Kenya wants more money to finance infrastructure, but can it actually spend it?

Saturday June 14 2014

The recent hype about the sovereign bond and the launch of the standard gauge railway are manifestations of the orientation of Kenya’s current government towards large infrastructure projects as key drivers of development.

This continues a discourse popularised under the Kibaki administration that the main constraint on growth is a paucity of capital investment. It is also linked to current efforts to tame the recurrent side of the budget by better managing the public wage bill.

As it is budget season, one of the questions we should be asking about the government’s budget proposal is: Has the development budget grown in line with this discourse?

A recent story in the Daily Nation suggested that the development budget is slated to rise from 10.3 per cent of GDP to 12.1 per cent. No timeline was provided, though we might assume this is a medium-term target (three years).

But what is the source of these figures?

The Budget Policy Statement has two different figures for the development budget: Ksh444 billion, which is said to be 10.7 per cent of GDP, and Ksh442 billion, which is said to be 9.5 per cent of GDP. The latter figure matches Annex Table 2. This table also shows a development figure of 10.6 per cent of GDP for 2016/17, meaning that, even over the medium term, the government is not expecting to increase development spending to 12.1 per cent of GDP.


The erratic figures continue in the actual budget tabled in April.

The budget summary shows that development spending will be just shy of Ksh480 billion. This means development spending would be 10.3 per cent of GDP in 2014/15. The text, however, refers to a higher figure of Ksh486.9 billion (10.5 per cent of GDP). Another table in the summary refers to yet a third figure of Ksh471.4 billion (10.2 per cent of GDP). Finally, the budget itself (line-item version) has still another figure: Ksh476.4 billion (just under 10.3 per cent of GDP).

These inconsistencies are small, but suggest some lack of seriousness in budget presentation. Even if the differences can be explained by minor differences in calculation (such as including net lending in some figures but not others), this is not at all clear from the documents.

Using these numbers, the development budget is increasing at most about 8 per cent total over the past two financial years (Ksh453 billion in 2012/13 to Ksh487 billion in 2014/15). Not trivial, but not particularly striking (and below inflation).

More important, development budgets appear to be systematically overestimated every year.

In 2012/13, the government estimated that it would spend Ksh453 billion on development, but actually spent Ksh199 billion (44 per cent of original budget).

In 2013/14, the government estimated that it would spend Ksh439 billion (according to BPS 2014; the budget books show an initial estimate of nearly Ksh447 billion). In the first half of the year, the government spent 16 per cent of this. After three quarters, it had spent Ksh207 billion, or about 46 per cent of initial budget.

This suggests that absorption of the development budget in 2013/14 may be better than 2012/13, but it is almost certainly still overestimated.

Where has development spending fallen short in 2013/14?

To be fair to the government, we should compare spending against target by quarter, rather than against the full year, because targets by quarter reflect more closely how government project cycles work. If we do this for the full development budget, third quarter performance was 59 per cent of target.

Unfortunately, several of the ministries that lagged behind the most are those providing infrastructure that is core to the discourse on capital investment: energy (26 per cent) and environment and water (44 per cent). Transport (68 per cent) performed somewhat better. Other key ministries also performed poorly, such as education (21 per cent) and health (45 per cent) and even Treasury itself (17 per cent).

At the same time, recurrent expenditure appears to be more realistically estimated. In 2012/13, the government spent 94 per cent of its original recurrent budget. By the end of the third quarter of 2013/14, it had spent 82 per cent of initial recurrent budget.

The net effect of all of this, of course, is that actual spending in Kenya is more recurrent than it appears in the initial budget.

Part of the problem may be that certain agencies do not report appropriations-in-aid properly, but even the government acknowledges in the Budget Policy Statement that absorption of grants and loans from donors has been the major issue on the development side of the budget. Poor projections of what is actually available are part of the problem.

So what should we make of all this? The 2014/15 development budget does not suggest a massive increase in capital spending. The actual performance of development expenditure has been far below budget in recent years.

If government purports to ramp up development spending using new sources of finance, is it certain that it can absorb the available funds?

Jason Lakin is senior programme officer and research fellow at the International Budget Partnership/ E-mail: [email protected]