On development expenditure
Last week I discussed non-development expenditure by Bangladesh government. This post is going to cover development expenditure. As usual, I’ll use a set of charts – like the posts on revenue and development expenditure, the data are from the Ministry of Finance and CEIC Asia.
Typically, the discussion on development expenditure in Bangladesh goes like this: the government (and its allies in the civil society and media) says “look, record annual development programme (ADP)”, while the opposition and skeptical pundits say “you never achieve your ADP target, and you’ll fail again this year”. Who’s right?
This chart of ADP realization rate (that is, realized ADP as a percentage of budgeted ADP) shows that the skeptics might have a point.
In this chart, the columns represent annual realization rates, and the lines are averages achieved under recent finance ministers. Compared with the 1990s, realization rate seems to have been lower in the 2000s. More recently, Mr Muhith seems to have done better than Saifur Rahman, but not as good as SAMS Kibria. And Mirza Aziz has been the worst ADP implementer in the past couple of decades.
What about the government claim of the record ADP? And relatedly, perhaps lower utilization is okay because the announced programmes have been much larger in recent years?
Well, looking at just the taka figure, every year the announced ADP is the record largest. But that’s like saying every day you’re the record oldest. The economy is growing, thus just to keep pace, the development expenditure ought to grow. What we need to look at is the ADP expenditure – both budgeted and realized – as a share of GDP. This is done in the second chart.
What this chart shows is that compared with the 1990s, ADP expenditure has actually been much lower relative to GDP. While development expenditure has been rising relative to GDP in recent years, they remain far lower than the averages achieved under Saifur Rahman and SAMS Kibria.
We’ve been looking at annual expenditure, but the pattern of expenditure in any given year also paints an interesting picture. Typically, the bulk of the expenditure happens towards the end of the financial year. Part of this is unavoidable, because bills are usually paid at the end of an accounting period. But part of it is to do with bureaucratic inefficiencies – no work in the first months of the year, then a mad rush to spend the money lest funding is cut off the following year.
This chart shows monthly development expenditure as a proportion of budgeted expenditure. The grey area is the range of outcomes achieved since 2002-03. The lines represent the averages under each government. It seems that the current government hasn’t been as bad as the BNP. That average, however, masks a worrying trend, which is shown in the next chart.
Here, the lines represent each of the three financial years under this government. While 2009-10 was one of the best in recent years in terms of the monthly profile of ADP realization – steady expenditure throughout the year, 2011-12 saw a spike towards the end, while 2011-12 saw quite poor spending in the first 10 months of the year.
There is another aspect to this, which goes beyond the performance of any particular government. To the extent that a major part of development expenditure is building stuff, a late rush is inefficient not just because of lack of due diligence, but also because of the onset of monsoon. Think of the a village culvert that is completed hurriedly in May-June, only to be damaged during the July-August heavy rains. Further, the monsoon stops the work for a few months, and given bureaucratic inertia, it takes a while for things to pick up again.
Suppose the financial years ended in April, instead of June. At least then the bulk of the late rush wouldn’t be washed away. Instead, a financial year would begin with couple of months of solid work before the rains come.
Our financial years are July-June because the British did things that way. But the Raj didn’t do development. Independent India has moved to May-April financial year. How come none of the Indophiliacs in Bangladesh want to adopt that useful practice*?
*Fugstar suggested something similar in a UV comment years ago. Sadly, I don’t remember the details, and can’t link the conversation.