A SIGNIFICANTLY large number of countries have taken repeated recourse to IMF assistance over the years (termed ‘prolonged users’ by the Fund, and ‘recidivists’ in a recent Forbes article!). Pakistan is one of them. While the immediate impulse is to lay the blame squarely on the serial non-performance of these countries with regard to economic reform, the truth is that many have ‘successfully’ completed IMF programmes in the past with limited success in actually addressing the fault lines in their economies.
Clearly, there are fundamental issues with IMF programmes in and of themselves too, ranging from deficiencies in programme design to misaligned incentives, agency conflicts and a lack of appreciation of the political economy. At the programme level, the concerns and questions raised are about the key objectives/end-goals being pursued, the policy instruments of choice, and how the programme is structured in terms of pace and sequencing. All these design issues have not just important ramifications for the macroeconomic performance of a receiving country, as well as serious social consequences, they also have a large bearing on the sustainability of reforms.
Pakistan’s engagement with the IMF from 2008 onwards is illustrative in this regard. In the 2013-16 programme, there was a heavy, front-loaded emphasis on tax collection from day one. However, this approach precluded meaningful reform of tax administration — which would have made the reform sustainable in the long run, but would have required giving authorities more time to implement serious measures (ie back-loading the deliverables under the programme). Without a meaningful reform of tax administration, reform of tax policy alone is insufficient.
This is not the only area where programme design has been at odds with Pakistan’s structural reform requirements. In the power sector, achieving ‘cost-recovery’ tariffs for electricity has been made a Holy Grail since 2008, if not earlier. The problem in the power sector (and in natural gas distribution to a large extent) is of massive revenue leakages that occur due to non-recovery from large consumers, and theft that occurs in collusion with the field staff of the utilities. In essence, this is a governance problem that requires strengthening institutional capacity. Yet the World Bank (which has ownership for tax as well as power-sector reforms) continues to attack the problem with the only tool it knows how to use — upfront price adjustments.
IMF programmes typically suffer from design deficiencies.
Not unsurprisingly, Pakistan has made multiple large adjustments to electricity tariffs under IMF/ World Bank ‘advice’ since 2008, with the result that the circular debt in the power sector has increased from circa Rs120 billion in 2008, to over Rs1,400bn in 2019. While part of this phenomenal increase in the circular debt represents an increase in installed capacity of the system, a large part represents the World Bank’s ideological — and quixotic — tilting at the administered price windmills. Similarly, in 1988, the World Bank got Pakistan to auction its public debt and liberalise interest rates in the mistaken belief that the price the country would have to pay (literally) for borrowing would force greater discipline in public finances. Instead, overnight, the country’s interest payments on public debt soared, adding to the structural fiscal deficit.
Despite an unblemished record of failure on this front, the World Bank continues with the same approach.
The foregoing discussion highlights the importance of sequencing in reforms. Many if not most of the reforms involve building institutional capacity to ensure that the end objective(s) are achieved. A minimum institutional ‘bandwidth’ is required to push through the reforms proposed under IMF programmes, especially since a majority of Fund assistance is tied to front-loaded conditionality. In our case, institutional capacity needs to be built at the minimum in tax administration (both federal as well as provincial), the energy governance chain (regulators, ministries, companies), and in key implementers such as the finance ministry and the central bank.
This is a tall order that requires several years of serious, committed effort for the best of reformer countries. And yet, the current IMF programme with its heavily front-loaded conditionality implicitly assumes this capacity exists across a wide range of government. This disconnect between required and actual institutional capacity in an implementing country is a fundamental defect within IMF programmes — and perhaps a leading cause for many recipient countries to relapse repeatedly into IMF ‘therapy’.
It also highlights another disconnect between the IMF/World Bank and recipient countries — an agency conflict to be precise. Since IMF programmes are typically structured around a one- to three-year duration, while a country’s needs for structural reform require a longer time frame, the Fund staff promotes ‘shock therapy’. Policy action is focused on dismantling the administered price regime, subsidies, the exchange rate, interest rates — actions that provide quick ‘tick-the-box’ successes for staff and management to report to the executive board.
The intrinsic ‘need for speed’ for the Fund staff can jeopardise the overall reform objectives of a country by undermining the political capital of an incumbent government by focusing on less important (and sometimes completely spurious) policy actions, while introducing serious sequencing errors.
If the current IMF programme were more aligned with Pakistan’s reform objectives, it would have been larger in size (at least $9bn instead of $6bn); contained more structural conditionality rather than a near-complete emphasis on upfront revenue collection and price adjustments; and be back-loaded rather than ‘front-loaded’ in terms of policy action on tax and power-sector reform. A more gradual adjustment in energy subsidies could easily have been accommodated by savings in interest payments in the budget by not undertaking the excessive increase in policy interest rates. This would have provided a more meaningful cushion against disruption for the middle class as well as the export sector.
The final point highlights another critical flaw in Fund programmes — the inability to gauge the actual burden of adjustment (or wider distributional effects) of recommended policy action, and their social effects, and hence an inability to insulate the most affected segments of society, despite the safety nets for the poor.
The writer is a former member of the prime minister’s economic advisory council, and heads a macroeconomic consultancy based in Islamabad.
Published in Dawn, August 16th, 2019