US Employment and GDP Tolerances

One of the lessons from the UK recovery is that sometimes employment is not simply a lagging indicator responding to changes in identifiable changes in consumption or investment, but also the source of those changes. This is no very striking or original insight, perhaps, but it is quite often overlooked or forgotten. Does this have any lessons for the US – particularly in the light of the shocking slump in 1Q GDP to just 0.1% annualized, and disappointing April data from retail sales (+0.1% mom) and industrial production (down 0.6%).  Consensus GDP forecasts have subsequently slumped to 2.5% for 2014.

But if we treat employment as being the main determinant of GDP growth, does that still seem reasonable?

Non-farm payrolls are currently rising at a rate of 1.7% a year, and have been doing now for two and a half years, and though there's no sign that this is resulting in any acceleration of wage growth, there's also no no sign that it is peaking out. The rate of job openings has not yet recovered to pre-crisis levels, but is steady at  around 2.8% (vs 3%+ pre-crisis), whilst the quitting rate (thought to be indicative of labour  market confidence) continues to plod slowly but steadily higher.  Meanwhile, initial jobless claims have fallen to their lowest levels since mid-2007.  So despite the shocks of the last few weeks, there appears no immediate threat to labour markets.

What would we expect from GDP growth if non-farm payrolls continue to rise around 1.7%-1.8%?  Here we confront the difference between the economy pre-crisis, and the economy now (the 'new normal'). Prior to the crisis (1992 to 2007),  a simple regression would associate a 1.7% non-farm payroll growth rate with a GDP growth rate of 3.4%. Since the crisis (2007 to the present) that growth rate would slip to 2.6%.  Unless there is a substantial upward revision in 1Q's dismal 0.1% rise, that would imply an acceleration during the rest of the year to 3.4%. It seems a stretch from here, but it's not impossible – annualized growth averaged 3.1% during 2Q-4Q13. 

But to get much beyond the 2.6%-2.7% range of growth, and into the 3%+ levels expected for 2015 and 2016 will demand either a sharp acceleration in employment growth or a rise in the productivity of those who are employed.   And here we get to the point: there's no mystery why the current level of employment growth is associated with a far lower rate of GDP growth than prior to the crisis: the fall in productivity is simply a reflection of the lack of investment spending, and an actual fall in capital per worker deployed.  I estimate changes in capital stock by depreciating all nominal gross fixed capital formation over a 10yr period. 

By that reckoning, capital stock grew just 1.9% in the 12m to 1Q14 – and that was the highest growth rate since the start of 2009.  But it also means that capital per worker is now growing by only 0.2%, and that the average amount of capital deployed per worker is currently 2.5% below its 2009 peak.  It also compares to an average growth rate of 4% during the 1992-2007 period.

So the conclusion is this: at present, the 1.7% employment growth is compatible with a central growth rate of around 2.6%. But this is likely to edge upwards provided growth of capital stock continues to outpace growth of employment. After all, it turns out that the prospects for 2015 and beyond depend above all on the emergence of a recognizably robust investment cycle.