More on fasheesmo and income distribution
I'm taking advantage of a peaceful and orderly digestion to satisfy some unsatisfied curiosities in some previous posts, particularly this one and this one , which focused on wage dynamics and income distribution (a central theme of our reflections). With your permission, I'll start with a curiosity I had, and then move on to yours.
The wage rate at the time of the colonels
In the previous post, we saw that in Greece the wage share dropped dramatically during the colonels' era. We're accustomed to thinking that the purpose of fascism was this: to subvert income distribution to the advantage of capital, and it doesn't matter whether the person doing it was the Mussolini Churchill liked or the Democratic Party Europe liked: it's always the same old story. The phenomenon, therefore, doesn't surprise us (with all due respect to those who still believe in the myth of the asocial left or right), but I was still curious: was the collapse of the wage share due to a repression of the numerator (wages) or an explosion of the denominator (GDP)? Let's be clear: if a ratio declines, it's clear that the trend in the numerator is less sustained than that of the denominator. The question, therefore, is whether we can discern in the data a clear structural change in one (and only, or predominantly, one) of these two elements.
The formula for the wage share in the AMECO database is this:
ALCD0 = [(UWCD:NWTD):(UVGD:NETD)]x100
where UWCD is the income from employment, NWTD is the employed persons, UVGD is the GDP (nominal) and NETD is the total employed persons, the variables are these:
(I have highlighted in blue the dates corresponding to the colonels' period) and yes, you can see that the growth of the numerator was lower than that of the denominator, but you don't see any particularly apparent structural changes in the two series:
To rationalize it according to our categories, I also calculated the trend of the wage share as the ratio between real wages and productivity:
where RWGD is the real wage calculated as (UWCD/NWTD)/PVGD and PVGD is the GDP deflator (note: we use the consumption deflator instead), and RVGDE is output per worker, calculated as OVGD/NETD and OVGD is real GDP. The relationship between these two series has exactly the same dynamics as the wage share ALCD0:
even if not the same units of measurement (it differs by a constant scaling factor that depends on the choice of price bases, I won't bore you with unnecessary details), but the point is that if we want to summarize the history of real wage and productivity growth in Greece by dividing it into not too arbitrary historical periods, what we get is something similar:
which is to say that you can see when the colonels left (from 1975 to 1981 real wages catch up with productivity, and therefore the wage share grows) but, unlike what one might think, you can't see when they arrived (in the period before their arrival real wages grew by about three points less than productivity, more or less the same as during the dictatorship, and therefore the wage share fell at a rate substantially similar to that which it would have shown during the dictatorship).
Incidentally, the other period of recovery in real wages was 2002-2007, from entry into the euro to the crisis, and I think you can see the connection (both with entry and with the crisis).
But this was just a curiosity of mine, I now turn to your esteemed curiosities.
The incidence of overtime hours
marcellocamaioni left a new comment on your post " Real Wages: An Update ":
A curiosity: For about three years, we've seen a markedly greater increase in hours worked than in employed workers, so during this period, on average, each worker has worked more hours per year than in previous periods (something similar was also seen in the years before Covid). This is presumably due to an increase in overtime hours, which, while resulting in an increase in hourly wages for the worker, likely costs the company less overall (the calculation does not take into account the 13th month salary, severance pay, and a reduction in social security contributions). Is this one of the ways to try to increase real wages per employee without compromising the company's competitiveness? What is the trend over time in the average real hourly wage or equivalent work units?
Posted by marcellocamaioni on Goofynomics on December 24, 2025, at 8:10 PM
We've already discussed this when we addressed the topic of choosing the labor input in calculating the real wage . The phenomenon is illustrated by this graph:
Indeed, given that the trend in hours worked is more procyclical than that of employed workers (for the simple reason that when activity slows, it may be rational to engage in labor hoarding rather than incur the costs of firing, and thus hours worked fall more rapidly than employed workers, while when activity accelerates, it may be rational to increase hours worked rather than incur the costs of hiring, and thus hours worked increase more than employed workers), we expect that, when measured per hour worked, real wages will grow less during economic expansions (since in this case their denominator grows more).
We had also noted this at the time:
but I am proposing again the two measures of real wages (per employee and per hour worked) updated and in a comparable manner.
Meanwhile, it may be useful to see how the three measures of labor input have moved, both in thousands:
which in the form of an index number:
It is noted that the acceleration in hours worked relative to employed people in recent years actually tends to close the gap that opened with the recession at the beginning of the century but above all with the Great Global Crisis of 2008 (and subsequent austerity), a gap likely determined by the spread of part-time work (while obviously the dynamics of hours worked is strongly correlated with that of equivalent work units).
We can then replicate the wage calculations in real terms by referring them to the hours worked:
and compare the trend of the two measurements obtained:
where obviously, since during the COVID era, many employees were paid not to work (and therefore the downward spike in hours worked is much more pronounced than that of employed people), we are witnessing the statistical paradox according to which the Conte II Government recorded an unusual spike in average hourly wages, which also affects the ranking of governments, which changes as follows:
always with Draghi in last place, due to the cumulation of the inflationary surprise and the statistical artifact that puts Conte II in first place.
So yes, there are obviously differences in the profile of real wages measured by hours worked or by number of employees, but these don't substantially alter the conclusions of our discussion, according to which the current government has recorded the best real wage growth (excluding the extreme experience of Conte II, which operated under exceptional conditions due to the lockdowns).
…and Spain?
Omar left a new comment on your post " The Wage Share in the Time of Fascism ":
I was curious to see how mythological Spain was doing.
Posted by Omar on Goofynomics on December 26, 2025, 4:42 PM
The answer might be: "Then go and look at it, since I've given you all the references to be able to do so!", but Omar is a careful reader and so I do:
Spain's dynamics are similar to Italy's but more pronounced: the decline was more pronounced after the 2009 crisis, the recovery was more pronounced during COVID, with one exception: the recovery in the wage share over the last three years is less pronounced (but since 2020, Spain has been at a higher level than Italy).
I don't think I've forgotten anything, but if you have any other questions, please let me know and we'll try to answer them (I couldn't find the wage quota for Russia and China anywhere…).
This is a machine translation of a post (in Italian) written by Alberto Bagnai and published on Goofynomics at the URL https://goofynomics.blogspot.com/2025/12/ancora-su-fasheesmo-e-distribuzione-del.html on Sat, 27 Dec 2025 15:23:00 +0000. Some rights reserved under CC BY-NC-ND 3.0 license.