Science and Nationbuilding

Revised paper presented to the Rotorua Branch of the Royal Society of New Zealand, 1 May 2002

Keywords Growth & Innovation; Political Economy & History

The Nationbuilders is a book about the economic social and cultural development of New Zealand from 1932 to 1984 when a group of visionary New Zealanders developed the nation. The story is told through a set of biographical essays, but while some have read the chapters separately for the individual stories, in fact the book has a series of themes, which the lives illustrate.

Thus while there may be no specific chapter devoted to science – although I shall add important caveats shortly – science permeates the book, as you would expect in any study of twentieth century national development. I did contemplate a biographical chapter on the DSIR, which would also illustrate the development of a government department other than Treasury, one of the themes of the book. The intention would have been to give prominence to the Grasslands Division which may be the most spectacular single scientific contribution to New Zealand’s development. However space was limited. My original plan was a 60,000 word book, but my publisher began to show signs of agitation as I cruised up to 100,000 words – 320 pages – so something had to go. And to be honest, I would have given higher priority to the Dairy Board.

But you will find a science story there in the book if you look. The first minister of the DSIR was Gordon Coates, the first full chapter biography. It is a typical Coates story. He had just become Prime Minister, he had a report from a British expert in March 1926, and the man ‘who gets things done’ implemented its recomendations by the end of the year, apparently without a lot of consultation, and he took the portfolio himself. (He also began the planting of the Kaiangaroa Forest and established Massey University – a truly remarkable New Zealander.)

The book also has two major applications of science in its story of Tasman Pulp and Paper which occurs in the chapters on Bernard Ashwin and James Fletcher, and there is a separate chapter on New Zealand Steel. Ross Galbreath’s history of the DSIR devotes an entire chapter to the scientist’s contributions to the processing of ironsands to steel. (I also add, that it gave me considerable pleasure to learn that the two business were both the result of tectonic activity. A Taupo volcanic eruption led to the Kaiangaroa Forest, and a much older eruption from Mt Taranaki generated the ironsands. In his summing up of my address, Roland Burton added to my pleasure by adding that a recent paper reports that the hydrocarbon depositis in the Tarankai area are also in part due too volcanic activity.)

Two scientists have chapter biographies. I shall use to Douglas Robb to describe the conduit role in science, so I merely mention here that he was a key New Zealand player in the transformation of New Zealand surgery from artisan craft to applied science, albeit one where craft skills remain vital.

The other scientist was Bryan Philpott, who followed the best scientific traditions of economics of progressing the subject by the empirical testing of hypotheses. One of the purposes of this paper is to assure you that economics can be a science, although many ideologues like to treat it as otherwise. The book uses Philpott’s science to explore the development and application of economic theory. I want to tell this story in a slightly different way tonight.

One of the book’s interests is the way that economic theory developed and became applied to New Zealand. Many commentators assume that economic theory is static and does not progress. In fact there is ongoing paradigm development, paradigm conflict and paradigm displacement. While at the time it is hard to see anything but a melee of controversy, one of the functions of the intellectual to make sense of it – at least with the advantage of hindsight. Nationbuilders describes the revolutionary paradigmatic clash, in which the institutionalist paradigm was displaced by neoclassical synthesis in the early postwar era. I am not going to describe this clash tonight – it is in the book – but I would caution that without understanding it, it is very hard to make sense of the late part of the nationbuilding period, in which Rob Muldoon operated from largely the institutionalist paradigm and was challenged by the neoclassical one. Rogernomics was, intellectually, an extreme – and incompetent – application of neoclassical economics, which has many attractions – including its mathematical rigour – but which does not dominate institutionalism on all dimensions.

In the 1950s the neoclassical paradigm had two major progressions which had significant policy applications some years later. The first might be called general equilibrium theory. Its basic idea is that prices are a signalling system which individuals use to make economic decisions. The best set of prices for making economic decisions are market one for which the government does a minimum of intervention, where there are no subsidies, where there are no taxes (or if they are necessary they are flat like GST), where there are no artificial barriers to entry such as a licensing regime or external protection, and so on.

The merits or demerits of this approach belong to another presentation. For our purposes we observe that there was a tendency to implement the policy conclusion – reduce government interventions – from the 1950s increasingly through to 1984, a tendency which Muldoon did not totally resist. After 1984 the policy conclusion – eliminate government interventions as much as possible – was implemented with a vengeance. The promise was that the existing resources would be allocated more efficiently with the result that there would an increase in the level of output for the available resources.

In fact there is little evidence for an overall boost in the level of economic activity after 1984 which the theory predicts should have occurred. The economy actually stagnated, although that might be attributed to other factors. Even more surprisingly there is no evidence of an increase in productivity which should have occurred if the theory on which the policy conclusions is based is correct, independent of whether the economy expanded or stagnated. The discrepancy between the prediction of the theory and outcome of its application is so enormous that it must qualify as a critical experiment, which so challenges the dominant paradigm that it has to be ignored until another scientific revolution resolves the inconsistency.

Even had there not been this paradox, the theory would have been vulnerable to quite a different criticism. General equilibrium theory is not a theory of economic growth, but a theory of a static economy, assuming there is a growth process, rather than explaining it. At best, the efficiency improvements that its policies promised would have lifted output and provided a foundation for economic growth, but in themselves they would not have generated any.

Fortunately, another progression in the neoclassical paradigm offered some account of the growth process. Neoclassical growth theory might be attributed to Robert Solow. A particular paper of his, written in 1957, is widely quoted in the New Zealand science debate, as if he wrote nothing else of significance, and there had been no progress since then. It has always astonished me that scientists who would be shocked if an economist used as their only source on scientific topic a paper that was 45 years old, should do the same for an economic research paper. I mention this because one of the oddities of the debate I am about to describe is that scientists did not tackle the policy issue with the sort of scientific discipline they expect in their own work.

Bryan Philpott did. He pioneered the application of the measurement of the aggregate production function in New Zealand, often literally with his own bare hands creating the data base, updating it for forty odd years until shortly before his death, and interpreting it right to the end. But he was not an uncritical user of that approach and he would not dissent from anything I have to say about it today. The only surprising thing about his record is how little recognition he got for this work and how, since he has died, there has been so little interest in preserving and maintaining his research program. I suspect that is because we are happy to rely on rhetoric to justify policy conclusions, and are not really interested in the scientific content of the research. Forgive me then if I talk about the paradigm a little, as free from the rhetoric as I can make it.

Solow’s central finding, replicated for many other data sets, for other periods and for other countries – including by Philpott for New Zealand over all the periods for which he had data – can be summarised this way. Suppose the amount of capital and labour and other inputs increase in an economy by 10 percent over a period. Then we might expect the economic output to increase by 10 percent too. In fact it increases by more than that, measurably more than that. So there must be something else which is increasing output over time on top of the additional labour and capital and so on.

Solow’s seminal paper is a beautifully written and worth reading for its elegance, but also for its scientific discipline. There is actually a mistake in the pre-computer days calculation, but unaware it is a mistake, Solow did not explain it away by an ad hoc theoretical adjustment , but remarks ‘for the present, I leave this a mystery’. How terrible it would have been had he completely explained the anomaly by a theoretical modification to learn later it was but a computational error. The scientist did not.

The paper described the other source of output, and hence the main source of economic growth, as ‘technology’, a term which has been seized on by the science community without any understanding what was meant, and how it was used in such expressions as ‘80 percent of economic growth can be attributed to technology’. Solow defined his concept:

‘I am using the phrase ‘technical change’ as a shorthand expression for any kind of shift in the production function. Thus slowdowns, speedups, improvements in the education of the labour force, and all sorts of things will appear as ‘technical change’. (Solow’s italics)

In fact the technical change – today it is called total factor productivity or TFP – is anything that cannot be explained by increases in labour and capital – a residual. A couple of British economists, Tommy Balogh and Paul Streeten went as far as saying that the residual was a ‘coefficient of ignorance’. You could say that those who think that we can increase economic growth by higher technical change are saying we should increase our coefficient of ignorance. Given the way they go about their advocation, they may well have.

Economists have of course tried to reduce the coefficient of ignorance by directly estimating the other factors contributing to economic growth. The results are largely unsatisfactory for various reasons, and even so often there remains a residual – albeit a smaller one. One result is that some of the residual can be attributed to the increasing economic quality of the labour force (although there are some theoretical difficulties with the measurement). While each generation is no better than its ancestors, each acquires more economically useful skills which contribute to it being able to perform more productively. This is a complicated area and I want to draw here but two points.

The first is that while those skills usually come from scientific knowledge (although in poorer countries there may also be substantial economics gains from a healthier work force), the knowledge is embodied in workers. Thus education and training can also claim to have a substantial contribution to economic growth. Not only do we need to be aware of the danger of double counting, but we can see there is a transmission and application issue. Even if knowledge is the primary driver of rising productivity, it is far from clear how it works. I will return to the transmission and application issue, but it is further complicated by the second point.

Just as knowledge is systematically embodied in human beings, it is also embodied in physical capital. Indeed all the complications of transmission and application of knowledge we see with human capital apply here too, and probably interactively between labour and capital, since capital with newly embodied technology is of little use unless there is the labour with the knowledge to use it and vice versa. Moreover it is not just the labour and the capital with their embodied knowledge, but a key factor may be how they are organised in the workplace, in the firm or establishment, in the industry and in the economy. General equilibrium theory implies that the organisation of the economy is very important, but sophisticated levels of organisation are probably as important, although harder to address in policy terms.

Ultimately the problem is the Solow approach is so aggregate it obscures the really interesting issues. For instance, the method, and much of the discussion based on it, assumes that capital is a well defined and readily measured notion, but how does one aggregate together a one horse shay with a 747 into a single index? Solow was aware of the problem of aggregation, neatly sidestepping:

I would not try to justify what follows [that is the measurement of technical change and the aggregate production function] by calling on fancy theorems on aggregation and index numbers. Either this kind of aggregate economics appeals or it doesn’t. Personally I belong to both schools. If it does, I think we can draw some crude but useful conclusions from the results.

‘Crude but useful’. Exactly. That is the best that we may hope for from such analyses. And Solow’s marvellous paper is just that. It points out the issue of economic growth is not just additional capital per worker. There appears to be some other important phenomenon which contributes to economic growth, and without which there would be little improvements in productivity. But we are far from clear what is this ‘technical change’. To become more enlightened we cannot simply pursue the Solow’s aggregate approach, we need to investigate at the microeconomic level.

There is a mass of research done at this level on the role of technology in economic growth, although virtually none of it in New Zealand. Other than Galbreath’s chapter, and that is history not science, We do not have a comprehensive study of the extraordinary success of the Grasslands Division in their research, transmission and application of pasture management principles. (Gerard Horgan, with whom I stayed, reminded me that the radiata pine project was also a great achievement, but there is even less written on that.) More generally, there are extreme difficulties in this research and economists do not yet have a comprehensive account of how knowledge turns into prosperity.

The problem is that it is very rare that economics or science policy issues can be studied via traditional experiments. Economic research has to use a different approach because no one will let us properly experiment on the economy, except by accident.

Consider the Rogernomics experiment. Before and after analysis of the New Zealand economic performance shows that not only did the promises for outcome of the reforms generally fail miserably – the exception is the disinflation where price stability was attained even faster than was expected – but very often the reforms led to a worsening of the outcomes the reformers highlighted. (Incidentally, Bryan Philpott was one of the few that worked on the careful measurement of these indicators). There is even a contrast for Australia approached the problem of economic reform differently to New Zealand from a similar baseline. The comparison shows Australia performed much more successfully than the extremism of New Zealand. Yet a decade later the reformers prance in the public arena as though there had been no test on their theories or that an impartial scientific one had found they were a success. This is Lysenkoism. Just imagine someone advocating a new fertiliser. The outcome following application is a reduction in crop yield. Yet they still promote it as if it would give a major boost to biomass, while the scientists who drew attention to the evidence of failure were condemned.

Hans Christian Anderson was wrong in the way he finished his story The Emperor’s New Clothes. By the time the boy draws attention to the Emperor’s sartorial inadequacies, the onlookers have followed the fashion and are cavorting around in the nude. So they beat the little boy. A few years later the fashion has changed, everyone is now heavily dressed and they studiously ignore the boy who had the courage to point out the past absurdity – after all who wants to recall making a fool of themselves, and in any case the current fashion may be as absurd.

I have spent a little time on the peculiarities of my own science in public debate, because there is a parallel development in science policy. For it was as ideologically led as was the economics reform one was, and the science leaders often let their scientists down. The problem was that science was absorbing substantial fiscal funds and demanding more. Aggregate production analysis’s conclusion was that technical change had a significant impact on the economic growth process, suggested that science was an investment in economic growth, so that by spending more on science there would be higher economic growth. (Educationalists, facing the same problems, came to the same conclusion. Treat education as an investment in economic growth to justify large and increasing expenditures on education. This lead them down a commercialization path too.)

Now the analysis is broadly true, but it leaves the critical policy issues totally unresolved. What science should we be investing in, and how do we transmit and apply the scientific findings into economic performance? The aggregate production approach cannot answer these questions, for that requires detailed microeconomic investigations. Unfortunately the profession’s leadership had seized on the idea of the science as investment without trying to understand the underlying theory nor seeing its weaknesses, and certainly not looking at the microeconomic evidence that was trying to address those weaknesses. This was the rhetoric of policy ideology not the rational analysis of science.

It is not surprising then, that the leaders linked up with another ideological rhetoric, that of the Rogernomes, who argued that if an activity was an investment then the most effective strategy was to treat it as private property in a competitive market environment – that is to commercialise science research. Basically, the approach was quack medicine, not just a placebo but something that made the patient worse, because it damaged the science industry rather than enhanced it.

There is not a lot of evidence that this commercialisation resolves the transmission of knowledge problem, except relatively late in the application cycle when property rights becomes practical – the effect of a patent is to privatise an idea. The industry in which this has been most successful is pharmaceuticals. However perhaps the pharmaceutical industry could have been better organised on different principles. In any case the way it is organised today is not a good general model for creating, transmitting and applying other science to promote economic growth.

A good counterexample is the Grassland’s Division, which was not organised on good commercial principles, which was extremely scientifically creative, and which transmitted its knowledge effectively to farmers who successfully applied it to the benefit of the pastoral industry and the New Zealand economy. But counterexamples rarely dissuade ideologists.

A nice illustration of this point appears in a recent NYRB by physicist Freeman Dyson. He wrote of the author whose book he was reviewing:

[His] argument makes sense if you accept the rules of theological argument, rules which are different from the rules of scientific argument. The way scientific argument goes is typically as follows: We have a number of theories to explain what we have observed. Most of the theories are probably wrong or irrelevant. Then somebody does a new experiment or a new calculation that proves Theory A is wrong. As a result Theory B now has a better chance of being right. The way a theological argument goes is the other way around. We have a number of theories to explain what we believe. Different theologians have different theories. Then, somebody, in this case [the author], declares Theory A is right. As a result, Theory B now has a better chance of being wrong.

The scientific problem seems to be that the neoclassical account which underpins the commercialisation policy assumes that price signals are sufficient to organise the production processes in the (competitive) industry. My current work in this area is more concerned with the health industry, where it seems that the conditions for good quality price signals in a competitive market do not exist, as is evident in the professional ethics of doctors. The same seems to apply to science, including that the reward system for scientists gives priority to early publication that being in the best interests of scientific progress if not commercial profit.

Perhaps it is not surprising that the hardline commercialist reforms of the government science sector (and indeed the health sector and elsewhere besides) soon fell apart. We are still trying struggling with how to incorporate science research and development into an overall economic strategy. A critical part of the answer may lie in the intricacies of the production process.

In order to discuss recent developments in knowledge policy, I want to identify a wider remit of science than just commercial policy. I can identify five Cs of science, only one of which is commercial.

Science is also a cultural phenomenon or, using the current government terminology, a ‘creative’ one. What that means is one does science not just for reasons of commercial outcome but because like reading and theatre and music and dance it enhances cultural life as it promotes the creative, curiosity and a sense of wonderment in the individual human being. Recall my pleasure at how tectonic plate activity linked the origins of Tasman and NZ Steel. Even if geology had no other use – it does of course – that better understanding of where our beloved land came from is a contribution to our welfare. If we spend public money on recreation – as we do for the arts and sport – perhaps we should spend public money on recreational science. For instance it seems highly unlikely that astronomy research in New Zealand will give any commercial benefits. But there exists throughout the country amateur astronomy societies whose members are passionate about the awe of the multitude above us. There is a case for supporting them, not by purchasing telescopes – the equivalent of cricket fields and theatrical facilities – but by funding a small but active national astronomy research program. It is not obvious how much the government should spend on science for such creative/cultural purposes but, then again, it is not obvious how much it should spend on rugby or symphony music. Whatever the practical answer, science betrays itself if it designs science policy which ignores the creative/cultural role of science.

The third C of science, is the need for a capability to deal with scientific issues which may arise, but not with certainty. It is said that the last bee scientist in the public service was about to be made redundant when the threat of the bee mite turned up. We will soon be spending $100m a year on quarantine protection of our external boundaries, and ten times that on a defence force we hope will never be used. Competence in science can have a similar protective role. Nuclear power is statutorily prohibited in New Zealand, and yet we need to have scientists who are knowledgeable – to advise on international issues and because we need a competency for the unexpected: suppose a nuclear reactor in a dying satellite fell on a piece of New Zealand; suppose a nuclear power ship in distress asked for haven in a New Zealand port.

The fourth C of science is the conduit, a mechanism which channels international scientific knowledge into New Zealand. Douglas Robb was among those who had this role in mid-twentieth century medicine. Scientific surgery was blossoming just before the First World War in a few centres such as the John Hopkins Medical School. Robb seems to have picked up the approach when he was trained in England in the mid-1920s. It certainly made him unpopular with the craft surgeons who flourished in New Zealand at the time of his return. They were most disturbed when he tried to introduce medical auditing into surgery practices and tried to train the house surgeons in new techniques in the 1930s. His writings advocating a public health service involves a collegial approach to medicine with research, libraries and ongoing continuing education. This may be conventional today, but it was new then. When the American forces came during the war, he organised demonstrations by their doctors from which our doctors could learn. He was constantly visiting overseas hospitals, and in the late 1940s visited John Hopkins to learn of developments in heart surgery which were soon applied here. He was vitally concerned with Auckland University, being on its Council for 35 years and it was his vigorous leadership which speeded up the foundation of the Auckland University Medical School. This was not a conduit which dumped the new science into New Zealand. He was energetically and sensitively adapting and applying it. The outcome of his Greenlane Heart Unit gained a world class reputation.

One thing that Rogernomics taught us was that even where knowledge is being conduited there is an application stage in New Zealand. It cannot be used raw. There was little attempt in the 1980s and 1990s to adapt the ‘international’ (that is American based) theories to a small open economy like New Zealand, and they were applied as if New Zealand was the United States of America. The problem of the failure to adapt sound foreign research for New Zealand circumstances persists in economics today. Last year’s Treasury report on the inclusive economy was primarily based on foreign research with a minimum of New Zealand input (the official’s own papers aside) and they also used an American as a referee. It would have been a much better report – more intellectually advanced and relevant – had it responded to the past New Zealand debate, rather than ignoring thirty years of New Zealand analysis.

The implication is that there is a need for local scientists to be working at the end of the conduit in order to understand the paradigm and adapt it for New Zealand. They may not be world class researchers, and do not need to be, although as Robb’s record shows sometimes they will end up that good. Different conduits will function in different ways. Sometimes the scientists will be based in the relevant business or public agency, sometimes they may be in a private or public research institution but contracted to do work for a business or agency, the need for the institution arising from the requirement of a critical mass of science activity. And the public agency may be a government research institution such as a Crown Research Institute or a tertiary institution such as university or polytechnic. There is no single arrangement that applies to everyone. Individual circumstances require pragmatic responses to the institutional and funding arrangements, not the one size fits all of commercialisation.

Sometimes science for creativity, capability or conduit of basic science will result in an unexpected commercial or other application which enhances economic growth. That is a serendipitous gain. While we should ensure that such potential beneficial outcomes are captured, it is important that the primary purpose – creativity, capability or conduit – remains the principal purpose.

However science also has a role of increasing our capacity to contribute to the generation of prosperity and progress. This is more than the commercial object of science, with its concerns of making a profit for private businesses and thereby, so it is argued, contributing to economic growth. That is both a too narrow an account of the growth process, and too narrow an account of how science adds to the nation’s capacity. If the Grasslands Division had been funded on a user-pays basis, our pastures would be thinner and the nation the poorer. That leaves open of how to get science to contribute to capacity, and how to get the right balance between that capability and its cultural, capability, conduit and commercialist roles (not to mention how to do so efficiently).

The government has been feeling its way to a capacity role of science, as in the recent SIAC report Innovators to the World, and the government response Growing an Innovative New Zealand. What strikes one about such reports is that they are stronger on general directions than they are on detail: one might say we are in still in the rhetoric rather than the research phase of science policy. There is too much use of the term ‘innovation’ as an undefined public good, and insufficient rigorous analysis about what it actually means in practical terms. Perhaps that is excusable in that we are still recovering from the damage that occurred when the rhetoricians exclusively defined the role of science to be a commercial one, and the seeking of a balance is commendable. However we cannot stop there. We need to think systematically about the creation, transmission and application elements of capacity science.

Actually we know a lot about how to promote scientific creativity, and quite a bit about how to apply it. It is the transmission that is problematic. For instance some of the most creative scientific cultures – Britain does well on most measures – have a poor record on the applications front as judged by the output of the economy. On the other hand there are post-war economies with poor pure science records – Germany and Japan come readily to mind – have often been among the most technically innovative.

One of the costs of the rhetoric that drove science policy in the past – by rogernomes and the scientists themselves – is that it was anti-science, as ideology must be. The result was to reduce our capacity to think systematically about science policy. It is extraordinary there is no centre for research into science activity in New Zealand. One of the indicators that we have moved out of the ideology phase is when a serious program of science policy research is instituted.

And we need to be clear that it is not simply a matter of how we organise and fund science. I was recently looking at an international comparison of post-secondary qualifications in Britain, Germany, Japan, Korea and Singapore. By far the least qualified workforce was the British, which suggests that one of the reasons British science has a poor transmission record may be to do with the quality of its workforce. Out of curiosity I added New Zealand’s record to the list. It seems that our workforce is as poorly qualified – perhaps slightly worse – than the British one. In a scientific framework this suggests hypotheses about where we might be able to improve the transmission mechanism by improving the workforce, but I would want to look more carefully at the evidence before I came to a policy conclusion. Ideologues will, of course, be able to immediately jump to their preferred policy conclusion on the basis of this one fact.

A related concern is the overall quality of the nations’s scientific understanding. Some crucial public issues are suffering as result – the genetic modification debate is an example in which blind prejudice has too dominant a role, but it is not unique. My concern is not whether New Zealanders know scientific ‘facts’, such as hydrogen is an element, although the population may be weak on them too. The concern is a lack of understanding of the scientific method. Could the anti-scientific approach of rogernomics have made as much progress as it did, if more than a handful had seen through its ideological content to its lack of scientific clothes?

So while the government’s evolving science policies seem to be in the right directions, there is still a lack of the rigorous thinking and application of research which, hopefully, is being addressed. The transmission process is failing again, is it not? Unless we add to science policy a more scientific disposition it will fail to achieve the aims the government has set itself. Invite me back in three years and we will assess how much progress and the nation has achieved. By then we should have some glimmer as to whether we have returned to the pragmatic nationbuilding described in my book, in which science played a vital role.