NNIP: Should we be worried about payrolls?

The 3-month moving average of monthly payroll gains slowed down to 167 K per month, well below the 200-250K range seen in Q2. What's more, the decline is even more pronounced for private sector payrolls and the slowing was fairly broad across sectors. In addition to this, employment growth in the household survey (on which the calculation of the unemployment rate is based) came down substantially as well. Despite this the unemployment rate held steady at 5.1% as the participation rate registered a 0.2pp decline and is now at a 38-year low. Meanwhile the U6 unemployment rate, which includes involuntary part-timers and marginally attached workers, fell to 10% which is 1.7 pp lower than a year ago. All in all, the labour market data over the past two months seem to point to a slowdown in employment growth coupled with a reduction in labour supply growth. At first sight this is not good news so it is worth investigating what could have caused this.

We see three possibilities:

I The slowdown could be just a statistical fluke. The monthly payroll data are very volatile with a standard deviation of around 100K so from a purely statistical perspective we cannot draw the conclusion that a slowdown has actually taken place. In support of this hypothesis we note that circumstantial evidence on the labour market continues to signal strength. The job openings rate in the Job Openings and Labour Turnover Survey (JOLTS) survey has accelerated markedly during the course of the year and now stands at 3.9% which is the highest since 2001. Meanwhile, the JOLTS hiring rate has been oscillating a tad above 3.5% over the past year which is only marginally below the average seen during the 2004-2007 boom. What's more, the employment component of the composite ISM has increased over the past few months and is consistent with trend payroll growth close to 300K. Also, the trend in weekly jobless claims remains around levels last seen in 2000. Finally, consumer assessment of the labour market measured in the Conference Board confidence index remains on an improving trend and is at the most benign level since the start of the crisis. This suggest that US households at least perceive a further tightening of the labour market.

II The slowdown in payrolls could be a genuine expression of increased corporate caution. We should not forget that US businesses have been facing some headwinds over the course of the year. The stronger dollar and weaker external demand have reduced the dollar value of overseas profits. In addition to this, the marginal cost of capital has increased on the back of the decline in equity markets and the widening of corporate credit spreads. Hence, there is little doubt that corporate profit margins are under pressure and even corporates which do not feel this very much (yet) could be subject to a moderation in confidence as result of this. All this could certainly lead to attempts to strengthen profit margins by cutting back capex and labour demand. Nevertheless, we should bear in mind that the level of US profit margins is still close to record highs which means that the corporate sector should be pretty resilient in the face of shocks, especially if these are perceived as temporary. In this respect, the dollar drag on the growth rate of corporate profits should wane once the dollar stabilizes and the same should hold for the drag exerted by weakening external demand and a rising cost of capital. In addition to this, the more domestically oriented sectors in the US economy are accelerating. Consumer spending is on track for a pretty strong performance heading into autumn and the same holds for the housing sector. These developments should act as a support for corporate revenue growth. Adding it all up, if corporate caution is the driver of the observed slowdown in payroll growth it may just be a temporary phenomenon. It will then essentially be an expression of the fact that the big shifts in relative prices (dollar and oil) have caused front-loaded corporate pain and back loaded and more widespread consumer gains. The latter should then in the end ensure that corporate appetite for expansion remains intact. Still, there is always a risk that corporate caution becomes so strong that the associated slowdown in hiring and capex triggers a negative feedback loop between rising corporate and household savings appetite.

III The slowdown in payrolls could have been driven by an acceleration in productivity growth. By this we do not mean an increase in productivity growth driven by corporate cost-cutting (this is basically the previous option) but a genuine improvement in the in the underlying trend. Labour productivity growth has been held down by low rates of investment which decrease the amount of capital per worker. In addition to this the relative price of labour versus capital has decreased over the past few years as real wage growth was very subdued while the SME sector faced tight credit conditions. This has probably caused the business sector to substitute away from capital towards more labour input. On top of this tight credit has impeded the transfer of resources from old low productivity firms to new high productivity businesses. Now that these factors are reversing we should expect an improvement in underlying productivity growth but this has not yet showed up in the data. If and when it does we should expect a slowdown in employment growth for any given level of GDP growth rate as well as a slowdown in the rate at which the unemployment rate declines. Of course we have not seen the latter over the past few months but one should bear in mind that we also saw a substantial slowdown in labour supply during this period. Taken in isolation the latter tends to push the unemployment rate lower. More generally, it will be impossible to tell whether or not an acceleration in productivity growth is the driving force for some years to come. Productivity is calculated by dividing real GDP by total hours worked. Real GDP is subject to a host of shocks not related to productivity which may have less influence on hours worked (especially if these shocks are perceived as temporary). Hence, we really need a few years of data to be able to distil the underlying productivity trend.

It is impossible to tell which factor is the main driver behind the payroll data. In fact, all three factors could have played a role. What we do know is that so far the payroll slowdown has been moderate and that the general trend among all labour market indicators still strongly suggests further labour market tightening. On top this, the acceleration in consumer spending and housing indicators combined with the stabilization of capex indicators points to moderately above trend domestic demand growth. In this respect there is no reason to change our base case story for the US which is one of domestic strength in the face of an increasing external drag (which is becoming more and more visible in the net exports data).