Walking The Tightrope To A Broken Fiscal Policy Analysis

KPMG International has released a report (I assume fairly recently) - Walking the fiscal tightrope: a framework for fiscal sustainability in government. In my view, this report is very useful, as it is an ambitious attempt to analyse a government like a corporation, and the weaknesses of this analytical approach is made obvious...

The report is 100 pages, and jam-packed with pretty pictures and graphs. It also has an introduction to the national accounting framework for governments (although the focus on "national balance sheets" makes no sense; see below). And the report has lots of recommendations for how governments should handle fiscal policy (spoiler: austerity programs!). But I was surprised but what was missing - I could not find a single theoretical justification for any of their policy recommendations.

At least the austerity supporters who cited Ricardian Equivalence attempted to create a theoretical justification for their views. (But they did the math wrong.)

The closest thing to a theoretical justifications are three sound bites: "intergenerational impacts","affordability" and "global economic interconnectedness".

1. Intergenerational impacts - this is completely unexplained and unjustified. What is the risk? Why? The idea makes no sense if you realise that the average voter is expected to be in the work force for about 45 years, and will be hanging around in retirement for decades after that. Meanwhile, almost no fiscal decision will leave a measurable impact after 20 years. Voters generally have to live with the long-term consequences of any decisions made by their elected representatives.
   For example, the "World War II" generation in places like the U.S. and Canada ran up a huge debt/GDP ratio as part of the war effort. The U.S. debt/GDP ratio fell to a low level until it was ramped up again by presidents who were part of the war effort in World War II (Reagan and Bush Snr.). 
2. Affordability - There is no coherent notion of what "affordability" means for a sovereign borrower, and no definition appears within the document (unless I missed it).
3. Global Economic Interconnectedness - I have no idea how this was supposed to work. My guess is that consultants just put in stuff about globalisation to make it easier to market their wares in emerging market countries.

I will attempt to be constructive here, and not list my many points of disagreement with their analysis. But I will note the root problem they faced - they attempt to analyse a central government like a corporation (which can make financial decisions without any noticeable impact on the rest of the economy). (No, I'm not making that up, they actually state that a government's finances is "ultimately" no different than a corporation's on page 28.)

In other  word, they are attempting to do macroeconomic analysis without using any concepts from macroeconomic theory. If they had spent just a couple of hours looking at how fiscal policy acts in a Stock-Flow Consistent (SFC) model, they would have realised that governments have limited ability to influence fiscal aggregates - unless they somehow change the propensity of the other sectors to accumulate financial assets. And that is not just a property of SFC models, it is what is also predicted by mainstream DSGE models.

I will now list three problems I see with their implied policy recommendations: the usage of national balance sheets, boosterism for infrastructure spending, and the efficacy of fiscal ratio targets for policy makers.

Usage Of National Balance Sheets

Obviously, you can define a balance sheet for a national government. But it is a completely meaningless exercise.

An asset should be valued based on the discounted stream of revenue that it generates for the owning entity. Balance sheets may use historical cost accounting, but that is only because historical cost is a relatively concrete concept. What matters is the "market value" of the asset, which is what the discounted cash flow represents.

What are the revenue generating assets for a national government? Well, those of you who are employees can go check your latest pay stub. Yes, you and every other taxpaying entity within a nation are the "assets" that generate revenue for the government.

National tax streams will rise at about the same rate as nominal GDP over time. If the expected long-term growth rate is higher than the discount rate, the discounted value of those tax streams become arbitrarily large (speaking loosely, "infinitely large"). It makes no sense to look at the balance sheet of an entity where the values of assets and/or liabilities can be infinite.

Infrastructure Boosterism

Infrastructure spending is like apple pies, motherhood and kitten videos on the internet - almost everyone is in favour of them. On page 30 of the report, the authors imply that infrastructure spending is one of the few justifications for the emission of government debt.

Having a properly maintained infrastructure is one of the basic justification for the existence of a modern state. And infrastructure spending has to be planned out over the long term. (As an aside, my home town of Montréal is a graveyard of white elephant projects that were based on what appeared to be sensible forecasts at the time. Long-term planning cannot defeat uncertainty.) However, infrastructure should be thought of as "stable core business" of a government, and it is not a great place to run counter-cyclical policy (just ask the Japanese).

Infrastructure spending is very similar to the "military Keynesianism" which was the post-war economic consensus (until the 1970's). Modern infrastructure is now heavily controlled by the private sector, so in addition to creating (expensive) jobs, the spending also has to cover high private sector return targets. This means that infrastructure spending ends up being distributed towards those with a low propensity to consume. This means that the "multiplier" on infrastructure spending will be lower than things like a tax cut or transfers, making it an inferior tool for macro stabilisation. The implication is that a bias towards using infrastructure spending for counter-cyclical policy will cause fiscal deficits to become much larger than if more efficient tools are used (as seen in Japan).

In summary, any modern government should have a long-term infrastructure spending plan. And any modern government will have net financial liabilities outstanding (government debt and money). But there is no reason to expect that there will be any relationship between the size of those two activities.

Budget Processes!

The key actionable recommendations for the report are for governments to engage in complex new planning processes on 3 horizons: 1-5 years, across the cycle 6+ years, and inter-generational (10+ years).

There are few obvious objections to this suggestion.

• The results of these planning exercises rest completely on the accuracy of economic forecasts of the fiscal planning bodies. Based on the accuracy of other fiscal forecasts by official bodies, the only value of these forecasts is that they will provide fodder for entertainment on bondeconomics.com. 
• A belief that deficits will be in balance across the cycle shows a lack of regard to the behaviour of stocks and flows in the economy.
• Fiscal policy is an integral part of what government is. Long-term planning implies abolishing politics from government.
• There are no observed macroeconomic implications from fiscal ratios, and so there is no basis to set any target level for them.
• In the extremely small chance that forecasts get aggregate economic forecasts right, the forecast outcomes will still be crushed by relative price shifts.

To explain the last point, this is yet another area where you should not think in terms of a single-good economic model. The true constraints facing government are not financial, they are real constraints - how much materials and labour do these policies consume? (Aside: this concept was developed as part of "Functional Finance".) At a given time, relative prices allow us to balance off the relative importance of inputs, as Austrian economists point out. We should thus not be surprised that government revenues and expenditures are roughly matched in dollar terms over time (although continuous "small" deficits are the norm). But relative prices change.

The "military Keynesian" Welfare State of the 1950-1960's ran into the rocks in the 1970's as the result of a relative price shift - labour got expensive relative to the price of goods. Although this represents a welcome increase in the standard of living for workers (real wages rose), this meant that labour-intensive policies that made sense in the 1950s became uneconomic, and fueled the inflation that we saw in the 1970's.

(Also see the discussion in Chapter 11 of Hyman Minsky's Stabilizing an Unstable Economy. Targeted government expenditures towards a sector of any economy breeds market power for private sector entities in that sector. Market power is invariably exploited, and inflation in that sector takes off. This point is applicable to spending in the infrastructure sector as well.)

My guess is that if the architects of the Welfare State had used the planning tools suggested in this report in the 1960's, the programmes would have looked perfectly sustainable by the financial criteria used. However, the programmes turned into "engines of inflation" (Hyman Minsky's phrase) as the non-forecastable relative price shift made the demands on real resources excessive.  

(c) Brian Romanchuk 2014