Most recently, "Rural livelihoods systems in Bangladesh: Changes and challenges" a forthcoming book (by Mahabub Hossain and co-authors) contains an academic as well as empirical analysis on the role of roads and electricity in the reduction of rural poverty in Bangladesh. Based on a data set generated by IRRI/BIDS and IRRI/PETRRA projects covering 62 villages, the authors tend to show how roads and electricity could become gateways to poverty reduction. Two points of time have been compared, 1987 and 2000.

In the cited book, villages have been categorised as: (a) 'developed villages' -- having access to both paved roads and electricity; (b) 'semi-developed villages' with an access to either paved roads or electricity and (c) 'under-developed villages'-- having been deprived of both.

Access to roads and electricity

During the 1987 survey, out of 62 villages surveyed, nearly one-tenth was found to have access to both electricity and paved roads (developed villages). By 2000, the share of developed villages stood at little over one-fourth. It seems that the share of developed villages in rural Bangladesh increased in tandem with the development in communication and rural electrification over the years. The change over time might sound reasonable, if not remarkable, in the light of developmental objectives. However, the proportion of sample villages with access either to electricity or paved roads (semi-developed villages) seems to have marginally declined implying that some of the semi-developed villages of 1987 have graduated as developed villages in 2000. On the other hand, the share of the villages without access to either electricity or paved roads (under developed villages) substantially declined over the same period indicating that a large part of the underdeveloped villages of 1987 graduated as developed villages in 2000. The trends are not unlikely events given the fact that between 1987 and 2000, Bangladesh witnessed appreciable growth in rural roads and electrification.

The differential access to roads and electricity resulted in differential economic performances and allow me to submit few of them in the following paragraphs.

On crop production

During 2000 survey, cropping intensity is estimated to be the highest in developed villages (179) compared to the same lower level (149) in both semi- and under-developed villages. Noticeably, just the reverse had happened in 1987: lowest at 159 in developed villages compared to semi developed (162) and under developed villages (173). Thus, it appears that, with increasing access to paved roads and electricity, developed villages stole the lead in cropping intensity while semi and under developed villages faced a fall in cropping intensity. Cropping intensity has relation with research and extension, marketing opportunities and processing activities. It is no surprise that with increasing access to paved roads and electricity, developed villages would perform better.

Developed villages retained the lead in terms of irrigation intensity -- the share of irrigated area to total cultivated area -- both in1987 and 2000. For example, about four-fifths of the cultivated area in developed villages was covered by irrigation in 2000 compared to 40 percent in 1987. However, irrigation coverage in semi- and under-developed villages also increased over time but that was far away from a respectable limit achieved by developed villages.

And finally, rice yield. All the sample villages witnessed a reasonable rise in rice yield between 1987 and 2000. However, in the case of developed villages, the increase was to the tune of 69 percent over the period compared to semi-developed villages (38 per cent) and under developed villages (36 per cent). This seemingly implies that rice productivity growth is positively associated with the development of rural infrastructure, especially roads and electricity. And one of the causal factors in this case could be the increased irrigation intensity and more areas under modern varieties of rice induced by paved roads and electricity.

By and large, the cause of apparent positive correlation between infrastructure and crop production is not, perhaps, difficult to detect. The development of infrastructure helps lower the marginal costs of production by shifting the marginal cost curve to the right. The supply curve shifts to the right when farmers face a fall in input prices and a rise in output prices. Both paved roads and electricity tend to contribute towards the fulfilment of the twin objectives.

On capital accumulation

Households in developed villages seem to own more agricultural fixed assets than semi- or under-developed villages. Noticeably, accumulation of agricultural fixed asset was negative for developed villages while semi and under-developed villages marginally gained over the periods. Besides infrastructure, some other factors might have contributed to narrowing down the gap over time.

The striking difference appears when one considers accumulation of non-agricultural fixed assets. Accumulation of non-agricultural fixed assets appears to be an increasing function of the development of infrastructure. Developed households possessed 1.3 times more fixed assets than semi developed villages in 1987 while the same households held 2.8 times more in 2000. Likewise, developed households had1.5 times more mixed assets than under developed households in 1987 while they held 4.5 times more in 2000. Thus, not only developed households endowed themselves with the accumulation of non-agricultural fixed assets at particular point of time, but also maintained a higher growth rate compared to others over the periods. In fact, roads and electricity tend to contribute more to non-farm activities and hence it is not unlikely that households with better access to these facilities would gain most in terms of accumulation of non-agricultural assets.

Interestingly, only three per cent of developed households had access to financial institutions in 1987 compared to 15 percent and 18 percent, respectively, by semi- and under-developed households. But by 2000, the share jumped to 29 percent for developed villages but feebly for others. Similarly, in 1987, institutional credit per household in developed villages stood at US$ 5 compared to US$ 15 and US$ 18 respectively for semi- and under- developed households. By 2000, all households witnessed an increase but faster improvements followed infrastructure. The phenomenon could be explained by the fact that it takes time for infrastructure to impinge positive effects and it is not surprising that given a longer time, they surpass all.

Productivity of assets

For 2000, per capita income estimated for developed households stood at US$ 309 compared to US$ 226 and US$ 202, respectively, for semi- and under-developed households. The per capita income of developed households is 1.4 times that of semi-developed households and 1.5 times that of under-developed households. In 1987, developed households had not had this edge over others. In fact, the per capita income of households in developed households was lower than semi-developed households and very close to under-developed households. It seems that, over time, developed households witnessed a growth rate of 7 percent per annum compared to that by semi-developed (about 3 percent) and under-developed (about 4 percent) households. The difference also widened in terms of total income per capita, and possibly, more in the case of non-agricultural income per non-agricultural worker.

On occupational mobility

With increasing access to paved roads and electricity, occupational mobility is evident from cultivation and agricultural wage labour to trade and business and to non-agricultural labour. The findings seem to be in consort with that observed in other countries, especially in India and China. When multiple occupations are taken into consideration, we observe that the degree of multiple occupations is relatively low in developed villages. In fact, it increases with under-development of infrastructure. The reasons are, perhaps, not far to seek. Paved roads and electricity drive households more towards specialisation and risk minimisation. Further, both mobility and malleability are likely to rise with the development of infrastructure like paved roads and electricity.

On poverty

It could be found that access to paved roads as well as to electricity contributed to a larger reduction in poverty in sample villages. In 1987, 33 percent of the households from infrastructurally developed villages were extreme poor compared to 31 percent and 27 percent, respectively, in semi and under-developed villages. By 2000, all villages witnessed a decline in extreme poverty but the pace of decline was much faster in developed than semi-developed and under-developed villages. For example, between 1987 and 2000, extreme poverty in developed villages fell by nearly two percentage points per year as compared to one percentage point in semi-developed villages and by half a percentage point in under developed villages. The similar kind of situation prevailed in the case of moderate poor. The poverty-gap index and squared poverty-gap index also seem to tell the same story.

On inequality

While poverty, reportedly, reduced in developed villages, inequality of income increased. In fact, inequality of income rose in all villages in 2000 but the gini ratio was much higher in developed villages (0.496) compared to semi (0.443) and under-developed villages (0.398). In 1987, the same pattern prevailed. Seemingly it sounds that inequality in income is an increasing function of infrstructural development. This could be true since sources of household income tend to change with access to paved roads and electricity -- say from crop production to business and trade that require more capital and education. Just take the case of rural trade and business. In 2000, in developed villages, this source accounted for 40 percent of the total gini ratio compared to only 13 per cent in 1987. This means that 40 per cent of the inequality originated from trade and business activities. In semi-developed villages, trade and business also accounted for a higher contribution to inequality but not so much as developed villages. Quite in contrast, however, under-developed villages witnessed a fall in the role of trade and business. By and large, most of the incremental inequality in developed villages seems to be linked to non-agricultural pursuits.

Policy palliatives

We were, perhaps, late in realising that rural roads and electrification could reduce rural poverty. At the time of writing this piece, I suppose, more than three-fourths of our rural households remain deprived of paved roads and reel under darkness. The research findings from Mahabub Hossain and co-authors -- based on a representative household level data -- should ring a bell for our planners and policy makers about the urgency of developing rural infrastructure like roads and electricity. As I mentioned in some of the earlier writings, one of the means to combat rural-urban migration and rural poverty is to take the urban environment to rural areas. Policies should not be populist such as "pro-poor" but pro-rural in spirit and action. Development of rural infrastructure is gateway to poverty reduction.

Inequality? That's likely to remain for a while. To contain inequality, access to education, credit and social services should be raised for lower income deciles. There should be human and financial capital available to the lower segment. If we fail to do that, we fail to contain inequality of income. Mind that inequality also increased over time in under-developed villages. But barring rural roads and electricity, possibly, we shall have to live with twin evils: poverty as well as inequality.

Abdul Bayes is professor of economics, Jahangirnagar University