13/2/2014: Sticky Wages, Job Effort and Jobless Recoveries

In two posts earlier I discussed some new studies relating to the problem of a jobless recovery (http://trueeconomics.blogspot.ie/2014/02/1222014-jobless-recoveries-post.html) and the ICT-driven displacement of workers (http://trueeconomics.blogspot.ie/2014/02/1222014-ict-productivity-employment-in.html). Here is another recent study dealing with labour markets outcomes in the case of a recessionary shock.

Here is another paper on employment adjustments, this time looking at cyclical shocks and wages rigidity: "HOW STICKY WAGES IN EXISTING JOBS CAN AFFECT HIRING" by Mark Bils, Yongsung Chang, Sun-Bin Kim (NBER Working Paper 19821: http://www.nber.org/papers/w19821, January 2014).

There is much evidence that wages are sticky within employment matches, so that incumbent workers face wages that do not adjust significantly fast downward in the downturns, thus creating a wage mis-match with entry of new workers. "For instance, Barattieri, Basu, and Gottschalk (2014) estimate a quarterly frequency of nominal wage change, based on the Survey of Income and Program Participation (SIPP), of less than 0.2, implying an expected duration for nominal wages greater than a year."

"On the other hand, wages earned by new hires show considerably greater  flexibility. Pissarides (2009) cites eleven studies that distinguish between wage cyclicality for workers in continuing jobs versus those in new matches, seven based on U.S. data and four on European. All these studies find that wages for workers in new matches are more pro-cyclical than for those in continuing jobs."


"Consider a negative shock to aggregate productivity. If existing jobs exhibit sticky wages, then firms will ask more of these workers. In turn this lowers the marginal value of adding labor, lowering the rate of vacancy creation and new hires. Note this impact on hiring does not reflect the price of new hires, but is instead entirely a general equilibrium phenomenon. By moving the economy along a downward sloping aggregate labor demand schedule, the increased effort of current workers reduces the demand for new hires.

The result of this mechanism of adjustment is that "wage stickiness acts to raise productivity in a recession, relative to a flexible or standard sticky wage model. Thus it helps to understand why labor productivity shows so little pro-cyclicality, especially for the past 25 plus years (e.g., Van Zandweghe, 2010)."

The authors set up two versions of the model for such an adjustment.

First model allows firms "to require different effort levels across workers of all vintages… During a recession the efficient contract for new hires dictates low effort at a low wage, while matched workers, whose wages have not adjusted downward, work at an elevated pace."

In the scone variant of the model, authors "impose a technological constraint that workers of differing vintages must operate at a similar pace. For instance, it might not be plausible to have an assembly line that operates at different speeds for new versus older hires."

The study finds that the second model "generates considerable wage inertia and greater employment volatility." In other words, if contracts do not allow firms to impose greater effort requirement on new hires against incumbent workers, there will be more shocks to unemployment and stickier wages for incumbents. Or in other words, there will be a more jobless recovery.

The authors provide an example: "Again consider a negative shock to productivity, where the sticky wage prevents wage declines for past hires. The firm has the ability and incentive to require higher effort from its past hires, in lieu of any decline in their sticky wages. But, if new hires must work at
that same pace, this implies high effort for new hires as well. For reasonable parameter values we find that firms will choose to distort the contract for new hires, rather than give rents (high wages without high effort) to its current workers. This produces a great deal of aggregate wage stickiness. The sticky wage for past hires drives up their effort and thereby the effort of new hires. But, because high effort is required of new hires, their bargained wage, though flexible, will be higher as well. In subsequent periods this dynamic will continue. High effort for new hires drives up their wage, driving up their effort in subsequent periods, driving up effort and wages for the next cohort of new hires, and so forth. By generating (counter)cyclicality in effort, this model can make vacancies and new hires considerably more cyclical." (Or put differently, it creates more unemployment at the shock and retains more unemployment in the adjustment period.

Now onto empirical evidence: "There is only sparse direct evidence on cyclicality of worker effort. Anger (2011) studies paid and unpaid overtime hours in Germany for 1984 to 2004. She finds that unpaid overtime (extra) hours are highly countercyclical. This is in sharp contrast to cyclicality in
paid overtime hours. Quoting the paper: "Unpaid hours show behavior that is exactly the opposite of the movement of paid overtime." Lazear, Shaw, and Stanton (2012) examine data on productivity of individual workers at a large (20,000 workers) service company for the period June 2006 to May 2010, bracketing the Great Recession. At this company a computer keeps track of worker productivity. They find that effort is highly countercyclical, with an increase in the local unemployment rate of 5 percentage points associated with an increase in effort of 3.75%."

The authors use their model to "show that sticky wages for current matches exacerbates cyclicality of hiring when effort responds. In particular, for our benchmark calibration with common effort, the effort response markedly increases the relative cyclical response of unemployment to measured productivity. It does so by increasing the response of unemployment to productivity, but also by making measured productivity less cyclical than the underlying shock."

The authors then look at the data to see if their "model is consistent with wage productivity patterns across industries, especially the cyclical behavior of productivity in industries with more versus less flexible wages. We measure stickiness of wages by industry based on panels of workers from the Survey of Income and Program Participation for 1990 to 2011."

The study finds that 

• "productivity (TFP) is more procyclical in industries with more flexible wages"; and 
• "this impact is much greater for industries where labor is especially important as a factor of production. 
• "However, we do not see that wages are more procyclical for industries with flexible wages, suggesting that frequency of wage change may not capture wage flexibility particularly well."

12/2/2014: Jobless Recoveries post-Financial Crises: Solutions Menu?

Next few posts will be touching on some interesting new research papers in economics and finance… in no particular order. Please note, no endorsement or peer review analysis from me here.

To start with: NBER WP 19683 (http://www.nber.org/papers/w19683) by Calvo, G., Coricelli, F. and Ottonello, P. "JOBLESS RECOVERIES DURING FINANCIAL CRISES: IS INFLATION THE WAY OUT?" from November 2013.

The paper discusses 3  traditional policy tools to mitigate jobless recoveries during financial crises: 

• inflation
• real devaluation of the currency, and  
• credit-recovery policies. 

The nominal exchange rate devaluation tool not being available to the euro area economies independently of the ECB, we have by now heard a lot about inflation (the need for). At the same time, real devaluation tool includes fabled European cost-competitiveness measures. 

Here's the pre-cursor to the paper: "The slow rate of employment growth relative to that of output is a sticking point in the recovery from the financial crisis episode that started in 2008 in the US and Europe (a phenomenon labeled “jobless recovery”). The issue is a particularly burning one in Europe where some observers claim that problem economies (like Greece, Italy, Ireland, Spain, and Portugal) would be better off abandoning the euro and gaining competitiveness through steep devaluation. This would be a momentous decision for Europe and the rest of the world because, among other things, it may set off an era of competitive devaluation and tariff war."

Hypotheticals aside, the study starts by "digging more deeply into the relationship between inflation and jobless recovery, also considering the possible role of real currency depreciation and resource reallocation (between tradables and non-tradables)."

As authors note: "This discussion is particularly relevant for countries that, being in the Eurozone, cannot follow a nominal currency depreciation policy to mitigate high unemployment rates 

(e.g. Greece, Italy, Ireland, Spain, and Portugal)."

First finding is that there is "some evidence suggesting that large inflationary spikes (not a higher inflation plateau) help employment recovery. Even in high-inflation episodes, inflation typically returns to its pre-crisis levels…" so the effects of the induced inflation wear out quasi-automatically.

Second finding is that "(independent of inflation) financial crises are associated with real currency depreciation (i.e., the rise in the real exchange rate) from output peak to recovery. This shows that the relative price of non-tradables fails to recover along with output even if the real wage does not fall, as is the case in low-inflation financial crisis episodes. This implies that, contrary to widespread views, nominal currency depreciation may eliminate joblessness only if it generates enough inflation to create a contraction in real wages; real currency depreciation or sector reallocation might not be sufficient to avoid jobless recovery if all sectors are subject to binding credit 

constraints that put labor at a disadvantage with respect to capital." In other words, there goes Argentina's fabled hope for recovery via devaluations.

Third finding extends the second one to the case closer to euro area peripherals: "Similarly, for countries with fixed exchange rates, “internal” or fiscal devaluations during financial crises are likely to work more through reductions in labor costs than changes in relative prices and sectoral reallocation obtained through taxes and subsidies affecting differentially tradable and non-tradable sectors." In other words, internal devaluations work, and they work via cost competitiveness gains and exporting sector repricing relative to domestic.

Tricky thing, though: "However, neither nominal nor real wage flexibility can avoid the adverse effects of financial crises on labor markets, as wage flexibility determines the distribution of the burden of the adjustment between employment and real wages, but does not relieve the burden from wage earners." which means that a jobless recovery is more likely under internal devaluation scenario.

Fourth finding: "Our findings highlight the difficulty in simultaneously preventing jobless and wageless recoveries, and suggest that if the goal is to avoid jobless recovery, the first line of action should be an attempt to relax credit constraints." Oops… but credit constraints are not being relaxed in the case of collapsed financial systems and debt overhang-impacted households in the likes of Ireland.

More on this: "Only direct credit policies that tackle the root of the problem seem to be able to help unemployment and wages simultaneously. …common sense suggests the following conjectures. In advanced economies, quantitative easing operations, especially if they involve the purchase of “toxic” assets, can have an effect on increasing firms’ collateral and relaxing credit constraints that affect employment recovery." But, of course, in Ireland these measures failed to trigger such outcomes - Nama has been set up for two years now and credit restart is still missing. May be one might consider the fact that targeting of bad assets purchases is needed? May be buying up wasteful real estate assets was not a good idea and instead we should have pursued purchases and restructuring of mortgages? Sort of what Iceland (partially and with caveats) did?

Overall, tough conclusions all around. 

7/6/2012: Irish Services PMI - May 2012

­­In the previous post (link here) I covered manufacturing PMI, showing a slight lift up in the growth rate from 50.1 in April (stagnant economy reading) to 51.2 in May (sluggish, but growth). More importantly, the 3mo average for March-May 2012 stood at 50.9 (weak expansion) compared to 48.9 average for December 2011-February 2012 (contraction).

Today’s Services PMI paints a weak picture in the other 48% of the private sectors economy in Ireland.

Headline Services PMI fell to 48.9 (contraction) in May from 52.2 in April. This marked the first month of sub-50 reading since January 2012. 12mo MA is at 51.2 and 3mo MA is at 51.1 in line with 12mo MA, slightly below 51.7 average for 2011.

This suggests that 5 months in 2012, growth conditions remain challenging. January-May 2012 average reading is 51.0, which, if sustained through 2012 will imply Services sectors growth of close to, but worse than a 2.15% real contraction in Services in 2011. Not exactly what I would call good news.

Of course, there are loads of various caveats to the above analysis, so don’t take it as some sort of a forecast.

New Business sub-index deteriorated from 52.7 in April to 49.6 in May, posting first usb-50 reading since January 2012. 12mo MA for the sub-index is now at 50.1, in effect implying that new business activity has been stagnant over the last 12 months. 3mo average is at 51.5 and the previous 3mo average was 50.2, some improvement on December-February period is still present. Good news, current 3mo average is ahead of same period averages for 2010 and 2011.

In line with broader indices, employment sub-index has fallen to 49.1 – returning to sub-50 level after March and April departures from the trend. Thus, 12mo MA for employment sub-index is now at 48.0 firmly signaling contraction in jobs in the sector. 3mo MA is at 50.3 owing to 51.9 spike in March, while previous 3mo average is 46.6. Current 3mo average and May level reading are both below the 3mo average for the same periods in 2011. 

Meanwhile, the giddy happiness signalled by the Services sector Confidence indicator bubbled up from 64.1 in April 2012 to 64.3 in May. The indicator runs on a silly scale well off the 50=neutral stance. Give you an example, in 2010, the indicator averaged around 66.7 and in 2011 it averaged 64.8. In both years, Irish Services sectors were, ahem… in a recession.

Output prices continued to fall, with the rate of decline accelerating to 44.4 from 44.9 between April and May. 3mo average through May is now at 45.4 and the previous 3mo average is 45.7. This marks continuation of below-50 readings in output prices since July 2008. Meanwhile, input costs rose at a faster pace (51.4) in May than in April (51.0), with 3mo average through May at 52.5, against previous 3mo average of 54.3.

Predictably, profitability was shot, again. Profitability sub-index fell to 45.8 in May from 47.5 in April.

More on profitability and employment in the following posts as usual.

5/3/2012: Services & Manufacturing Employment - PMI data for February

In previous posts I have covered new data on Manufacturing PMI and Services PMI. In this post, I will look closer at Employment sub-indices by these two broad sectors.

As before, all original data is courtesy of NCB, with analysis provided by myself. Some of the indices reported are derived by me on the basis of proprietary models and are labeled/identified as such.

Chart above shows core PMIs for Services and Manufacturing, highlighting the following changes:

• Manufacturing PMI moved from 48.3 in January to 49.7 in February, remaining below 50 line, signaling weaker contraction mom. 12mo MA is now at 50.3 and Q1 2012 average running is 49.0 against Q4 2011 average of 49.1.
• Services PMI has improved from contractionary 48.3 in January to expansionary 53.3 in February, with 12mo MA at 51.0 below february reading. Q1 2012 running average is 50.8 and it is almost identical to 50.9 average for Q4 2011.
• Volatility of Manufacturing PMI had risen from the STDEV of 4.48 in 2000-present sample to 5.62 for 2008-present sub-sample (crisis period), while volatility of Services PMI had fallen from 7.75 in 2000-present to 6.60 in 2008-present.

The chart below summarizes Employment sub-indices for Services and Manufacturing PMIs:

• Employment index in Manufacturing has deteriorated from 49.5 (contractionary) in January to 49.3 in February, with 12mo MA now at 49.9, Q1 2012 running average of 49.4 and Q4 2011 average of 48.6.
• Employment index in Manufacturing has become more volatile during the crisis, with STDEV rising from 4.41 for the sample of 2000-present to 5.51 for the crisis-period sample.
• Employment index in Services has improved from contractionary 44.5 in January to still contractionary 47.9 in February, with 12mo MA at 47.7 and Q1 2012 running average of 46.2 against Q4 2011 average of 47.3.
• Employment in Services is less volatile since the crisis on-set, with STDEV of index running at 6.71 for the sample of 2000-present against crisis period STDEV of 5.64.
• Overall, Employment index in Services is virtually as volatile during the crisis period as the Employment index in Manufacturing. However, before the crisis onset, and historically overall, employment was much less volatile in Manufacturing than in Services. This suggests, given strong growth of our exports in Manufacturing compared to Services, that most of our current exports boom is explained not by real economic activity, but by transfer pricing - a conjecture supported by my analysis of the trade data here. Note, that this is also consistent with lower overall employment and lack of jobs creation despite the relatively strong singlas coming from the PMIs in both sectors.

Charts below clearly show that our 'exports-led' recovery is not creating jobs and is instead associated with overall net jobs destruction continuing to rage across the economy.

So what is going on? we can only speculate, but in my view, 


Reasons why our Services PMI growth is not translating into jobs creation are: 
(1) much of growth is due to transfer pricing via IFSC & likes, 
(2) Maj of services exports are not labour intensive (hours worked) but skills intensive (high-end skills generating high value added), 
(3) Domestic services continue to shrink (retail etc), 
(4) Profit margins are very severely strained - so profitability has ben shrinking since end of 2007 every month, implying cuts in employment to raise productivity, 
(5) Many of jobs in services exports are NOT employing domestic workers as lack of skills drives these jobs into international markets. And these are the growth areas, while domestic employment sectors are shrinking. 


Incidentally, this is not new. 


Since the beginning of data series, in Manufacturing, we had 33 months characterized by rising unemployment and rising exports (exports-led jobless recovery) against 43 months of jobs-creating exports-led growth. So there is a 43.4% chance that any recovery in Irish manufacturing will be jobless. This chance is much higher during the current crisis, with 20 monthly episodes of jobless recovery against just 8 jobs-creating recovery episodes.


Similarly, in Services, since the beginning of the data history, we had 31 episodes of jobless recoveries against 32 episodes of jobs-creating exports growth. So probability of 49.2% is associated with seeing jobless recovery if a recovery is exports-driven. Since the beginning of this crisis, there were 26 jobless exports-growth episodes against only 1 month when jobs growth coincided with exports growth.


The above, of course, show exactly how fallacious it is to anticipate exports growth to translate into jobs recovery.

04/11/2011: October PMIs - risk of recession rising

Continuing with the analysis of the latest PMI figures for October 2011 for Ireland, this post is looking into the relationship between employment, PMIs and exports-led recovery both over historical horizon and the latest performance. The previous two posts dealt with detailed data on Manufacturing (here) and Services (here).

Manufacturing PMI posted a rise from 47.3 to 50.1 between September 2011 and October 2011, moving above 50 reading for the first time in 5 months. However, as explained in previous post this increase does not signal expansion, as 50.1 is statistically insignificant relative to 50. At the same time, employment sub-index for Manufacturing PMI remains in contraction at 47.1 (statistically significantly below 50) for the second month in a row.

Services PMI posted a slight improvement in the rate of growth at 51.5 in October, up from 51.3 in September, but once again, given the volatility in the series, these readings are not statistically different from 50 (no growth) mark. Meanwhile, Employment sub-index of Services PMI remains below water at 46 - same reading for both October and September.

Charts below show two core trends:

The trends are:

• Both manufacturing and Services PMIs are flatlining around 50 mark, signaling stagnation
• Both in Manufacturing and Services, there are no signs of easing in jobs destruction

Consistent with these trends, overall Services sector has moved from the position of relative jobless recovery signalled at the beginning of 2011 to border-line recession and jobs destruction in October. Manufacturing sector has moved from the optimal growth area (jobs creation and recovery) in the beginning of 2011 to a recession in October 2011.

In addition to weaknesses in employment and overall PMIs, October figures show deterioration in exports growth, with Manufacturing New Export Orders sub-index at 49.8 and below 50 for the second month in a row (note that 49.8 is statistically not significant compared to 50) and Services New Export Business sub-index at 50.1 (down from 53.1 in September). Both sub-indices show stagnant exports performance in the sectors. Chart below shows that we are now in a recession (albeit border-line) - vis-a-vis exports-led recovery in Manufacturing and are getting close to a recession in Services.

Economics 1/07/2010: Live Register - no recovery here

Live Register figures for June are truly depressing, folks. Regardless of what our QNA numbers telling us about real GDP growth, unemployment is continuing to climb.
We are now at 444,900 and climbing. In the year to June 2010 there was an unadjusted increase of 37,420 (+9.0%), down from an increase of 43,788 (+11.1%) in the year to May 2010. But that offers little in terms of consolation - most of people on LR in 12 months to May are still there - unemployed or underemployed.
A snapshot of weekly numbers above. Depressing. The average net weekly increase in the seasonally adjusted series in June 2010 was 1,450, which compares with a weekly increase of 1,650 in the previous month. But unadjusted things are looking much worse (figure above).

There was an increase of 4,800 males and 1,100 females in the seasonally adjusted series in June. Undoubtedly strengthening contraction in construction activity in June is not helping here.
The standardised unemployment rate in June was 13.4%. This compares with 12.9% in the first quarter of 2010, the latest seasonally adjusted unemployment rate from the Quarterly National Household Survey. We are firmly on track to reach 13.7% before the end of this year.

Rates of change in LR are also accelerating - a disheartening feature:
As I said in my previous post on QNA data (here): we are having a fake recovery.