It is known that employment has taken a deep hit in this last recession, and has been much slower to recover than output (McKinsey Report on jobless recoveries).
This is related to the ongoing debate as to whether the crisis has permanently shifted the level of potential output or whether output will eventually meet the pre-recession trend level. Are we facing structural or cyclical shifts in the labour market? Structural and statistical estimates tend to produce disparate results depending on the country under scrutiny.
If we plot Portugal’s GDP trend growth (see footnote) and corresponding output gap using quarterly data between 1970-Q1 and 2011-Q1 we obtain the figure below.
It is clear that the “golden” days are over and that the so-called “lost decade” seems to have arrived to stay… Not only do we have negative trend growth, despite some mild signs of moderation/reversion(?), but also negative output gap.
Output can be decomposed into a product of population (p), the labor force participation rate (lfpr), the unemployment rate (u), hours per worker (h) and labor productivity (lp): Y=pop*lfpr*(1-u)*h*lp. Let’s just (briefly) analyse some of these components.
Plotting the trend growth of employment, labour force and unemployment we get the following:
One does see a downward evolution of the employment trend growth up to a point where around 2006 it became negative. Evidence seems to suggest than the trend may be reverting and the inflection point reached. A similar, but lagged behaviour, can also be seen in the case of labour force. Simultaneously, unemployment trend growth has been positive (since the early 2000s) having reached a peak in 2009-Q1 and signaling a downward movement. In any case, it seems a “New Normal” has arrived, characterized by higher unemployment and lower employment across all economic sectors and age-ranges which, most likely, won’t go back to pre-crisis levels.
One certainly needs to revert these variables’ behaviour in the short to medium-run for which a sine qua non condition is getting the Portuguese economy back on the growth track, increasing its potential output and closing the gap. All in all, some of these stylized facts should encourage policy-makers and the new government to "do something".
Footnote: For the interested reader, trends were computed by means of a univariate fitted Structural Time Series Model with unobserved components allowing for a stochastic trend, cycle and seasonal and estimated by maximum-likelihood via the Kalman filter. Source: OECD data; author's calculations.