Evidence from Tanzania
Poverty and prosperity are multi-dimensional. Using single indicators of poverty will produce distorted pictures of well-being. We show that, in rural Tanzania, one of the most common measures of poverty can systematically exclude important forms of wealth. We also show that, while the perspective change in assets offers is useful, assets are in themselves not a good general indicator. We need to add their story to the picture and bring in other elements too.
One of the clearest trends in wealth and poverty in Tanzania is shown below: national economic growth is considerable, but rates of rural poverty decline are slower.
This gap between economic growth and poverty decline poses awkward questions, as Mashindano and Shepherd observe
‘Why has rapid growth in Tanzania not been accompanied by a corresponding fall in poverty? Why have the numbers of impoverished risen?’ (page 3)
Most explanations focus on the fact that the sectors which are growing – infrastructure, tourism, mining and service industries – are not rural. Rural areas depend on agriculture which is perceived to be relatively unproductive, changes slowly, suffers from extensive backwardness and risk averseness.
But this is an unsatisfactory explanation. It cannot explain inconsistencies in the data. The same data that report slow declines in poverty also show a sustained increase in the quality of housing in the country.
How is it possible for a persistently poor rural population to construct such (relatively) good houses? Some suggest that remittances play a role. But expenditure from remittances should show up in household budget surveys. Perhaps the paradox is not the failure of economic growth to make rural people wealthy. Rather it is why do poverty line data not capture Tanzanian’s ability to build relatively good houses?
If we are more sceptical about the data then the paradox diminishes. First, we have to be cautious about data on agricultural productivity given how hard it is to tell what Tanzanian farmers grow and what they do with their harvests. Second there are restrictions of poverty line data based on measures of consumption. These data exclude outliers – isolated large items of expenditure – that would make poor families look wealthy. They are unlikely to capture unusual expensive items (such as metal roofs on houses).
But, in addition to that, these measures also systematically exclude all investment in productive assets. So whenever a farmer invests in her herd, or land, or buys inputs such as fertiliser, improved seeds or a plough – all these are missed out of the expenditure records used to calculate poverty lines.
There are good reasons for these omissions. If you are a shopkeeper and spend $500 a week on goods for your shop then that expenditure is plainly not part of your household budget. We cannot confuse the turnover of a business with household consumption. By the same logic we exclude expenditure by smallholders on productive assets for their farms.
But this is a problem. For most rural Tanzanians their prosperity is best indicated by precisely these assets. Assets provide income streams in the future and security from shocks. As Andrew Shepherd put it the ‘[k]ey to success in agriculture is accumulating assets – land, oxen, and ploughs’ (page viii). So where rural populations are getting richer, and investing in new assets, then poverty line data will, unfortunately, systematically miss these changes.
You can see this clearly in the schedule of ‘COICOP’ codes used in Tanzania’s 2011/12 Household Budget Survey. This provides an exhaustive list of all different types of expenditure recorded in a consumption survey, including food, clothing, house costs, furniture, education, water, electricity, insurance and even money spent on prostitution. But it does not include any purchase of land. Nor does it allow for the purchase of ploughs, power tillers or tractors. Livestock purchase is only recorded if the animal is used for meat. Veterinary services are only recorded for pets and fertiliser for gardens, not farms. It is impossible to mention investment in productive assets because there are no codes for them.
Our research explores long term trends in rural Tanzania based on asset data. We have recently published in three open access papers in the Journal of Peasant Studies (as well as other studies available on this page). This method is difficult (as this paper discusses), yet important because assets matter so much to rural Tanzanians (as this paper shows). We have found numerous cases – although not in every case – of families prospering in ways that poverty line data could not capture. As this paper shows, in Rukwa, we found people are farming larger areas of land, growing more cash crops (sesame), and investing the returns in assets (shown in the table below). In Dodoma we found that people were building better houses (shown below) after farming larger areas of land for sunflowers. There are also cases in which assets have not flourished through agriculture. As Anna Mdee’s research shows, in such cases only remittances and rising land values can sustain asset gains. But the general trend is clear: tracking assets provides important additional insights that complement poverty line data.
It is useful to track assets where rural populations are becoming more wealthy. Prosperity dynamics are visible in changing asset portfolios. But note these cautions. First, where people are getting poorer then assets are a poor indicator of stress. Rural people sell them reluctantly. Second, growth in assets does not mean that rural Tanzanians are ‘really’ wealthy. Poverty, and prosperity, are multi-dimensional. Saving for assets is hard, as our interviewees explained: ‘tumebana matumizi’ – we tightened our belts.
And neither consumption nor assets are not necessarily good proxies for morbidity, mortality, education, gender relations and so on. Assets provide an important complication in our understanding of change in rural societies. Where poverty in terms of consumption persists, or grows (see Jason Hickel) then this is no less real for any accompanying growth in assets.
Nonetheless asset data to allow us to look at Tanzanian rural economies afresh. Commonly-used statistics are likely to underestimate these farmers’ contributions to national change and local development. We need to look at assets to see how smallholders create wealth, and the forms of consumption that their activities allow.