Balanced Government Budgets: When It Makes Sense…And When It Doesn’t

Pakatan Rakyat has promised to balance the government budget…eventually. Barisan Nasional has committed to balancing the budget by 2020. Given the government debt situation and whether you believe in political manifestos, it all sounds good, right?

Problem is, I find it hard to fully justify this as a policy objective.

But to understand my way of thinking, let’s lay down some ground rules first:

1. At an aggregate economy level, income equals expenditure. This is an accounting identity and is always true. Intuitively, any expenditure by one household or firm constitutes income for another household or firm.
2. Fiscal policy is monetary policy by other means, and monetary policy is fiscal policy by other means, at least in the context of keeping the economy on its optimum growth path. In other words, fiscal and monetary policy can be considered substitutes.
3. Both fiscal and monetary policy actions are subject to lag effects – it takes time for policy changes to have an impact – but lags for fiscal policy are longer than for monetary policy (I know, market monetarists will object to the assumption of any lag in monetary policy).
4. Economies are subject to business cycle and commodity price cycle shocks i.e. they are not always at equilibrium.
5. The goal of macro policy is to maximise full employment, non-inflationary growth.

Now that that’s out of the way, let’s take three different scenarios – a strict balanced budget rule; a cyclical balanced budget rule; and a full employment (unbalanced) budget rule.

Strict Balanced Budget Rule

Under this rule, the government only spends what it earns in an annual budget cycle. The budget is always in balance, and no further debt is borrowed except to rollover existing debt.

This has the advantage of rapidly reducing the government’s debt ratio, as any growth will increase the denominator of the ratio relative to a static numerator. The lag effects of fiscal policy are not an issue. The problem here is that given assumption 1 (income=expenditure) and 4 (the economy is subject to exogenous shocks), government spending will automatically be pro-cyclical.

In boom times, revenues increase as companies make more profits and households experience stronger wage growth, leading to increases in tax collection. Since a strict balanced budget rule is in effect, government spending will also increase, which leads to even higher corporate and household incomes. This will boost growth precisely when it is least necessary, and leads to inflation.

The opposite happens in a downturn, as lower corporate profits and household income reduce tax revenues, which in turn forces down government spending. This leads to even lower private sector incomes and deflation, right at the point where government spending would actually be most useful.

The upshot is that, as far as fiscal policy is concerned, the objective of maintaining full employment growth devolves entirely on monetary policy. Moreover, because fiscal policy is so strongly pro-cyclical, monetary policy needs to be both more pro-active as well as more volatile.

A strict balanced budget rule only make any sense if you assume that the economy is always on or rapidly returns to the full employment growth path. Alternatively, it makes sense if you live in a world with no economic growth at all. In other words, you’re a classical or new classical economist, or a pure monetarist. Economies are always in equilibrium, markets are always self-regulating and welfare maximising, and government intervention is always welfare destroying.

Cyclical Balanced Budget Rule

Under this rule, it is assumed that the budget has to be balanced over a set period, but not necessarily in any given year. Singapore for instance follows this rule, with the budget to be balanced across every five year term of Parliament.

You can thus limit the pro-cyclical impact of a balanced budget to a degree, while still keeping the debt-busting effect of a balanced budget. The problem is that this assumes that the balanced budget cycle matches the fluctuations in the economy, and that downturns are largely of the “Freidman” variety i.e. downturns don’t shift the path of full employment growth. Also, while varying deficits and surpluses across the business cycle helps stabilise economic growth, the major burden of maintaining that growth on track still devolves on to monetary policy.

But the assumption that downturns are all Freidmanite is simply not true. Financial crises, for example, tend to be long and protracted and depress economic growth below its initial path. Commodity prices (which tend to exacerbate fluctuations in government revenue) also tend to have long, multiyear cycles that don’t conveniently coincide with a five-year parliamentary cycle. I think its safe to say that cyclical budget rules still entail some volatility in economic growth.

This sort of cyclical balanced budget approach makes sense to a mainstream Keynesian economist, who automatically assumes that (1) government intervention is possible and advantageous in spots, and (2) economies also tend toward equilibrium, but don’t always do so due to market failures. It might also appeal to monetarists who think monetary policy is the better stabilisation tool anyway, but accept the notion that government spending might be inelastic to income i.e. there are a lot of “sticky” components which are hard to vary over the short term.

The Full Employment Budget Rule

This is the old Keynesian paradigm, that governments can always counteract fluctuations in the path of economic growth through active intervention.

The principle is hopelessly wrong of course, at least in terms of the extremes it was taken to in the 1960s and 1970s as far as interference in the economy. But there are, IMHO,  some redeeming features of this approach, at least in the way it views government budgets and debt.

The main underlying point here is to adjust government spending based on the gap between potential and actual output. In theory that means that net government spending should increase in downturns, and decrease in booms. This makes no reference to the actual balance between government spending and revenue, which can diverge or converge as circumstances require.

That would be more intellectually appealing to me, if it weren’t for the problem that it assumes perfect foresight and coordination by government planners. If fiscal policy is characterised by longer lags in policy effectiveness than monetary policy, the choice should always be to use monetary policy in the first instance, and maybe varying fiscal policy just enough to ensure that it is not pro-cyclical.

Another problem is the unspoken assumption that there is a trade-off between employment and inflation – the (in)famous Phillips curve. But historical experience and the Lucas Critique show the dangers of relying on seemingly stable historical economic relationships.

The other issue is that the underlying principle of intervention gives open license to governments to interfere in any and all economic activity, which is not always a good thing especially since populist demands will almost certainly increase in times of economic stress.

Obviously this unbalanced approach would appeal most to Post-Keynesians and proponents of Modern Monetary Theory (MMT).

Some Thoughts

From the foregoing, there’s obviously no ideal budget rule…each has its advantages and disadvantages. Most reasonable people would probably opt for the middle ground of a cyclical budget rule – a little economic volatility might be an acceptable price for restraining public sector expenditure.

If you had to pin me down, I’d probably plop for an unbalanced, full employment budget rule, not so much because I believe implicitly in government fiscal intervention, but because I think:

1. Balanced budget rules ignore the need for the continuous provision of public goods and investment, irrespective of good times or bad;
2. I think the MMT perspective of sectoral (im)balances has a lot of validity, which means that fiscal imbalances directly reflect imbalances in other sectors (and vice versa); and
3. If the goal is to maximise social welfare and economic development, an unbalanced approach is more appropriate even if monetary policy is both more effective and works faster.

In fact, monetary policy has little to say about either the first or second points mentioned.

But there remains the problem of government debt. I’m not going the whole hog and accept the MMT notion that deficits and public debt don’t matter at all. If Reinhart and Rogoff were wrong in their cut-off point of 90% of GDP, the fundamental point that public and private debt crises impact long term economic development shouldn’t be lost, nor that debt crises are a recurring feature in global economic history.

There’s also the tricky and unresolved problem of picking out what “full employment” actually means. Fluctuating rates of structural unemployment – where people are out of work because of changes in the economic structure, or due to labour market friction – has more than a little empirical backing and helped lead to the breakdown in the predictions of the Phillips curve.

So to cut a long story short, I’m thinking that something approaching the fiscal equivalent of Market Monetarism (MM) might be worth exploring. In MM, the idea is for monetary policy to aim at stabilising nominal income around its optimal growth path, using policy to set expectations of future income growth which then (through multiple channels) become a self-fulfilling prophecy.

To support this, fiscal expenditures should also be aimed at a rate of increase in keeping with the economy’s growth potential. In effect, that means aiming for a steady, consistent and predictable increase in government expenditure along the same growth path as potential (not actual) national income.

Since government revenue will vary with the growth of the economy, this will automatically make fiscal policy counter-cyclical as it would tend towards producing deficits in downturns and surpluses in upturns. The bigger the shock, the bigger the dislocation in revenue, and the bigger the surplus or deficit produced.

More importantly, we won’t have fiscal policy and monetary policy at odds with each other. And we won’t have the pro-cyclical effect that boom-bust cycles in global commodity prices have had on government finances. This is a budget rule that is truly balanced to the whole of an economic cycle, not to the temporary ups and downs of that cycle.

Note that this kind of budget rule says nothing about the level or rate of increase in fiscal deficits or government debt, or how extensive government intervention in the economy should be. Those are social and political questions about the right mix between public and private sectors. But what it does do is, I hope, put fiscal policy under the same type of discipline that a balanced budget rule would have, without the costs of the latter in terms of exacerbating economic cycles.

Obviously, this needs to be modelled and tested out, which is beyond my capabilities at the moment. Also, the similarities with Friedman’s 3% rule for money supply growth (and its failure in practice) gives grounds for pause.

Nevertheless, I think this is an idea well worth exploring.