(introduction...)

How can these lessons be used to interpret Bangladesh's experience and push the country farther down the road to rural industrialization? Before addressing this question, some essential background on nonfarm activities is needed relating to scale composition and distribution. This information is not easy to supply because data are sparse, especially for recent years, and do not specifically demarcate rural nonfarm activities, particularly rural industry.

The most detailed recent source of information is the Annual Economic Survey of 26 nonfarm activities conducted by the Bangladesh Bureau of Statistics (BBS) in 1989-90. It estimated that about 1.5 million households were engaged part or full-time in nonfarm activities. In all, these households contributed four million workers-a tenth of the rural labor force. The nonfarm sector accounted for 13 percent of GDP. Much of the employment was in trade, commerce, and services. Education services claimed the lion's share of the employment -59 percent-and retail trade was the source of 45 percent of the value added (table 22). Other important sub-sectors in this category were wholesale trade, restaurants and hotels. About 1.5 million people were employed in some form of manufacturing, and the rural share of small-scale industries was a respectable 60 percent. Rural manufacturing in Bangladesh is almost synonymous with textile production from handlooms, which absorbs 27 percent of nonfarm labor. Also prominent are wood products, including sawmills, furniture making and bamboo cane products, and food processing, including rice and oil mills and manufactures of sweetmeats, dairy products, baked goods, flour, and gun (Table 23). Because agriculture is still largely unmechanized, repair, engineering, and metal working are limited in scale and oriented toward producing hand tools, blacksmithing, and servicing simple farm equipment.

The absence of comparable data for the mid- 1980s, makes it difficult to gauge how fast nonfarm activities have expanded or see trends in composition. A survey conducted by BSCIC (the Bangladesh Small Scale Industries Corporation) in 1978-80, provides information on cottage and small-scale industries without differentiating by sector. If it can be assumed that the rural-urban distribution of industry changed only modestly over the decade, then growth of these enterprises are indicative of the performance of rural industry. Throughout the 1980s employment in small-scale manufacturing has increased at close to 5 percent per year. Interestingly, the fastest growth was registered by metal products and machinery (9.6 percent), albeit from a small base. Employment in agricultural machinery rose 20 percent, but was still just 1 1,000 in 1990. Oil expelling mills and other food processing industries also expanded at nearly this rate. Another industry that grew sharply was ready-made garments, but this industry is almost exclusively concentrated in Dhaka and Chittagong, having very few linkages to rural 27 manufacturing Likewise, hosiery and knitted fabrics, which has also expanded capacity, remains an almost exclusively urban activity.

Overall, increase in employment since 1980 has been low (tables 24 and 25). Nevertheless, a few industries performed well even though they accounted for a small share of employment. Among these, machinery, bamboo furniture, and jewelry appear to be the most promising for future growth and exports. If urban-based garment producers 29 could be induced to integrate their production with the rural textile industry, it could help to modernize this segment of manufacturing activity, substantially enlarge its domestic market, and possibly initiate overseas sales.

As indicated above, rural nonfarm activities are influenced by the availability of infrastructure, especially transport and communications. The next section analyses the relationship between infrastructure and rural development in Bangladesh.