Paper to the National Measurement Conference, 14 July, 2000.
Keywords Business & Finance
In October, the Mars Climate Orbiter spacecraft burnt up in the Martian atmosphere because the acceleration data for controlling its thrusters had been provided in pounds of force (US customary units) but entered into the space craft’s computer as newtons (the SI unit). Little information was obtained from the trip, so most of its $US240 million was a complete waste. This is a spectacular example of how measurement failure can be costly, but in some ways it is misleading. The costs of a failure to have a sound measurement system are generally more subtle than that, as are the benefits. In total, a system failure from an inefficient measurement system may be relatively more expensive than the loss of a single spacecraft.
Measurement Systems as Economic Infrastructure
In recent years economists have become aware of the crucial role that the commercial infrastructure plays in an economy. We all understand how poor roading and other transport links, or poor telecommunications, directly adds to business costs. But the intangible infrastructure plays an important role too. Without a sound system of commercial law and effective means of legal enforcement private commercial transactions become clumsy, expensive to transact, and difficult to obtain redress. These do not only add costs to business, but they also discourage entrepreneurs from seeking new opportunities which benefit consumers. Economies with high transaction costs are inefficient, whether those costs arise from inefficient transport systems or inefficient legal systems.
Another part of the intangible infrastructure, which can add or subtract to transaction costs, is the national system of measurement. Suppose there was not one. Then businesses involved in any transaction which involved measurement would have to spend time (and so money) agreeing on some common standard. (Imagine the complication for a business if each business it traded with required a different common standard.) Or suppose each supermarket had its own weights and measures. Consumers would be unable to judge the relative prices in the various chains, and so the competitive system, which drives better quality at lower prices, would be less effective.
What would happen is that there would develop a system (or systems) of private measurement, so that two businesses trading together or consumers when purchasing would use the common standard it set. Hopefully there would be only one system, in which case it would be a monopoly. Moreover in an era of increasing international trade it would have to connect with similar systems overseas. There are gains from a single international system of measurement.
In practice the international and national system have to be government ones, or at least heavily regulated b the government. To see this, consider the many transactions where measurement is important and in which the government is involved – for example excise duties on alcohol and tobacco. The statute which sets the involvement – in this case that which sets the excise duty rates – requires authoritative government set measurement standards. Otherwise the private transactor could manipulate the measurement in its own interests. A similar issue arises in international trade. If two countries do not have a common measurement standard (or a means of linking them) trade negotiations would quickly become very complicated.
Such issues have become more acute with the increasing complexity of the modern economy. This means greater anonymity between transactors, and an increasing use of technology. Today, economic transactions (trading) typically occur between people or agencies who are not well known to one another, so that the accuracy that fairness and confidence in trade requires becomes important. Often the transaction is one-off with the transactors never trading together again (as when a traveller purchases petrol from a service station in an otherwise unvisited location), but even where the transaction is repeated (the regular visit to a supermarket, say) there is normally some personal distance between those involved.
Where the technology is simple, exact measurement may not be required. Conversely as it becomes more complex, measurement accuracy becomes more significant. Consider the precision required for the hardware of modern informational technology.
Note that while measurement is often discussed in terms of length and mass, this applies to many other areas including the requirement of a nationally determined calendar and clock. Some economists would argue that the goal of price stability is equally a matter of having a constant reliable and authoritative standard of value. For a transaction to occur smoothly, the seller, receiving money for the goods or service, needs to be confident of the value of the cash received or the reliability of the credit offered. Thus, there has developed a sophisticated monetary system, even at the retail end where there is an officially guaranteed money (supported by measures to eliminate counterfeiting), together with private provisions such as banks, international credit card agencies, and credit advice agencies. One of the functions of money is, of course, as a measure of value.
A national system of measurement is not static. As technology becomes more sophisticated, precision becomes increasingly necessary, and so therefore must the common standard. As measurement extends into other technological areas, common standards must follow.
The significance of accurate measurement is well illustrated by considering transactions where it is not possible. When purchasing a mattress, say, the buyer can be sure of its size to fit the bed frame. However, on the crucial matter of the degree of hardness and softness (recall a person will spend about a third of their life lying on mattresses), there is no agreed quantitative standards. As a result the purchasing process for a mattress can be much more complicated than for a bedframe. Or consider the measurement system becomes unstable changing over time so that inter-period comparisons are difficult. A good example of this is the fierce inflation which New Zealand experienced in the two decades after 1966, when prices rises were sometimes averaging one percent a month, and were irregular so it became hard to compare infrequent purchases. The usefulness of money as measure of value was reduced.
It would, however, be wrong to confine the importance of measurement only to economic transactions. When a couple of friends agree to meet at a certain time, they are dependent upon a national convention on how to measure time. Amateur sport, indeed much recreational activity, often requires precise measurement.
In summary, good standard measures underpin much of modern human endeavour. Where they do not exist (usually because they cannot) the transactions of economic and social life may be more cumbersome. A universal standard measure can simplify those transactions adding to the efficiency of economic activities and the quality of human life.
Measuring the Measurement System
Thus far I have argued the value of a sound comprehensive measurement system in qualitative terms. Ironically it is difficult to quantify its value, as is true for other elements of the commercial infrastructure. However there are some quantitative measures which give an indication of the importance of the measurement system.
It is estimated that total trade measurement transactions in Australia – commercial transactions requiring explicit measuring – amounted to $A320b. It is tempting to compare this total with Australian GDP of around $A480b, but the measurement transactions include those which occur throughout the chain of production and distribution, and so the same item (in various stages of processing) may be measured up to seven times. An intriguing conclusion of the study was that only a fifth of the total involved final consumer transactions, so that is four fifths were between businesses or businesses and government. While the public may see the problem as the butcher with unreliable scales, measurement issues are far more pervasive.
One method of quantifying the system is to estimate the amount of resources being used in the economic activity of measurement. This is sometimes called a cost estimate, although that may be misleading because cost is generally seen as a penalty, so the measure does not indicate the benefits of metrology to the economy, since the resources are incurred because their benefits are thought to exceed their costs.
T.J. Quinn reports that ‘measurement and measurement-related operations are estimated to account for between 3 percent and 6 percent of the GDP of industrialised nations.’ If, as seems likely, the range also applies to New Zealand, then in 1999 the country would be spending between $3b and $6b on measurement and measurement related operations. Although the comparisons are not quite valid, similar size industries include tourism (a 3.4 percent contribution to GDP), agriculture (5.6 percent), construction (3.6 percent) and communications (3.0 percent).
Because there is no measure of net output, it is not possible to accurately estimate the amount of labour the measurement activity uses. But suppose net output is about half gross output, a not uncommon ratio in New Zealand industrial activity. That suggests that measurement activity employs the equivalent of around 26,000 to 52,000 full time workers (say 1 to 2 million hours per week). This, of course, is not the number of workers actually involved in the activity, since most of those who are spend only part of their time directly measuring. The number involved at some time in their working day is far larger.
One further indicator of the economic significance of measurement was that between 1970 and 1976 New Zealand, like some other countries at around this time, went through a process of metrication whereby the traditional ‘imperial system’ was replaced by the ‘international’ (or metric) system of weights and measures. The seven-year timetable was to reduce the cost of the transformation. There seems to be no estimate of the total cost of the change. Even so, one is unlikely to include the discomfort to individuals of changing the measurement system they grew up with, and some older people still have difficulty with the new-fangled measures (often not having an intuitive feel about their magnitude). One might justify the change by arguing that the metric system is intrinsically simpler and more efficient than the imperial system but, instructively, the justification for the decision to change ‘was based almost entirely on the necessity to keep in steps with overseas trading partners.’ It was international rather than domestic commerce which forced the requirement on New Zealand.
In summary, there is no doubt that the measurement activity is of considerable economic significance, whether it be assessed by:
– that which takes place at the individual work place, or
– by the value or equivalent employment in New Zealand output, or
– by the nation’s willingness to disrupt habits of a millennia to convert to a better system of measurement.
However no matter how valuable a national (and international) measurement system is, the economist is always worried about whether compliance costs – the costs of being required to comply with measurement standards set by the state – over-costly.
Firms value their reputation and even set measurement standards at a higher level than the statutory minimum. This involves costs. On the other hand failure involves costs too. It is not just a matter of a prosecution and the damage that could do to the business’s reputation among consumers. Consider the situation of a supermarket in an environment where there were no standards. Shoppers would be distrustful of the measure of their purchases, and likely to complain. The complaints would take up the shop’s time, and not be easily resolved. In contrast, one supermarket manager told me that he had but two complaints about weights and measures in three years. Each involved a person whose home scales proved inaccurate. The Trading Standards Service receives only a few hundred complaints a year – fewer than the transactions a single household might make in a fortnight. Only a fraction is substantiated. People trust the existing system.
The point of this counterfactual alternative is that it does not at all follow that were there no measurement infrastructure, transaction costs would be less. Arguably they would be more. That is the danger of quoting the 3 to 6 percent of GDP figure as the costs of measurement activity to the economy. The reality is that without incurring those costs there would be even more economic resources involved in the related activities of self-compliance, searching and litigation.
It is true that government regulation of measurement requirements can impose costs on businesses. Government policy is to minimise these compliance costs, although care has to be taken in doing so they do not raise transaction costs elsewhere. I have not come across any example of businesses considering metrological compliance requirements to be excessive. To the contrary, some supermarket chains require a higher standard than the law requires, which suggests they do not think the regulations are onerous. Even so, government policy needs to be conscious of compliance costs when reviewing or extending the regulatory framework. Ultimately, excessive compliance requirements are a burden on the efficiency of the economy, and raise prices to consumers.
The Public Understanding
While I have seen no survey, my impression is that the broad understanding of the public is that somewhere or other there is something or other which ensures that the products they buy are subject to accurate weights and measures. Their confidence is such that they rarely bother to test this belief. That supermarkets have few weights and measures complaints suggests that most of its customers do not check the weight of what they buy (for a high proportion of them must have inaccurate scales, and yet they do not complain).
Is this all the general public need to know? Perhaps having a vague confidence in the system is sufficient if they were only shoppers. However as workers, people are actively involved in measurement as a part of their jobs. There is an increasing recognition that for a worker to carry out a task properly (as is required by total quality management – TQM), he or she needs a thorough background in the fundamentals of their worktasks, and not just in its operational routines. Moreover, they meet measurement problems in their everyday non-commercial life. This suggests a need for a public understanding of metrology somewhat higher than merely having confidence there is a government backed measurement system.
There is also a wider public policy issue. While it is common to see measurement issues as being primarily of consumer interest, and only secondarily production matters issues (although the Australian evidence suggests the balance is the other way around), there is also the matter of international trade (the reason for the metrication upheaval). More insidiously, measurement criteria have been used as a non-tariff barrier to foreign competition.
Probably the most important long-term strategy is to ensure an integral part of the core curriculum of the education system includes a basing in measurement issues. Historically, students got some training in their mathematics, general science, home science or handicrafts courses. But not all students now do these, and one wonders whether in any case there was sufficient for an understanding of the increasingly sophisticated issues where metrology touches everyday life. (For instance, the increasingly inherent problem of measurement error in medical diagnosis – summarised in the notions of false positive and false negative.) Additionally, teachers of specialist subjects need to include in the performance checklist whether the course has adequately covered its measurement issues. If this sounds mundane, the history of metrology has some spectacular illustrations of its importance. A student doing advanced secondary school physics should revel in the metrology of the discovery of relativity; a cooking class has the opportunity to discuss the consequences of failures in recipe specification and interpretation; measurement in sport provides some examples, with riveting videos to illustrate them.
Issues for Further Consideration
1. Should the remit of the Trade Measurement Unit in the Trading Standards Services of the Ministry of Consumer Affairs explicitly recognise metrology’s importance to the external sector of the economy?
2. Should the government should review whether its role in the International Organisation for Legal Metrology (OIML) should be upgraded from corresponding to full membership, given the increasing importance of metrology in food, health, and environmental issues (and their impacts on key elements of New Zealand international trade)?
3. While there is no evidence of an unnecessary burden of compliance in regard to metrological issues is onerous, should the Trading Standards Service explicitly acknowledge that one of its functions is to reduce compliance costs, providing they do not add to transaction costs elsewhere in the economy?
4. Is there a need for raising public awareness of measurement issues in the long run? The program could include:
– ensuring that key formers of public opinion are aware of metrological issues;
– using news events which involve measurement issues to provide a background of the wider issues;
– ensuring there is adequate coverage of measurement issues in the core primary and secondary curriculum;
– expecting all specialist courses at secondary and tertiary (including vocational) levels to cover systematically the relevant metrological issues of the subject;
– approaching NZQA to assist in attaining the previous two objectives;
– providing the public with material to assist with these objectives.
Conclusions: New Directions in Metrology – The Economists’ Perspective
Transaction costs have been described as the friction in the economic system. Because of these costs the system operates less efficiently than it would in a world without transaction costs. One of the functions of government policy is to reduce that friction, to – in effect – put some oil into the machinery so it runs more smoothly.
Measurement is one of the sources of transaction costs. The metrological infrastructure is a means of reducing those costs, of putting some oil into the system to reduce the friction that would arise from inaccuracy and inconsistency of measurements. It is not possible to evaluate the contribution of the engineer which supplies the lubricating oil, nor the oil itself. Clearly, the contribution is not the same thing as the cost. Rather, we can imagine what would happen were there no lubrication. At best the machinery might come to a grinding halt. The users would hurry to find some oil, but because it was not designed for the particular use and because they were not trained in its application, the machine would still function poorly.
Metrologists are among those who design and apply some of the lubricating oils for the economic system, and ensure that it is applied as effectively as possible. Their value, like that of the engineers with machinery oil, is incalculable. But nonetheless, the value of a properly organised system of international measurement is palpable.
1. Committee of Inquiry into Australia’s Standards and Conformance Infrastructure (1995).
2. Quinn (1994).
3. The quotation comes from NZOYB (1982:854). (Despite the claim of efficiency of the metric system, we should acknowledge the linguistic merit of our forefathers’ simpler ‘pounds’, ‘feet’, and ‘miles’.)