Latin America Economics Update
19 February, 2016
Brazil: anatomy of a budget meltdown
• We estimate that only one-sixth of the jump in Brazil’s budget deficit over the past two years has been due to a drop in tax receipts. Instead, the bulk (five-sixths) has been due to a rise in expenditure. Most of this rise in spending has been due to an increase in interest payments and losses on FX swap contracts but there has also been a substantial increase in so-called “primary expenditure”.
• Having bumped along at around 3% of GDP or so for much of the period since the 2008 global financial crisis, Brazil’s public sector budget deficit began to blow out from mid-2014 as the economy slumped. Over 2015 as a whole it was equivalent to 10.4% of GDP.
• Despite the alarming size of the deficit we don’t believe that Brazil is facing an imminent fiscal crisis. (For more, see “Brazil: in a fiscal hole, but default fears overdone”, 23rd September.) But understanding why the deficit has blown out is important to assessing both future fiscal prospects and the measures necessary to put the budget on a stable footing. The problem is that the official data do not provide much help in this regard.
• Between 2013 and 2015 the public sector budget deficit widened by 7.4%-pts of GDP – from 3.0% to 10.4%. At the same time, the “primary” balance (that is, excluding spending on debt servicing) worsened by 3.6%-pts of GDP, from a surplus of 1.7% to a deficit of 1.9%.
• Put differently, 3.8%-pts of the deterioration in the budget deficit must have been due to spending on debt servicing. This partly reflects rising government bond yields. But we know from other data that around 1.5%-pts of this is losses incurred by the central bank on FX swap contracts. These contracts were issued to try and stabilise the real – when that policy failed and the currency sank further, the central bank suffered losses on the contracts and these were passed to the government.
• Interest spending aside, we don’t get a full breakdown of primary expenditure (i.e. expenditure not including debt servicing costs) and revenues for the full public sector budget. However, we do get a breakdown of the central government budget and, on that basis, it’s clear that, while there has been a steady decline in tax revenues as economic activity has fallen, a jump in primary spending has played a far greater role in pushing up the deficit.
• Putting all of this together, Table 1 shows our estimate of the contributions to the rise in the overall public sector deficit since 2013. Of the 7.4%-pts increase in the deficit over this period, we think that only 1.1%-pts (i.e. one-sixth) can be explained by a loss of revenues. In contrast, 6.3%-pts (i.e. five-sixths) can be explained by a rise in expenditure. Of this, 3.8%-pts can be explained by a rise in interest and debt servicing costs while 2.5%-pts can be explained by an increase in “primary” spending.
• The crucial question for the government is how much of this blow-out has been caused by economic weakness that should be reversed as the cycle turns and how much has been due to deeper structural problems that require a more fundamental tightening of fiscal policy. The truth is that it’s difficult to answer this without a having time series of the output gap – and estimating the output gap with any accuracy in a country like Brazil is next to impossible.
• Nonetheless, we can have a stab at separating the cyclical from the structural. On the revenue side, the close correlation between revenue growth and GDP growth suggests that almost all of the drop in receipts has been due to economic weakness. The problem for the government is that the new normal for GDP growth is likely to be much lower – meaning that revenue growth will be correspondingly weaker. We think as much as one-third of the revenue lost since 2013 could be permanent.
• On the expenditure side, the losses on FX swap contracts should fade as the real eventually stabilises. Interest payments on government debt should fall too as bond yields normalise. But with the debt pile now larger than before, around 1.3%-pts of the rise in debt servicing costs could be permanent.
• It’s more difficult to assess a structural/cyclical split on primary spending, but a breakdown of the contributions to the increase over the past couple of years helps. The rise in social security spending could reasonably be attributed to rising unemployment. But increases in other areas of the budget spending are more difficult to blame on economic weakness. All told, we think that around 1.5%-pts of the rise in primary expenditure is likely to be permanent.
• In other words, of the increase in the budget deficit of 7.4%-pts of GDP since 2013, around 4.2%-pts may be attributable to the economic downturn or temporary factors, with the remaining 3.2%-pts due to a more permanent loss of revenue or increase in expenditure. The former should resolve itself as the economy recovers; the latter will require a fiscal squeeze to eradicate.
• Put differently, Brazil needs budget consolidation equivalent to around 3% of GDP in order to return the deficit to pre-crisis levels. There are two major barriers to overcome. The first is constitutional mandates on spending, which mean that the bulk of any squeeze is likely to be through an increase in (already high) taxes. The second is the backdrop of a chronically weak economy. As we’ve noted several times before, it’s extremely difficult to repair the public finances in times of economic downturns, since the ensuing budget squeeze adds to the pressures on growth, and thus government revenues.
• In this context, the current debate over proposed changes to the fiscal framework – which argue for a shift towards a target range, rather than a fixed point, for the primary budget – misses the point. This amounts to fiddling around the edges. Brazil’s fiscal problems run much deeper – and will be long-lasting.