Wednesday, June 7, 2017

Bloated budget, challenging implementation

It was published in The Kathmandu Post, 5 June 2017. Here is an earlier blog post on FY2018 budget (same in NEF blog). The latest issue of Himal magazine includes an analysis on the same in Nepali language. 

Bloated budget

Fiscal transfer to local bodies warranted a bigger budget, but its implementation will be challenging

On May 29, Finance Minister Krishna Bahadur Mahara presented the budget for fiscal year (FY) 2017-18 to Parliament. The projects and programmes are pretty much a continuation of last year’s budget, since the election code of conduct barred the government from rolling out new initiatives that could influence the outcome of the second phase of the local elections.

Some of the notable features of the budget are its focus on the devolution of spending authority to local bodies, the progress of, and adequate funds for, large-scale infrastructure projects, and post-earthquake reconstruction.

Macro overview

The total expenditure outlay is capped at Rs1.3 trillion, which is equivalent to an estimated 45.5 percent of gross national product (GDP) in FY2018. This is an increase of 36.7 percent over the revised expenditure estimate for FY2017. The government argued that the increase in the size of the budget is justified owing to the fiscal transfers to local bodies in the federal setup. These transfers are clubbed under recurrent spending, which constitutes 62.8 percent of the total budget.

Meanwhile, 26.2 percent of the budget is earmarked for capital spending, which includes public expenditure on new buildings, bridges, roads and civil works, among others. The government is planning to meet 64 percent of expenditure from tax and non-tax revenue, foreign grants and principal repayment.

This leaves a budget deficit of NRs461.3 billion, which the government expects to cover by a combination of foreign loans, domestic borrowing and cash balance from FY2017.

Unrealistic targets

At the aggregate level, three important issues in the budget require scrutiny. First, GDP and revenue targets of 7.2 percent and 25.7 percent respectively are ambitious. The economy grew by an estimated 6.9 percent in FY2017 due to a low base effect, favourable monsoon, improved power supply and normalisation of supplies after two years of disruption. Achieving 7.2 percent growth from such a high base would be challenging, because the usual drivers of growth have to be much stronger than last year. 

Even with good monsoon rains, continued improvement in power supply, acceleration of reconstruction works and capital spending and election-related expenses, GDP growth may be restricted between 5 and 6 percent. Furthermore, expecting revenue growth higher than in FY2017 with pretty much the same revenue policy is a bit of a stretch. FY2017 revenue was higher because of the normalisation of imports, which faced an unexpected dip following the trade embargo in FY2016. The expectation that tax revenue from trade will be the same as in FY2017 (47.5 percent of total tax revenue) is a bit misplaced. 

Second, since the likelihood of meeting revenue growth is pretty slim, the government will aggressively engage in domestic borrowing, which is expected to be about 4.3 percent of GDP. This will undoubtedly put upward pressure on retail interest rates, which will increase the borrowing cost of households and businesses, as deposit growth is still sluggish (amidst higher credit growth) due to the deceleration of remittance inflows and the government’s inability to spend the mobilised revenue on time. A large deficit financing also tends to exert inflationary pressures.

Third, recurrent expenditure is ballooning too fast and needs to be contained. It is expected to be a whopping 110 percent of revenue in FY2018. The previous government sharply increased recurrent budget by inserting all sorts of pet projects and distributive programmes in the ‘red book’, which is a catalogue of all approved projects and programmes. It would be difficult to discontinue such projects once they make their way into the red book and hence this government had no option to exclude them. This, combined with the need for large unconditional fiscal transfer to local bodies, led to an almost 43 percent increase in recurrent spending. It is now time to curb recurrent spending in real terms and cull unproductive, repetitive and “zombie” projects to maintain fiscal prudence.

Execution challenges

In the absence of a Natural Resource and Fiscal Commission to oversee fiscal transfer to local bodies, this budget allocated transfers based on a number of factors such as population, development status and cost of operation. Fiscal transfer and conditional grants to local bodies now constitute about 50 percent of total recurrent spending. This is unprecedented in the sense that the transfers will now be directly credited to the bank accounts of local bodies—bypassing the maze of bureaucratic approvals needed at Singh Durbar—and the local bodies will have the authority to prioritise, design and implement their own projects. These include programmes and projects on agriculture, livestock, irrigation, roads, drinking water and sanitation, schools (including payment for teachers), tourism and sports, among others. 

Despite such large transfers and early spending authority to line ministries, execution of the budget will be challenging. The budget mandates project offices to award contract by mid-October to accelerate capital spending, which averaged 75 percent of planned budget in the last five years. This particular reform addresses expenditure delays attributable to lengthy and repetitive approvals needed at various layers of ministries and the National Planning Commission.

However, other issues contributing to chronically low spending pattern such as lack of project readiness, high staff turnover and poor contract management will need to be addressed swiftly to accelerate capital spending. Furthermore, most of the local bodies do not have the capacity to design, assess, appraise and execute projects at the local level. Hence, an unconditional fiscal transfer to local bodies could be an asset or a liability. If spent in the right way and with proper accountability, it could be an asset and a significant contributor to economic development.

However, with weak oversight, lack of capacity to plan and execute projects and misappropriation of funds, it could become a liability and eventually contribute to fiscal stress. The centre needs to be careful in providing handholding support to local bodies to address execution issues, but it should not try to micromanage projects as in the past.

Overall, the budget gives a mixed picture: bloated in size justified by the compulsion to continue previous projects and direct fiscal transfer to local bodies; ambitious GDP and revenue growth targets; large deficit financing, which would potentially increase interest rates; faster spending authorisation by cutting down lengthy and repetitive approvals; and unsustainable pattern as well as size of recurrent budget. The nature of this budget would also mean that there is no room and justification for a supplementary budget after the local elections.

Thursday, June 1, 2017

Quick thoughts on Nepal’s FY2018 budget

Here are my quick thoughts on the FY2018 budget.

Deputy Prime Minister and Minister of Finance Krishna Bahadur Mahara presented FY2018 budget (mid-July 2017 to mid-July 2018) to the parliament on 29 May. The budget focuses on implementation of federalism (particularly at this stage, providing initial financing to local bodies), post-earthquake reconstruction, continuation of previous programs and policies (because the government cannot roll out distributive programs and pet projects during the election time), and ensuring adequate funds for large infrastructure projects.

The budget is unique in three ways: 
  • First, it tries to bypass the usual Singha Durbar related procedural hassles by directly transferring funds to local bodies so that they can initiate institutional setup and some local level projects. Local bodies are supposed to get NRs225.1 billion and provinces are supposed to get NRs7.1 billion. On top of this, there is the usual transfer to local bodies (as grants under recurrent expenditure), which is about NRs174.7 billion.  
  • Second, the budget is pretty much a continuation of the FY2017 programs and policies as the ongoing local election bars the government from introducing new projects that are of distributive nature or that can influence the outcome of elections.
  • Third, the expenditure level is growing very fast and needs to be stabilized until the ministries figure out what is working and what is not (zombie and unnecessary/duplicate projects need to be culled line by line). Time has come now to seriously rationalize recurrent spending. The FY2017 budget was prepared in an irresponsible way by substantially jacking up expenditure. It will be hard to remove such expenditures from the 'Red Book' once included. The FY2018 budget has become a victim of that fiscally irresponsible/populist move. Recurrent expenditure in FY2016 was 16.5% of GDP, which was jacked up to 23.7% of GDP in FY2017 (revised estimate) and in FY2018 budget it is around 27.4% of GDP (see the figure below). 
FY2018 budget overview
GDP growth target (%)

Inflation target (%)

Budget allocation 

Rs billion
Budget allocation
Financial provision

Projected total revenue
Foreign grants
Principal repayment

Projected budget surplus (+)/deficit (-)

Projected deficit financing
Foreign loans
Domestic borrowing
FY2017 cash balance

1. Budget outlay:

The total expenditure outlay for FY2018 is NRs1,279 billion (an estimated 43.5% of GDP), which is 21.9% higher than the budget estimate for FY2017. The FY2018 outlay comprises NRs803.5 billion for recurrent expenditures (62.8% of the total outlay), NRs335.2 billion for capital expenditures (26,2%), and NRs140.3 billion for financial provision (11%).

The substantially larger size of the budget is due the sharp increase in recurrent and capital spending, particularly the increase in recurrent spending (which as stated earlier includes large transfer to local bodies). The outlay for recurrent expenditure (equivalent to 27.8% of GDP) is 43.1% higher than the revised estimated expenditure in FY2017. The planned capital spending has been increased by 27.9% over the FY2017 revised estimate (an estimated 11.4% of GDP). About NRs146 billion is set aside for post-earthquake rehabilitation and reconstruction.

2. Revenue target:

A total revenue target of NRs817.7 billion (27.8% of GDP) has been set for FY2018, including projected foreign grants of NRs72.2 billion (2.5% of GDP) and principal repayment of NRs15 billion. The revised estimate for revenue mobilization (including grants) in FY2017 is 25.2% of GDP. Revenue (tax and non-tax) growth target is 25.7%.

3, Deficit financing:

The budget deficit is to be financed by foreign loans equivalent to NRs214.4 billion, domestic borrowing of NRs145 billion, and FY2017 cash balance of NRs102.7 billion. Net foreign loans and net domestic borrowings are projected to be 6.3% and 4.3% of GDP, respectively. Overall, fiscal deficit is projected to be about 3.5% of GDP.

4. Where is the recurrent budget going?

Almost 29% of planned recurrent expenditure of NRs803.5 billion is going to local bodies (plus provincial bodies)  as fiscal transfer and another 23% is going as grants (including some social service grant) to local bodies (some of this money is used for local level capital projects as well, but then it also includes scattered pet projects of politicians).  The other big ticket item is the compensation of employees, which takes up about 16% of total recurrent budget. These amount to an estimated 14.2% and 4.4% of GDP respectively.

6. Where is the capital budget going?

Almost 60% of the planned capital budget of NRs335.2 billion is going for civil works. About 23% is allocated for building work. These amount to an estimated 6.8% and 2.6% of GDP, respectively.

7. How much are the local bodies getting?

The government has not formed Natural Resource and Fiscal Commission to oversee fiscal transfer, revenue distribution and grants to local bodies. So, a number of factors (population, development status, cost of operation, etc) has been considered to allocate transfer to local bodies. The distribution is as follows (see the details here):
  • Rural municipalities: NRs100 - NRs390 million (plus conditional grant NRs12 - NRs172.2 million)
  • Municipalities: NRs150 - NRs430 million (plus conditional grant NRs39.5 - NRs312.7 million)
  • Sub-metropolitan cities: NRs400 - NRs630 million (plus conditional grant NRs148 - NRs310 million)
  • Metropolitan cities: NRs560 - NRs1,240 million (plus conditional grant NRs281.2 - NRs783.9 million)
The local bodies will directly get the budget in their bank account by mid-August. A local council will approve the programs to be initiated by local bodies in their locality. These include program and projects on agricultural, livestock, irrigation, roads, drinking water and sanitation, schools (payment for teachers as well), tourism and sports, among others.

These are substantial fiscal transfers and grants that were approved and monitored by at Singha Durbar in the past, leading to substantial delays in project completion. Now, these can be directly used by the local bodies, which means the local residents will need to keep an eye on project selection, contract award and project progress. The local residents now have to point their fingers first at their local representatives instead of the bureaucrats and politicians at Singh Durbar.

These transfers are clubbed under recurrent spending. Fiscal transfer and grants to local bodies together constitute about 50% of planned recurrent spending in FY2018. In FY2017 grants to local bodies alone accounted for about 44% of estimated recurrent expenditure. So, in a way there isn't much increase in transfer to local bodies. It just that a little bit more has been added and some of the conditional grants to local bodies has been diverted as unconditional fiscal transfer so that the local bodies are not at the mercy of the lethargic bureaucracy at Singha Durbar. Some of it is used in capital accumulation or capital maintenance projects. But, a large portion of this for now will go to meeting recurrent expenses of local bodies. 

8. What about implementation?

Budget under-execution is a chronic issue. This budget tries to facilitate budget execution by cutting away several approvals needed even after the projects are included in the Red Book
  • Expenditure authorization will now be approved based on trimester work plan outlined in LMBIS of the MOF. Spending authority will be given to line ministries prior to the start of FY2018.
  • The government commits to establish project offices by mid-August. There will be a detailed work plan for project chiefs and relevant staff (this is nothing new, but it hardly get followed). 
  • Project preparation and tender calls should be completed by mid-August and contract should be signed by mid-October (after completing bid evaluation). Contracts will be canceled if the contractor does not adhere to the timeline agreed beforehand. Contractors will be barred from sub-contracting works (beyond their capacity but they nevertheless bid for the project) unless approved by project chief. 
Capital spending is affected by six major factors. First, structural weaknesses in project preparation and implementation remains unresolved. Barely any substantial homework is done before the inclusion of projects and programs in budget, leading to allocative inefficiencies to begin with. 

Second, low project readiness is another recurring problem as pork barrel and populist projects are inserted without feasibility studies and detail design, and time bound procurement plans and land acquisition plans. 

Third, bureaucratic hassle in approving and reapproving projects at various layers (sector ministries, Ministry of Finance and National Planning Commission) and weak intra and inter ministry coordination delay the full and effective realization of planned capital spending. This is the issue FY2018 budget is trying to resolve by directly transferring funds to local bodies and by doing away with approvals needed even after projects are listed in Red Book

Fourth, infrastructure projects of any scale and nature are riddled with poor project management, especially due to high staff turnover (which erodes institutional memory), lack of staff capacity to administer project implementation, lengthy procurement process, subpar capacity of contractor and weak contract management.

Fifth, high fiduciary risks in project implementation in suburban and rural areas when projects are implemented through local government having limited human resources and administration capacity not only delays spending, but also makes it inefficient. The funds allocated to parliamentarians to spend in their constituencies fall in this category.

Finally, political instability and interference both at planning and operational levels hinder timely completion of projects.

9. Here are the main takeaways from FY2018 budget:

The budget is introduced one and a half months prior to the start of FY2018, as per the constitutional provision. It is expected to boost capital spending. However, looking at last year's expenditure absorption rate, the progress is not as expected. Although the government is estimating 84% of the planned capital budget in FY2017 will be spent, the progress so far points to a much slower absorption rate. As of 29 May, only 34.1% of the planned capital budget is spent. 

This raises doubt over the commitment by the government to accelerate capital spending even though the government has committed adequate funds for infrastructure projects. The FY2018 budget commits to approve spending for projects from the beginning of the fiscal year and also directly transfer funds to the local bodies, bypassing the lengthy approvals needed at various layers of line ministries, NPC and MOF. However, the funds cannot be spent just like that. The local bodies need to follow procurement guidelines and develop project proposals. It is going to take time as the local bodies lack human resources and expertise to design, appraise, approve and execute project. Completing procurement by mid-October will be a huge challenge. The pace of capital spending, hence, depends on how well the center facilitates the local bodies in tackling these issues. Furthermore, the two more rounds of elections by January 2018 (on top of second phase of local elections on June 28) will require government employees to focus on conducting elections, which will deprive them of properly assessing and execute projects outlined in the budget. 

A robust, credible and a time-bound implementation plan to spend the earmarked money is partially addressed (the MOF says it is somehow included in the LMBIS, the inter-ministry budgetary information system). We will have to wait and see until the first trimester to evaluate its efficacy. 

The GDP growth target of 7.2% is a bit ambitious because it is easy to grow from a low base than from a higher base. FY2017 growth rate was high because of the low base effect, good monsoon, improved supply of electricity, some pickup in reconstruction and capital expenditure, and normalization of supplies. However, maintaining a high growth rate will be much more challenging. The rationale for 7.2% growth rate is not convincing even with a good monsoon, prospect of continuation of improved power supply, expected acceleration of reconstruction works and capital spending, and election-related expenses (provincial and federal elections after the local elections by January 2018). These drivers need to be much stronger than in FY2017 to attain a higher growth rate. This is unlikely to be the case as of now. A growth target between 5% and 6% looks reasonable if budget execution (including those at the local level), monsoon rains and elections-related expenditures happen as expected. 

The target to limit inflation within 7% is seems reasonable. Inflation are expected to be subdued in India and global oil prices are not projected to rise dramatically. General prices may heat up due to the increased recurrent budget (but then if the past is any guide not all of the planned budget will be spent). The downside risk to inflation target could be supplies disruption due to political tension in Terai region (as election day approaches and constitution amendment bill does not materialize). We need to see what kind of monetary policy the central bank will roll out in the coming months. 

Furthermore, net borrowing of about 4.3% of GDP (higher than 3.6% in FY2017) may put upward pressure on interest rates (especially when bank liquidity is drying up due to the deceleration of remittance inflows and the large government savings in NRB) and inflation (if actual spending is close to the planned one, which in all likelihood won’t be the case). Overall, fiscal deficit is projected to be about 3.5% of GDP. Net foreign loans are projected to be about 6.3% of GDP. 

The problem with such a large spending plan (met through increasing deficit financing) is that revenue is increased by all means, but then actual spending falls far short of planned one. This leads to substantial government savings, which are used to bloat the budget in the next  period (by committing to finance a part of it using cash balance). This is a bad budgetary practice. The government savings from previous years should ideally be used as a basket fund to finance large-scale infrastructure projects instead of financing a part of recurrent expenses. Budgetary practices should be clear to avoid ambiguity and misappropriation. FY2017 cash balance used in FY2018 budget is almost 22.2% of total planned deficit financing. 

There is no substantial change to existing tax policy. So, achieving revenue growth of 25.7% looks ambitious. If the revenue target is not met, then the government will for sure try to borrow a large amount of money from the market (usually it doesn’t exhaust its borrowing target unless there is a necessity to tweak the budget figures: for instance, to ensure that the fiscal budget is not in surplus like in the past).

Macroeconomy-wise, its a mixed picture: (i) a bloated budget justified to ensure reasonable initial financing and grants to local bodies in the federal setup; (ii) ambitious GDP and revenue growth targets; (iii) widening of net domestic and foreign borrowing; and (iv) a potential surge in interest rate due to the sharp increase in net domestic borrowing. 

With this kind of a bloated budget, which is practically beyond the absorption capacity of the bureaucracy and institutions, the government should not try to bring a supplementary budget after the local elections are over. The priority should be to fully execute the budget. Furthermore, time has come to stabilize the recurrent budget at this level and then start culling unproductive, repetitive and zombie projects. Revenue (tax and non-tax) is not even able to finance recurrent expenditure: in FY2017 recurrent expenditure was 97% of revenue and in FY2018 it is expected to be 110% of revenue.