balance between the rich against the poor. While there were extremes of wealth and poverty in New Zealand at that time, the New Zealand income distribution was then compact by international standards, in which case income may not be a powerful discriminator of voting behaviour.(6) Instead Chapman focuses on the voter’s place in the economic structure, as well as income and class.

Much of what Chapman’s writing about social behaviour in the mixed electorates is seat-of-the-pants conjecture, based on personal observation, rather than the results of systematic analysis. Over two decades later, anthropologist Elvin Hatch’s study of a rural South Canterbury community, broadly confirmed Chapman’s account of how social class was practically moderated by the community and economic requirements of pastoral farming.(7)

Large Town Constituencies(8)

Chapman repeats the broad outlines of his mixed town and country seats in his discussion of the provincial towns.

“The large towns balance within themselves grades which in a city might have a constituency largely to themselves. Add to the urban accompaniments in a large town and immediately the city-style sectional grades is affected. The same addition of say, fertilizer works and its labour force, will also alter the balance between rural and urban orientations. Thus two kinds of analysis interest, two categories are found, two perimeters parallel one another in the large towns: rural/urban on the one hand and higher and lower socioeconomic grades on the other. Indeed there is an overlap, almost a fusing of viewpoints in the large towns between the attitude of those who think with the countryman, being themselves proprietors, agents, professionals and service employees, and those who, belonging to the same groups, base their political judgements on opposition to the party of the unionists, industrial workers and of the lesser grades generally. The converse is rather less true. Unionists and industrial workers give little evidence of thinking at all along the urban/rural dimension or in any way taking up a position governed by hostility to or identification with the rural community and its interests. Rather their attention and their politics seem rivetted to the vertical dimension of status and possessions visible in the cities. They treat their environment as though it were altogether urban, which in their factory and dormitory suburb environment it is, and recognize `our party’ accordingly. `Our party’ for the other side in the large towns may be `ours’ in the first place because it is the party of people like us or because it is thought to recognize enterprise in the small owner and keep an eye on the militant unions, or primarily because it is envisaged as understanding the farmer and we all depend on the farmer. Such attitudes, of course, sit easily in combination and together they make it possible for the National party to battle strongly in what at first glance ought to count as wholly urban and probably as enemy territory. The result is an overriding likeness in this group of electorates which straddle the dividing lines of New Zealand sectionalism, predominantly service towns slightly to the Right, more industrial towns more to the Left, but all within reaching distance of the 50 per cent line. In consequence the class of large town electorates perform nearest to the New Zealand political norm for they contain both the major divisions of New Zealand life. Some are microcosms of all New Zealand, rurally oriented and urban minded in balance, some merely concentrate the sectional appositions of our cities; but together the swings of large town opinion contribute much to the fate of the parties and altogether they may be truer mediators between city and country, matching the inclinations of both, than the mixed electorates, committed as they are to one pole.”(9)

Chapman ranks the ten seats by their voting behaviour.(10) His tabulation is an early version of the bottom of a electoral swingometer, which did not become popular until the mid 1970s.(11) Perhaps had Chapman had the computing power that became available only a few years later, the 1962 chapter would have had a full blown swingometer.

The Underlying Political Economy

Chapman has an New Zealand political economy implicit in his studies. Economists later called it the “two-legged” economy.(12) The rural pastoral farm leg earned foreign exchange for importing and debt servicing, while the urban industrial leg used imports to produce goods and services for the domestic economy, creating jobs for those whom the pastoral sector and its satellites could not directly employ. One of the strength of Chapman’s analysis is to remind us the rural leg included a host of other activities and occupations which were dependent upon the farm sector, as well as the farmers themselves.

Out of this bifurcated economy there had evolved two distinct political parties – almost equal in size – with very different social bases reflecting the two main components of the pastoral settlement political economy, modified a little by income levels. National was the party of the countryside and the farmer with support from urban commerce: Labour was the party of the city and the industrial worker, with a component of the poor rural worker.

The policy debate about economic strategy reflected those two legs. Those in the rural sector argued for low – preferably zero – protection because, they said, high levels reduced economic efficiency. The urban sector argued for protection to generate jobs, pointing out that economic efficiency applied only to the employed, and free trade had no means of dealing with the inefficiency of unemployment. This is not the place to evaluate the efficacy of the two prescriptions. Here we note that it seems likely that the special circumstances of the pastoral sector probably meant that any efficiency losses from protection – assuming there was full employment – were small. Probably the main effect was to transfer the high land rents from pastoral farming to the nation as a whole, and the landless in particular, through higher wages and broader government services. The allocation of this rent was at the heart of much of the policy debate at the time.(13)

There is a yet unresolved question of when the two-legged economy evolved. At the turn of the century the economic structure was probably still dominated by the pastoral export sector. There was a manufacturing sector, but when Gary Hawke divided the secondary sector into those who were in “industry” and those who were in “handicrafts” – the first category covered those who appeared in the factory statistics, while the second were those manufacturing workers in the population census who did not work in factories – there were more handicraft than factory workers until the beginning of the twentieth century.(14) As the Caversham studies show, even factory workers were then still organized more on a craft basis than an industrial process basis.(15)

It seems likely that the industrial worker evolved in the middle part of the twentieth century, rather than the nineteenth. The big change in the freezing works from craft butchering to the industrial chain really occurs in the late 1920s and early 1930s, car assembly plants occur a little later. This transformation of manufacturing was crucial in the growth of the Labour Party. Obviously there is no day in which the two-legged economy sprung fully clothed into the world, but at some time – by the 1930s – it became sufficiently large to have a substantial impact on the political system. This contrasts with the myth of Labour’s origins in the mines of the West Coast. Historically that is true, but Labour had to move outside there into the urban factories to become government. It is sometimes commented that Labour under went a transition as it moved from the leadership of mine worker Harry Holland, MP for Buller, to brewery worker Michael Joseph Savage, MP for Auckland West. The change symbolizes the changing political economy to which Labour was responding.

Suppose we turn the swingometer upside down. It now looks like two legs of the economy, with the one the import substituting domestic economy, the other the export oriented farm economy? The marginal seats – especially those mixed ones – being the crotch, where the bowed legs join together.

The Urban Seats

Yet it would be misleading to say that Chapman had theory based only on economic structure. A central part of his account is the rising role of the urban seats. He sees cities allowing diversity but it in that diversity there is specialization. Given his prime source of data is electoral booths, the specialization he observes is a stratification by location. There are safe Labour seats, safe National seats and marginal ones, which he argues were differentiated by income. Given the primitive data base it was a heroic effort to trace this.

The result was a number of social groups found themselves in ambiguous locations. Freezing workers were politically with Labour, but their economic interest is in the pastoral export leg. Manufacturers were in the domestic sector, but politically with National. These effects moderated the purity of the political divide being simply along political economy lines, because each party had an interest in both legs, even if there was a preponderance in favour of one. Thus there was not great differences in economic policy between the parties.

About the time that Chapman was writing various people claimed that New Zealand was a classless society. For instance Arnold Nordmeyer, then leader of the Labour Party, said to the 1963 conference of the Federation of Labour, “We are one people and the sooner we realize there are no real class divisions in this country the better it will be for us all.” It was even claimed that Keith Sinclair said so in his immensely popular history. (17) In fact Sinclair wrote “New Zealand is not a classless society. It must be more nearly classless, however, than any other society in the world.”(18) Since for logical reasons there exists one society for which this statement is true, Sinclair may be accused of no more than suffering the LBW syndrome – “leading the bloody world” – which he himself first identified. (He may even have been correct.)

Sinclair’s statement is backed only with a series of anecdotes and otherwise unsubstantiated impressions. Chapman offered an analytic account of why class might not be so important in New Zealand as elsewhere; why the imposition of a European account of class structure to New Zealand was not entirely appropriate.Moreover social scientist Chapman had an elaborated enough model to make predictions about future developments of the class structure. He expected the urban centres would grow, and as they did class stratification would increase. As it happens the prediction never had the opportunity to be properly tested.

The New Political Economy

In the mid 1960s the two legged economy began to fragment. The terms of trade, the price of the pastoral exports relative to the price on imports had been under pressure since the mid 1950s, but the collapse of the wool price in 1966, together with ongoing weaknesses in the price of red meats and butter throughout the post-war era, meant that the pastoral sector could no longer carry the foreign exchange earning burden the economy required to maintain a high standard of living.

Fortunately a number of other industries developed rapidly as exporters: tourism, forestry, fishing, horticulture, manufacturing. (At the same time – socially and politically critical as well as economically so – the exporters developed new markets especially in Australia and East Asia). The story of this diversification dominates the political economy in the last third of the twentieth century (although there is also a diversification of social behaviour of which the feminist revolution and the Maori renaissance are the most spectacular examples). The economic diversification was not just a matter of a new bunch of industries, occupations, and activities joining the export leg. Now there was no significant land rent to transfer to support the domestic leg, and so it collapsed. Just about all industries became a part of the export effort – either directly or indirectly by supplying exporters.

Inevitably the political economy of the post 1966 diversified economy has yet to be delineated in any detail.(19) My remarks here are confined to the impact on the electoral swingometer, which I use as a crude summariser of Chapman’s broad thesis.

The swingometer worked in the Chapman model because while there were two significant dimensions: urban-rural which were represented in the marginals by the provincial seats, and rich-poor represented by the marginal city electorates. Either because the two dimensions were independent of one another, or because of a high degree of association, they could be mapped onto the same scale.

There is a further requirement for the swingometer to work. It is not just a matter that geographical clusters have a broadly uniform swing, something which Chapman demonstrates for the seven elections, but the swing must be able to be characterized by a single parameter. This is true only where there are two parties. The switch matrix which characterises electors changing their party vote has two parameters, but they may be condensed to a single one. Once there are more than two parties the number of parameters increases, but regrettably not in proportion to the party number but roughly proportional to the square of that number.(20) Thus when there are three parties, three parameters are needed. For four it is six. So it becomes impossible to represent a mutli-party race by a simple swingometer, unless the empirical conditions are exceptional.

The first-past-the post electoral system maintained the hegemony of two parties till 1993, so the swingometer could continue. Even so if Chapman were to repeat his analysis for, say, the first seven elections after 1960 the pattern he found would not be so evident.(21)

However with the collapse of the old political economy, the electorate could no longer be strung along a single dimension in the way that Chapman and the swingometer could locate them. More recent electoral studies have had to use two and three dimensions.(22)

This may not just be the bigger data base and the greater computing power. It almost certainly reflects the greater social and economic diversity. That certainly is the view of a good number of today’s political parties who, no longer constrained by FPP, seek niche groups.

But if the Chapman account of political behaviour, and its successor swingometer, is dead we have to ask what is the alternative? It is not just the new voting environment. At the moment we are overwhelmed by data, but there is no compelling organizing theory to go with it. It is ironic that if Chapman was handicapped by a lack of data and a lack of computing power, he was also blessed by it, because it forced him to seek a simple – yet powerful – theory to go with the facts available to him. It is probably too soon to expect an overall coherent account of the politics of the new political economy of New Zealand to replace that which Robert Chapman gave us in the early 1960s.

Endnotes
1. Chapman’s collected works are published in E. McLeay (ed) New Zealand Politics and Social Patterns: Selected Works of Robert Chapman, Victoria University Press, 1999.
2. R .M. Chapman, W. K. Jackson, & A. V. Mitchell New Zealand Politics in Action: The 1960 General Election, Oxford University Press, London, 1962.
3. op cit p.235.
4. In 1960 they were Ashburton, Bay of Plenty, Marlborough, Marsden, Rangitikei, Rotorua, Tauranga, Wairarapa, Waitaki, and Westland.
5. op cit p.236, 238.
6. B.H. Easton Income Distribution in New Zealand, NZIER Research Paper No 28, Wellington, 1983.
7. E. Hatch, Respectable Lives: Social Standing in Rural New Zealand, University of California Press, Berkeley, 1992.
8.In 1960 they were Gisborne, Hamilton, Hastings, Invercargill, Napier, Nelson, New Plymouth, Palmerston North, Timaru, and Wanganui.
9. op cit p.242.
10. op cit p.240. There is also one for the Marginal City Seats at p.246.
11. A. McRobie & N. Roberts, Election ’78, John McIndoe, Dunedin, 1978.
12. B.H. Easton, In Stormy Seas, Otago University Press, Dunedin, 1997, p.47,83.
13. The argument is elaborated in “Prescription or Poison: `New Zealand can be Different and Better’ by Wolfgang Rosenberg”, New Zealand Books, December 1993, p.5-6, and Easton (1997) op cit Ch 5.
14. The terminology is a little misleading, since the handicraft category includes self employed workers not employing labour (e.g. a tailor). See G. R. Hawke, Disaggregation of the New Zealand Labour Force, 1871-1936, VUW Working Papers in Economic History, 79/1, Victoria University of Wellington Economics Department, 1979.
15. E .A. Olsen, Building the New World: Work, Politics and Society in Caversham, 1880s-1920s, Auckland University Press, 1995.
16. Reported in W.B. Sutch, The Quest for Security in New Zealand: 1840 to 1966, Oxford University Press, Wellington, 1966, p.405. Nordmeyer’s first seat, incidentally, was Waitaki, one of Chapman’s mixed country and town electorates, and close to where Hatch did his fieldwork.
17. e.g. D. Pitt (ed) Social Class in New Zealand, Longman Paul, Auckland, 1977.
18. K. Sinclair (1959) A History of New Zealand, Penguin, London. p.276.
19. But see Easton (1997) op cit.
20. The number of parameters required to characterise a switch matric is n(n-1)/2, where n is the number of parties.
21. In practice the exercise would be very difficult because electorate boundaries were less stable, that itself being indicative of the rapid changes going on in New Zealand society.
22. e.g. J. Vowles, & P.Aimer (1993) Voters’ Vengeance, AUP, Auckland Press; J. Vowles, P. Aimer, H. Catt, J. Lamare, & R. Miller (1995) Towards Consensus: The 1993 Election in New Zealand and the Transition to Proportional Representation, AUP, Auckland.