There is a similar development in the proportions of the population engaged in commercial and professional services. A certain proportion of the primary production (a proportion which is increasing) is directed also to the local market. This is particularly so in Canterbury with wheat and potatoes. Indeed, statistics show that Canterbury imports more than it exports overseas, and obtains the balance of its purchasing-power by supplying the rest of New Zealand with foodstuffs and manufactures. All these factors, however, while they render caution necessary in the use of statistics of foreign trade as an indication of progress in the Dominion, do not invalidate the use of such an index, for the following reasons :— (a.) The statistics of occupation show that the primary exporting industries still engage the largest proportion of the workers. (b.) Of the industries and occupations which are not directly concerned with exporting, a very large but not exactly measurable proportion depend directly and almost completely upon the purchasing-power made available by the exports. (c.) Research undertaken by Mr. A. H. Tocker, M.A., of Canterbury College, has shown conclusively that the policy of the banks in controlling credit advances in New Zealand for all industries varies in the closest possible manner with the balance of trade, which builds up a fund of purchasing-power in London upon which the banks work as through a “credit-exchange” standard. * “Monetary Standards in Australia and New Zealand,” Economic Journal, December, 1924. For all industries and occupations, therefore, the credit policy upon which expansion or contraction largely depends is regulated by the balance of trade, which is taken in this paper as an indication of the economic health of the Dominion. 2. The Calculation of an Index of Productivity from the Export Returns. The total figures of imports and exports by themselves do not convey any clear impression, except perhaps to expert eyes, of the changes in economic conditions which they record. The accuracy of these official figures, not perhaps in details but certainly in totals, the writer knows from experience in the Departments of Customs and Statistics to be very high. † The chief defect of the import statistics is a consistent tendency to undervaluation because of the practice of adding only 10 per cent. to invoice values of ad valorem goods in order to arrive at landed value; probably 20 per cent. would be nearer the mark. But this under-valuation does not destroy the validity of the following argument, since it is consistent from year to year; it does, however, make the balance of exports over imports appear more favourable than it really is. (N.Z. Official Year-book, 1925, p. 272; also League of Nations' Memo. on Balance of Trade Payments, 1910–23, p. 372, footnote.) The variety and nature of the Customs tariff make the production of original invoices indispensable, and the total values, as distinct from the values of particular items, must be accurate within a negligible percentage. There is less accuracy in the figures officially given for the values of exports, which, in some important cases, must be estimated here before being sold in London. But in this case, fortunately, the nature of the exports is such that the quantities exported are easily and accurately obtained. The official statistics have therefore been accepted as the basis of calculation. There is, however, some difficulty in interpretation, even when the trade statistics are compared with the corresponding figures of population, and the chief impression gained is one of constant and unvarying progress (see Diagram A). There are two main confusing factors which must be eliminated before the essential facts can be made clear—viz., the