Bob Gilhooly, Gene Kindberg-Hanlon and Dan Wales.
The dramatic fall in the price of oil has had a marked effect on headline inflation across the world. In contrast, measures of core inflation (ex. food & energy) have been more stable suggesting, that once the base effects from oil drop out, headline rates of inflation should bounce back. However, while inflation rates around the world will mechanically pick up in the near-term, it is not clear that global labour markets are strong enough to drive inflation fully back to target.
In June last year the price of a barrel of oil peaked at a little over $115, before falling to around $45 by January 2015, one of the largest ever falls without a clear trigger. Of course, this decline should provide a significant boost to many economies: as consumers spend less at the petrol pumps they could opt to spend more on other goods, boosting inflation in the medium-term. But, in the short-term, this decline in the oil price has lowered headline inflation across the world by around 2pp since mid-2014 (Chart 1).
Chart 1: Oil prices vs average CPI and core CPI across the world
Sources: Thomson Reuters DataStream, CEIC and Bank of England calculations.
* Global equates to 70% of world PPP GDP; 20 country sample.
Measures of core inflation that attempt to strip out shocks to energy and food prices have seen much more modest moves. So, while the decline in oil makes judging inflationary pressures much more difficult, this could suggest that underlying inflationary pressures remain strong enough to bring inflation back to target.
But the relative stability of average core inflation across the world may give false comfort. Core inflation had started to slide in many countries before the fall in oil prices. And, although in aggregate global core inflation appears stable, in the post-crisis period the relative strength of core inflation in many emerging markets has offset low core inflation in many advanced economies.
In the past, high commodity price inflation enabled authorities to reach headline inflation targets with relatively modest core inflation. Using the weights of food and energy in the CPI baskets for advanced economy inflation targeters and the average rates of food and energy inflation over the past, we can construct implied targets for core inflation for each country which are consistent with headline inflation at target. Indeed, such an implied measure arguably provides a better benchmark for assessing inflationary pressure, given that central banks will almost always look through volatility in commodity prices. Using this framework, the current rate of core inflation across advanced economies is around 0.4pp below its implied target, and has been so for over two and a half years (Chart 2, blue line).
Chart 2: Implied deviation of advanced economy core inflation from target
Sources: Thomson Reuters DataStream, CEIC and Bank of England calculations.
*AE equates to 35% world PPP GDP; *implied inflation targets assuming food & energy inflation of 3.8% pa, which is equal to the OECD average since 2000; ** assuming food & energy inflation of 2.5% pa.
Much of the surge in commodity prices between 2000 and 2014 was triggered by a step change in emerging market growth. This picked up from an average of 3.7% over 1980-2000 to 6% post 2000. Partly because of this, emerging market economies have accounted for all of the 20 million barrels per day increase in global oil consumption over the past 15 years. But more recently prospects for emerging markets have deteriorated. Several major emerging markets, including Russia and Brazil, have entered deep cyclical recessions and Chinese growth has continued to decelerate. The IMF expects growth in emerging markets to only recover gradually, averaging just 4.5% over the next three years. This could be associated with much lower commodity price growth, and highlights the possibility that the extraordinary commodity price growth which occurred over the past two decades will not be repeated.
Assuming this change in emerging market growth weighs on commodity price growth by around 1pp, we can re-examine just how problematic the current weakness of core inflation could be in returning inflation to target. In that case, the shortfall between the current rate of advanced economy core inflation and that consistent with the headline inflation target increases to around 0.7pp (Chart 2, red line).
To reach global inflation targets, any reduced contribution of commodity prices to headline inflation going forward would require domestic cost pressures to shoulder more of the burden than they did in the pre-crisis period. The labour market may provide an indication of whether the underlying drivers of inflation are sufficiently strong to return inflation back to target after the base effects holding inflation down drop out. Wage growth has been weak both in the UK and abroad, even in countries where slack in the labour market is now thought to be very small. Of course, much of this is likely to reflect (perhaps temporary) damage to productivity growth caused by the financial crisis and, as such, lower average wage increases could be associated with the same degree of inflationary pressure. Indeed, unit labour cost (ULC) growth rates across OECD countries are around 0.5pp lower than their pre-crisis average, and the interquartile range has shifted down, but they are not unprecedentedly low (Chart 3, blue line). But, assuming lower growth in emerging markets weighs on commodity price growth by around 1pp, current advanced economy ULC growth would be around 0.9pp beneath the rate required to hit inflation targets (Chart 3, red line).
Chart 3: Unit Labour Costs (ULCs)
Source: OECD and Bank of England calculations. *Assuming the all of the weakness in commodity prices inflation is offset by higher ULC growth to maintain headline inflation targets.
We are able to explain some of the recent weakness in wages as reflecting the remaining slack in advanced economy labour markets. Simple wage equations attempting to control for the degree of labour market slack and other factors, such as productivity, account for the weakness in wages reasonably well in recent quarters, even if the residuals on these equations have been negative for most G7 economies for around two and a half years (Chart 4). As the remaining slack erodes domestic cost pressures should increase. However, these equations have been estimated over a period when unusually strong commodity price growth supplemented weaker domestic cost pressures to keep inflation at target. Therefore, if lower commodity price growth persists, wage growth may need to strengthen by more than you would otherwise expect for inflation to return to target.
Chart 4: Unexplained component of nominal wage growth in G7 economies
Sources: Thomson Reuters DataStream Eurostat, ONS, OECD, national sources and Bank of England calculations.
Note: Wage equations estimated over 1995-2015 using logged variables as follows:
As the past falls in oil prices drop out of the annual comparison, rates of inflation will mechanically pick up across much of the world. But underlying inflationary pressures in the G7 remain weak, despite many countries having low and falling unemployment rates. While the relationship between the labour market and inflation is hard to pin down, it is difficult to believe that this underlying link is truly broken: we will need to see the labour market heat up further to be confident that inflation across the G7 is really headed back to target.
Bob Gilhooly works in the Bank’s International Surveillance Division, Gene Kindberg-Hanlon works in the Bank’s Global Spillovers & Interconnections Division and Dan Wales works in the Bank’s International Surveillance Division.
If you want to get in touch, please email us at bankunderground@bankofengland.co.uk. You are also welcome to leave a comment below. Comments are moderated and will not appear until they have been approved.
Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.
Great post guys! Totally agree that a key question in the coming months will be the further pick-up in inflation after the oil price effect drops out.
Deflation has and is taking hold across the Globe. Commodity Prices have and are leading the way. We are in a long term Bear MKT for commodities.
Peak to Peak in Commodities 20 year cycle, next top around 2030.
Velocity of Money is at record lows going back 50 years and going lower.
The oil price decline may not mechanically drop out. A further decline may create more disinflationary effects in the near future. Some are projecting oil at $20 a barrel soon. The St Louis Fed produced a report on whither inflation if oil was at $100, $52 or $20. $20, they suggest, would be deflationary not just disinflationary. Their projections are only for the USA. But their would be a similar oil impact on inflation world wide.
https://www.stlouisfed.org/on-the-economy/2015/october/predicting-impact-oil-prices-inflation
If true then more heating up of the labour factor would be required in G7 countries and elsewhere to achieve target inflation.
A most welcome and candid assessment of recent ‘noflation’, not least the move beyond commodity prices to unit labour costs. And agree with the conclusion that that labour market needs to ‘heat up further’ before inflation goes back to target.
But the worry is that these techniques still abstract from more fundamental problems. Unit labour costs tell us mainly that earnings growth has conformed to productivity. The long run average premium of just under two per cent is in line with inflation targets. The persistent shortfall since the crisis might be a worry, but is far from the main problem.
ULC tell us nothing about productivity. While done to death since the crisis, the ongoing slump in productivity across nearly all OECD countries continues to defy conventional explanation.
The TUC has argued that low productivity figures can be straightforwardly explained as a result of a prolonged weakness in aggregate demand. The argument is in three steps:
• reduced demand growth has meant lower economic growth, which the labour market has been forced to accommodate;
• it has done so through price – i.e. major reductions in wages (and potentially the quality of work) – rather than quantity – with employment numbers holding up;
• the productivity calculation sets disproportionately higher employment growth against lower economic growth, so that low productivity is simply an arithmetical result.
A fuller account is available here: http://touchstoneblog.org.uk/2015/03/productivity-no-puzzle-about-it/
So productivity can be straightforwardly explained from the point of view of demand, not as a result of labour hoarding but as a result of wage adjustment. Causality is reversed relative to conventional theory, with wage behaviour determining productivity not productivity determining wages. Productivity is therefore a symptom of these wider economic conditions rather than a quantity with causal force.
Since 2013 there has been a revival of demand (resulting first from interventions in the housing market and next as the government reversed cuts and began to spend, notably from 2014), but wages growth remained largely contained. As a result productivity remained weak. Only in 2015 was there finally some reversal, with wages growth finally picking up and correspondingly some gains in productivity. But wage gains are already looking short lived, not least in the context of resumed weaker aggregate economic growth, especially in nominal terms (which seems likely to be the more relevant view at present, with four quarter growth down around 1.5 percentage points on a year ago in contrast to real figures which are down only 0.5 ppts).
The wage adjustment is common to most countries. The paper above includes a decomposition of the reduced growth in ‘labour income’ since the crisis. Updating the analysis to 2015 and taking an average, the share of the adjustment on wages is more than half. The UK is striking as one of the countries where more than 2/3 of the adjustment was on wages, along with France, Norway, Sweden, the US and a number of eastern European countries.
The demand logic in turn contests the way capacity is assessed. From the perspective of the output gap, weakness in GDP outcomes are regarded by the Bank (and OBR) as structural because of the weakness in productivity. The output gap has therefore been allowed to close to near zero. But if productivity is a mathematical artefact, as above, this mechanistic interpretation is inappropriate. Likewise the logic of the NAIRU is based on employment and wage growth going hand in hand, and is opposed to the outlined labour market processes where they are moving in different directions. Arguably economic outcomes are exposing these conceptions as inappropriate in present conditions, ultimately giving a misleading signal of the extent of capacity.
The financial crisis made it fairly obvious that conventional reasoning does not capture the way economies actually operate, yet assessment of aggregate outcomes since the crisis has remained rooted in the same thinking with the supply side dominant. But it seems obvious that demand might be important, given the extent of the cuts in government spending (growth rather than levels) as well a wider conditions in the aftermath of the financial crisis. As we show in our paper, weak outcomes can be explained perfectly easily when using realistic multipliers. The cumulative effect of ongoing shortfalls in outcome is increased spare capacity, likely on a very large scale. Inflation outcomes (including commodity prices) should therefore be taken at face value, as supporting this interpretation.
While I can understand policymakers’ desire to break out of the seemingly extreme and unending monetary stimulus, on this reasoning the likelihood is that a reduction in stimulus (as just seen in the US) is inappropriate given the likely scale of spare capacity (leaving aside the legacy of indebtedness and ongoing asset inflations). I’m sure you know where this is going. The inevitable policy conclusion must that the danger of noflation giving way to deflation will not go away unless policy counteracts this ongoing weakness of aggregate demand . The obvious conclusion must be to expand not contract public spending?
great article thanks! an interesting question is: what if due to technological progress and peak population globally (in a few decades) we get commodity prices flatlining or even declining as the default??? We will have to look at the weights of the CPI basket as well. For example in Switzerland, one of the richest countries, energy has a significantly lower weight in the basket, so relatively, the drop in commodities has a smaller effect. So maybe far enough down the road, commodities may be just irrelevant for inflation!