Peter Drysdale’s editorial reminded us of Colin Clark’s contribution to providing an empirical basis for what is sometimes called the ‘Clark–Fisher hypothesis’. This is associated with Clark’s Conditions of Economic Progress (1940) and AGB Fisher’s Economic Progress and Social Security (1945), though the idea itself had earlier origins (as Clark later acknowledged) in William Petty’s Political Arithmetick (1690). The hypothesis generalised the sequence of changes in the proportions of agriculture, manufacturing and services in employment, and in output, as development proceeds, with services becoming dominant in a maturing economy. That idea is now uncontroversial, although in a largely forgotten spat from the 1950s it came under fire from Bauer and Yamey (The Economics of Under-Developed Countries, 1957). Their unsuccessful attempt to refute this ‘law’ turns out to have unexpected value, however, in assisting our understanding of the informal economy.

Bauer and Yamey contended that, due to a combination of data deficiencies and inadequate observation, we were missing the high degree to which labour-time is actually allocated to tertiary or ‘service’ occupations during early stages of development. Note the reference to labour-time, rather than labour, for their contention was based on the time-budgets of peasants, who were invariably recorded as ‘farmers’ in official colonial returns while actually spending much time (especially in the dry season) transporting and distributing agricultural produce, as well as pursuing other off-farm economic opportunities. Thus, ‘the economic activity of many people in under-developed countries is better described as the performance of a number of different tasks than as the pursuit of a definite occupation’.

To explain this phenomenon, Bauer and Yamey pointed to the existence of ‘imperfect specialisation’ among such workers and to their ‘occupational fluidity’. The first concept originated with Adam Smith and referred to the multi-activity livelihood strategies (as described above) of cultivators and others in poor countries. The second refers to a phenomenon, also difficult to capture in the official data of 1950s colonies, which was the result of easy entry to many service tasks requiring only low levels of skill and capital for their performance. This also led to potentially-multiple economic activities being conducted by individuals. Together, these two elements were said by Bauer and Yamey to contribute to a greater allocation of labour to tertiary or ‘service’ occupations, during the early stages of development, than was anticipated by either Clark or Fisher.

It should be pointed out that the ‘tertiary’ activities identified by Bauer and Yamey were of a different order from those which Clark saw as becoming predominant with the progress of development (and to which Peter Drysdale refers in his editorial). But the character of the activities they described was consistent with their assertion that ‘tertiary’ employment, far from being the consequence of economic development, was a necessary element in the transition from subsistence agriculture to market exchange. In retrospect, the importance of their assault on Clark and Fisher did not lie in its attempt to demolish a well-established generalisation concerning growth and structural change, but rather in its pioneering account of indigenous trading and service activities. In effect, they provided the first analysis of the economics of informal economic activity, though the concept itself eluded them.

Unfortunately, Messrs Batten and Mellor made no mention of informal activity in their recent account of the service sector in Vietnam and Papua New Guinea. They noted that both have underdeveloped service economies by ASEAN standards. Thus while in ASEAN the sector’s contribution to GDP averages 45 per cent, in Vietnam the share is 35 per cent and in PNG 29 per cent. It would be interesting to know how well these numbers capture the contribution of informal service activities to gross output. The two countries differ in that Vietnam has a large and highly differentiated informal service economy while in PNG this is relatively underdeveloped. A researcher from CIRAD, a French agricultural and development research centre, has suggested that informal markets provide around 50 to 80 per cent of food supplies in Hanoi, and about the same percentage of employment in food distribution. PNG’s informal service sector is much less advanced, having struggled to develop efficient mechanisms for the distribution of domestic foods. Small ‘producer-traders’ are still dominant in many marketplaces, though the much-maligned trade in betelnut, a mild narcotic, shows the way to more professionalised marketing arrangements.

Vietnam is classified as ‘agrarian’ in the World Development Report 2013 (WDR), though it has been transformed in the last quarter century and increasingly exhibits characteristics of what the World Bank calls an ‘urbanising’ country. Batten and Mellor’s call for a larger service economy could be satisfied in part by action to formalise Vietnam’s present informal service sector, and by increasing the productivity of its activities. In the process, what the WDR called ‘successful urbanisation’ would play a part in raising productivity overall. But evidence suggests the instinct of Vietnamese authorities is to repress informality rather than deal with it constructively. CIRAD’s experience was that efforts to increase transparency around operations of the informal food economy in Hanoi had a perverse outcome: making officials better informed about it triggered ‘a mix of legalisation and repression’.

PNG on the other hand, while it has some characteristics of what the WDR calls a ‘resource-rich’ economy (it is experiencing a mineral resource boom), continues to be overwhelmingly ‘agrarian’. The large majority of its people still operate in agriculture, with a substantial continuing subsistence (that is, non-marketed) element in production. The informal economy is in general further advanced in certain rural areas whereas the urban informal service sector is relatively underdeveloped. I have suggested that what PNG needs at this stage of its development is more informal economic activity, not less. This would assist in broader diffusion of the benefits of the minerals boom among the population and, by giving more people a stake in the market economy, could reduce civil disorder and contribute to social cohesion. It would also contribute to the emergence of a national class of entrepreneurs at the SME level. Batten and Mellor quite properly stress the potential for economic reform in PNG to stimulate high-productivity, high-wage formal service sector employment. It should not be forgotten, however, that an even larger number of PNG citizens could also benefit at present from a policy environment facilitating productive self-employment in a range of the services which feature in the consumption baskets of low-income people.

John Conroy is a Visiting Fellow at the Crawford School of Public Policy, the Australian National University.