In recent years, the application of expenditure baseline projections has become an important element in medium-term budgeting. Baselines provide information about the spending pressures of maintaining current policy commitments and facilitate an early assessment of the available fiscal space for new policy initiatives. They are also a key reference point for setting expenditure ceilings, developing budget requests and identifying areas to conduct spending reviews. This article supplements the IMF note on How to Prepare Expenditure Baselines[1] by providing some practical lessons learned through my own experience in the Catalan Government and the Barcelona City Council but also from years of international consulting.
Overall, the implementation of spending baselines has markedly improved the budgetary systems of Catalonia and Barcelona, in particular the allocation of fiscal space and the setting of initial expenditure ceilings for budget preparation. In addition, baselines have helped improve control over the generation of new multi-year expenditure commitments during budget execution (Catalonia), enabled a performance-oriented approach to processing budget requests (Barcelona), and improved the quality of information disseminated to the Parliament leading to better informed budget debates (Catalonia, through its Pre-Budget Report). But implementation of baselines has proceeded slowly and by trial and error, and has not been without its challenges. Some of these challenges are discussed below.
A key feature of expenditure baselines is that they are technical in nature and by definition do not reflect any new policy decisions. A potential risk to mitigate is the misconception that baselines are protected from cuts, thus encouraging the inflation of these projections to “camouflage” new initiatives. Dealing with this risk requires developing clear communications and training of budget officials, as well as providing enough time for baselines to be reviewed by the government’s central budget office. For example, in the first years of implementing expenditure baseline in the Barcelona City Council, several million euros were “cleaned” from draft baseline projections after their review (which usually lasted around 2-3 weeks).
The probability of such problems arising is higher if baselines are prepared along with spending agencies’ budget requests. In the early stages of implementing expenditures baselines in Catalonia, spending agencies were allowed to provide several inputs and propose large amendments to their draft expenditure baselines during the budget request process. This initially increased the risks of introducing new policy initiatives camouflaged within the baselines, thus confusing the dialogue between spending agencies and the central budget office. Later, the process was changed so that spending agencies and the budget office reached agreement about the baseline figures at an early stage of the budget preparation process, a change that also benefited the implementation of performance budgeting (in Barcelona).
Regarding the often-debated trade-off between developing aggregate or detailed expenditure baseline projections, my experience suggests that the detailed approach offers greater benefits, and allows an easy integration with the annual budget estimates and the development of forward estimates. In Barcelona, the development of baselines were conducted at the highest level of the budget and accounting classification (budget items, subprograms or organizational units). Such an approach helped prepare a zero draft of the budget. It also reduced the workload of finance officers, by avoiding the need to reconcile baselines prepared at an aggregate level with much more detailed information required for the draft budget, and by reducing the risk of errors and inconsistencies in the data.
In both Catalonia and Barcelona, the automation of information on expenditure baselines, and the integration of these data with other financial information systems has been critical to the successful implementation of expenditures baselines. For example, the semi-automated loading of data on budget execution and multi-year expenditure commitments, along with relatively low-grade analytical and manual work, helped the Catalan Government to quickly generate initial expenditure baselines for current goods and services, capital investments, transfers and subsidies. In Barcelona, links with the human resource IT system helped to integrate projections of current personnel costs, with a high level of granularity.
Finally, Barcelona and Catalonia adopted different approaches on whether baselines should include workload (service delivery levels) adjustments due to changes in demand. These changes can complicate the calculation of baselines, especially in the initial stages of the reform when it may be considered desirable to follow a simple and easily implementable approach. For this reason, Barcelona decided not to apply such adjustments. Catalonia applied them for several services within its education, healthcare, and social services programs (e.g., long-term care and universal basic income). These adjustments significantly increased the complexity of the system, but without having a robust methodology to estimate the corresponding marginal costs or to provide reliable information on current and estimated output levels and demand coverage, thus leading to potential errors and inconsistencies in the projections.
