In the early 198Os, barter arrangements among developing countries and between them and industrial countries began to attract attention. By the mid-1980s, the estimates of its share in world trade were escalating, with even specialists and normally careful international institutions agreeing on lO-15%, although the IMF remained more cautious at about 1%. These estimates, unlike the studies quoted later in this chapter, appear to have been based on little more than other estimates (plus a little for growth since the previous estimate), and a general feeling that if it was attracting attention, it must be large. There was little precision about whether the estimates were of new arrangements or included a large proportion of old ones, now newly given a common name. Although traditionally barter trade was a way of trading with and among the non-market, centrally planned economies, the new flows were considered to be largely by, and with, the developing countries. Any increase in barter’s share in total world trade approaching these estimates would have meant a significant proportion of their trade (10% of world trade in 1985 equalled 35% of their exports). This would have been much larger than the estimates discussed in Chapters 2 and 3 for such constraints as preferences or NTBs. Even if initially any increase had come as the result of individual countries’ own policies, a share as high as this in total developing country trade would have been a major alteration in the trading system for all.