Published in the Daily Star on 13 June 2010.
This piece analyses the size and implementation of the annual development budget.
THE 2010-11 (FY11) budget envisages an Annual Development Program (ADP) of Tk.38,500 crore. Some media commentaries have focussed on its record size, without recognising the fact that both our economy and population are larger than ever before. To note the record size of ADP in isolation is like saying I am older today than at any time in my life!
Meanwhile, others have criticised the ADP for being too big or too ambitious. Pointing to the facts that only once in the past decade has an ADP been fully implemented and the implementation rate has been particularly poor from FY02 to FY09 (Chart 1), many commentators cynically predict that whatever the announced size, the actual ADP will likely fall short of the target.
Is the ADP too big? Let’s explore this question with a couple of charts.
(Top Left) Chart 1: ADP implementation rate (actual as percentage of original ADP) Source: Ministry of Finance, Bangladesh Bureau of Statistics, author’s calculation. Chart 2: ADP as percentage of GDP Source: Ministry of Finance, Bangladesh Bureau of Statistics, author’s calculation. (Bottom left) Chart 3: ADP per capita The dotted lines represent the trend path achieved between FY92 and FY01. Source: Ministry of Finance, UN Population Database, author’s calculation. (Bottom right) Chart 4: Monthly ADP implementation (as percentage of annual originalADP) Source: Ministry of Finance, author’s calculation.
Chart 2 presents original (that is budgeted) and actual (that is implemented) ADP since FY92 as percentage of GDP. Averages of these ratios between FY92 and FY01 are also presented.
Between FY92 and FY01, original ADP was on average 6.8% of GDP, while the actual ADP used to average 6% of GDP. In the past decade, both original and actual ADP had shrunk relative to GDP. In FY09, original ADP was 4.2% of GDP, while the actual ADP was 1 percentage point less. Under the current government, original ADP has risen relative to GDP. But at 4.9% of GDP, it is still far below what was regularly implemented in the 1990s.
Is the ADP too big? Not compared to the size of the economy.
Another way to compare the size of the ADP is to analyse it in per capita term. This is done in Chart 3. Our measure of the population is the UN Population Database medium variant estimate, which puts the country’s population at 164.4 million in 2010. In per capita terms, original ADP had risen consistently since FY92, except for FY08 and FY09, when it dipped from the trend (for both FY92-FY01 and FY92-FY07 periods). In FY10, ADP per capita merely returned to what would have been the case had the past trends been maintained. Meanwhile, in the last decade, actual ADP per capita has consistently slipped below the trend achieved in the 1990s.
The budgeted ADP for FY11, compared against the past trend, does seem big, particularly given the gaping hole between the budgeted and implemented ADP in recent years. But there is more to this picture than meets the eye. In per capita terms, the FY11 ADP as budgeted is 2,310 taka. Suppose the 1990s trend in the implemented ADP had continued throughout the past decade — the realised ADP in FY11 would have been 1840 taka. If the government implements 1,840 taka per capita of ADP in FY11, it will have implemented only 80% of the announced ADP.
Is it an ambitious ADP, or a conservative assessment of the government’s own realisation capacity? After all, between FY02 and FY09, only about 77% of announced ADP had been implemented on average a year.
Could it be that the government is adopting an approach where a large ADP is announced, with promises made to various constituencies, lobbyists, and supplicants, all the while knowing that the marginal projects will be discarded in the implementation phase? Could this be a “second best” policy outcome — the “best practice” of selecting and fully implementing projects based on a meritocratic criteria being unavailable due to political economy reasons?
If this “second best” approach is what the government is following, then the focus on the implementation rate might be misplaced. Indeed, in this scenario, if the government feels compelled to fully implement the announced ADP, there is a risk that large sums of money will be spent on poor quality projects towards the end of the financial year.
This is illustrated in Chart 4. In this chart, the shaded area represents the range of monthly ADP implementation expressed as a percentage announced ADP over the past five financial years. Typically, there is little implementation at the beginning of a financial year, while the month of June sees a bulk of sums spent (a reflection of end of year settling of accounts).
Ideally, ADP would be implemented in a way such that monies are spent steadily in the autumn, winter and spring months — this pattern seemed to have been holding for FY10 until January. Against this, a poor outcome is when relatively little is spent in the post-monsoon months, and by the third quarter of the year there is a lot of commentary on how ADP remains underutilised, causing a wasteful spending spree in the final two months of the year — this is essentially what happened in FY08.
Instead of debating whether the ADP is too big or too ambitious, or calling for the “best practice” project design that might be unattainable in our weak political and institutional culture, it would be better if attention is focussed on making sure that right projects are implemented at the right time.