- While the labour market is looking tight, we believe it is too soon for the Fed to seriously apply the brakes
- Hidden labour market slack and accelerating business investment means there is plenty of room for the economy to grow
- Helped by the recent tax reform package, we believe increased investment is driving a renaissance in productivity
- Higher productivity will dampen unit labour costs, and keep a lid on inflationary pressures
- The balance of hawks and doves on the FOMC remains in flux, with four voting members still to be appointed. Should these voters tilt hawkish, the risk is for more hikes than the two we project for 2018
The US is currently enjoying its third longest economic expansion in history, and should it continue past mid-2019 – as we project – it will be the longest. Growth is not just sustained, but has also accelerated: following a 3.3% annualised expansion in Q3 17, Q4 GDP is tracking 2.8% according to the Atlanta Fed’s GDPNow model – rates that are well above those of 2016 (1.5%) and the Fed’s own view of trend growth (1.8%).
What started as a tepid recovery from the 2008-9 recession has gradually gained momentum, and, having slowed over 2015-16, investment is again leading the way. After a mildly negative drag on growth in 2016, non-residential investment is likely to have contributed 0.6pp to growth in 2017, and – consistent with the surge in leading indicators such as the ISM new orders index – we project it will boost 2018 growth by 0.7pp. The recent corporate tax cuts and reforms to interest deductibility and depreciation accounting – which incentivises (potentially front-loaded) capital expenditure as opposed to the M&A and share buyback activity that has prevailed in recent years – could provide a further tailwind for investment, and poses upside risks to our 2.7% 2018 GDP growth forecast.
Grower for longer
What this means, for an economy that according to many measures is at full employment, is that the recovery can likely be sustained for longer before inflationary pressures significantly build. The rebound in investment is beginning to make an impact on productivity growth, which has been exceptionally weak in the post-crisis period. Having grown on average 1.2% q/q annualised between 2008-17, productivity growth in the past four quarters has picked up to 1.6%. The relationship between investment and productivity is a complex one, with variable lags and even an inverse relationship during the crisis period (investment fell sharply but so did the denominator of productivity – employment and working hours). However, over long time periods we do find that higher investment is consistent with higher productivity growth (see figure below). Should the acceleration in investment be sustained, past experience suggests this would be consistent with productivity growth returning to its pre-crisis average of 2.5-3%, and potentially higher.
A long overdue renaissance in productivity would create room for stronger wages, which have so far grown at a tepid rate despite headline unemployment and the oft-cited U-6 measure (which includes involuntary part-time and discouraged workers) both falling back to pre-crisis levels. However, the implications for inflation are less intuitive. A key support for core inflation is unit labour cost growth, but this has turned negative on a 4 quarter moving average basis, for the first time since Q4 2010 – reflecting the recent recovery in productivity growth. Until wage growth begins to significantly outpace productivity growth, we are unlikely to see a resurgence in core inflationary pressures. We suspect the protracted period of slow productivity growth after the crisis means there is likely pent-up room for a rebound, which could keep inflation lower for longer than the FOMC is projecting.
Hidden slack could also keep inflation in check
Labour market slack is also a key component driving unit labour cost growth. Although most indicators show slack is back to pre-crisis norms, one measure – labour force participation – remains an outlier. Overall participation averaged 66% between 2005-8, but in recent years has averaged around 63%. An interesting trend has emerged particularly in the 25-54 year-old category, where female labour force participation has recovered towards pre-crisis levels but the rate for males remains well below pre-crisis levels. While the causes of the decline in labour force participation are unclear, it is possible – likely, even – that workers are encouraged to re-enter the labour market as opportunities broaden, and income prospects brighten. Should this happen, the increase in labour supply would bear down on wage growth. Indeed, in the December FOMC minutes, “a few” participants suggested that the reductions in personal tax rates introduced in the Republicans’ tax reform package could lead to an increase in labour supply, suggesting the potential for labour force participation to recover. This, too, would keep a lid on unit labour cost growth.
The drag from weak (or even negative) unit labour cost growth is compounded by the persistence of structural drags on core inflation. Core goods inflation has been negative since 1995 and looks likely to remain so for the foreseeable future – a reflection of both globalisation and technological advances. Alongside the recent drag from mobile phone tariffs, key recent supports – shelter and medical costs – have also begun to cool. All told, we expect core CPI inflation to remain subdued in 2018 at 1.7% (2017f: 1.8%), before rising modestly to 1.9% in 2019 as wage growth accelerates and pushes unit labour costs higher.
(Nearly) All change at the Fed
While we believe inflation is likely to remain benign, one factor that could alter our current projection of just two rate hikes in 2018 is the composition of FOMC voting members this year. Of the twelve potential voters, just three (Powell, Quarles and Brainard) from 2017 will also be voters this year. Alongside the incoming Chair Powell, who we expect to chart a similar course of gradual rate hikes to the dovish-leaning Yellen, Donald Trump recently nominated the hawkish Marvin Goodfriend to the board, while ex-McKinsey executive Thomas Barkin – whose views on monetary policy have not been made public – is taking the helm at the Richmond Fed. There remain three additional board vacancies, including that of the influential Vice Chair (previously held by the hawkish Stanley Fischer), while NY Fed president Dudley is due to retire in the middle of the year (this role will be appointed internally rather than by President Trump).
While it may be too early to judge at this stage, so far there are fewer voting doves than in 2017, creating room for a potentially more hawkish-tilting FOMC. Hawks could prefer to pre-empt a build-up in inflationary pressures, perhaps pointing to the fact that overall financial conditions remain highly accommodative, despite the Fed’s rate hikes so far in the cycle. However, even if the remaining voting member vacancies were to be filled with hawks, we believe they would be reluctant to dissent against a dovish-leaning Chair Powell, absent a material trend-change in inflation.
That said, we acknowledge the risks to our view on inflation – that wage pressures could build up more quickly than productivity growth, or that productivity growth peters out once again, pushing ULCs and core inflation higher. While not our base case, in that scenario a more hawkish FOMC is likely to hike at a quicker pace than our current two hike projection for 2018.
Mueller could be a fly in the ointment
A final risk to our constructive outlook on the US is politics. As the Mueller probe into the Trump campaign’s Russia links continues, the fallout has already started, with a bitter war of words breaking out last week between President Trump and his former strategist Steve Bannon, who has called Donald Trump Jr’s meeting with a Russian lawyer prior to the election ‘treasonous’. The investigation could potentially – almost at any moment – plunge the administration into a crisis, and we must acknowledge the risk that the uncertainty this creates could impact business confidence, as well as financial markets.
Our chief economist, Han de Jong, last week sketched out a potential path towards impeachment and removal from office of President Trump, with reference to the cases of Andrew Johnson (1868), Nixon (1974) and Clinton (1998). We believe the risk of Donald Trump’s impeachment has increased, and is significant, but that it would still be extremely difficult to remove him due to congressional arithmetic.
How the President might be removed
The House of Representatives is the body capable of impeaching a president, by a simple majority. While the Republicans currently control the House, the midterm elections – where all 435 seats will be contested – offer an opportunity for the Democrats to regain control and impeach the president.
However, to subsequently remove an impeached president from office requires a 2/3 majority in the Senate. In the midterm elections just 33 of 100 seats will be contested, of which the Democrats already hold 25. Even in the unlikely event the Democrats held on to all their current 25 seats and won all 8 of the contested seats they do not already have in November, they would be unable – alone – to remove the president (a more realistic scenario, based on polling, suggests the Democrats could gain three seats for a total of 50). Thus, the charges against President Trump must be so serious that a significant number of Republicans also turn against him.
 We have marked Marvin Goodfriend as hawkish in our table because of his stated preference for a rules-based monetary policy framework, and early calls for rate rises in the post-crisis period. However, he is also a supporter of negative rates during economic downturns, suggesting he may be more pragmatic than he is usually portrayed.