Day 2: Country Session – Uganda

The session was jointly hosted by Drs. Louis Kasekende (Deputy Governor of the Bank of Uganda) and Richard Newfarmer (Country Director of IGC Uganda).

The first paper, on poverty dynamics, was presented by Prof. Andy McKay (University of Sussex) – joint work with Dr. Sarah Ssewanyana (EPRC Uganda) and Dr. Marguerite Duponchel (IGC Uganda). Prof. McKay began by outlining the recent economic history of Uganda and highlighted that increases in living standards have not been progressing since the financial crisis and its economic fallout. Lower growth rates have partly been the result of frequent droughts.

The data used in the study comes from the Uganda Bureau of Statistics and is a four wave panel survey conducted between 2005/06 and 2011/12 with 3,000 housholds (HH). The key measure of poverty in this context is consumption data. In general terms, the North is the poorest region and the East of Uganda has the highest absolute number of poor. According to the panel, poverty numbers were up in all regions except Central, with the highest increase in the East.

In terms of the effects impacting poverty numbers, the availability of community infrastructure such as health facilities is correlated with lower poverty figures. Other factors reducing poverty are education and credit access, whereas drinking alcohol and being engaged in agricultural labour are associated with higher poverty. A major barrier to poverty reduction is the high dependency ratio.
Dr. Albert Musisi (Ugandan Ministry of Finance, Planning and Economic Development) was the discussant, presenting results of the government’s own poverty analysis. Strikingly, those figures show a sustained reduction in poverty that is quite at odds with the largely stable or rising overall poverty evidence emerging from the panel data set. Income inequality and the panel nature of the data may be two reasons for the divergence of results. He concurred though that poverty is a serious issue and that skills and infrastructure are good poverty reduction measures.

Marguerite Duponchel then looked at household survey evidence of a gender gap in agricultural productivity. The data is for 630 HH and around 7000 plots of which around half are female-managed. With her co-authors from the World Bank she found that there is a gender gap of around 17.5% without controls and 37.7% when controlling for inputs, land quality etc. These are the results of an Oaxaca-Blinder decomposition approach. In terms of the policy implications, child care constraints are an issue that could be addressed with low cost options. Also, the promotion of high value crops for females and bringing extension closer as transport costs are high would be promising approaches. Inputs apparently don’t matter so much in this context.

Instead of discussing the paper, Prof. Jakob Svensson (Stockholm University and IGC Uganda Lead Academic) then presented preliminary results of his research. Interestingly, these showed that all the fertiliser samples analysed contained less nutrients than necessary for good effects, on average 40% less so. If good fertilisers were used, triple the yield could be achieved, but actually given the quality and their expectations of the same it turns out that farmers’ non-adoption of this input is a rational choice.

Prof. David Bevan (Oxford University) then delivered the third and final presentation of the day on joint work with Prof. Chris Adam (IGC Tanzania Lead Academic and Oxford University). The core of the paper concerns a general equilibrium model initially developed by the International Monetary Fund to analyse debt sustainability in a dynamic context rather than the more static approach of traditional debt sustainability analyses. The model has been extended to take account of tax distortions, maintenance and operations expenditure and incomplete appropriability of public capital investment returns. It is calibrated to Uganda and draws on inputs from a wide variety of sources.

The results of the model were at an early stage and expressed in graph form, but some striking results stand out. Especially, inefficiency in public investment that results in less than a dollar’s worth of capital for a dollar invested is a major issue that lowers long term outcomes significantly. Compared with no such inefficiency, a 20% level of inefficiency lowers GDP by about one third over a thirty year time horizon. The results also showed that a very large increase in tax take would be necessary to fund investment plans, suggesting in practice that increased external borrowing will be necessary to finance current spending plans. The next steps will be to iron out the last kinks of the model. The authorities will then be invited to give their views on desired modelling scenarios before final simulations are run.

By Tim Ohlenburg, Country Economist, IGC Uganda