How iSalari™️ is calculated (part one)
(… taking advantage of the undertow of the bridge, I resume this post started on April 30th and then interrupted for these reasons , with the aggravating circumstance of this incident – incredible: someone considers information operators to be intelligent life forms! – and I try to bring it to an end, so we can put an end to the objections of ChatPD users and move forward with the discussion on the essentials …)
Beppe88 left a new comment on your post "Darling, my wages have shrunk!":
Good morning Professor, a question (perhaps a bit of an amateur, but I am): what changes compared to the graphs on the trend of average unit wages shown in other blog articles and which seem to highlight a different trend in our wages (i.e. totally flat)? I am referring for example to this article:
https://goofynomics.blogspot.com/2024/11/la-kuestione-salariale.html
Is the difference related to productivity? That is, in the graph of this post only gross wages are considered without weighting them for productivity growth, while in the previous ones the trend in labor costs per unit of product is considered?
Thanks as always and have a nice day,
Joseph
Published by Beppe88 on Goofynomics on April 30, 2025, 11:14 am
A more than legitimate question, and indeed sorry if due to the rush I don't always post the links or even mention the databases I use. We go through all the steps, so you will learn to use them too, and this will allow you to do a more accurate review (which is not trolling) than what I tell you.
Meanwhile, I'll tell you where the salaries are not found. You won't find them in the IMF's World Economic Outlook database. It is not easy to find them in the International Financial Statistics , which have just changed the interface, improving it (i.e. making it worse for those who have been used to the old one for decades, like me): once upon a time there was an index of nominal contractual wages, there will still be, but I only have two hours and a limited number of invocations of the deity available, I don't have time to try my hand at what some engineer has found was a more logical method for reaching information that was once just a click away. They are not in the World Development Indicators , where if you search for "wage" you get this:
i.e. statistics on child labor and inclusion, but not the level of wages.
However, there are in the OECD Data explorer , in fact, I would say that there are too many, because by searching for "wage" in the search engine:
you get 92 databases:
and to arrive at something similar to what we need we must necessarily limit the research to national economic accounts:
but let's not continue this miraculous catch today. They are in the AMECO database , where you can find them by selecting the definition of GDP on the income side:
(the last time we explained what GDP is was here , incidentally, those who don't know it do well to take a look at it and those who know it do better, because in reality they believe they know it; however, I refer you to the Italian Stock Exchange website for an effective summary , so you will understand why to find salaries it is better to look for GDP…). However, the AMECO database has the limitation of only offering annual data. The quarterly data are found for example in the Eurostat database , which is what I used, selecting in the tree
the breakdown of the quarterly GDP. Then of course there are in the Istat database , which combines the dubious advantages (let's call them that) of the Ifs (the interface has just been "improved") and of the OECD (it provides too much data, so one gets lost), with an undoubted disadvantage: it only reports Italian data, as is right for a national statistical institute to do, but we need to be able to compare the situation in different European countries with homogeneous data.
Referring to Eurostat also offers a tactical advantage towards citizens: for them everything that is "European" is better, so let's defeat "nationalism" by keeping ISTAT aside for a moment, and try to quickly remember what we are looking for.
Let's start from the Italian definitions, remembering that, in a somewhat counterintuitive way at least for me, things are like this:
The cost of labour, including social security contributions (in particular those borne by the employer), is called "income", while the income received by the worker (net of social security contributions) is called "salary". In English things are like this:
Income from employment understood as labor costs is called "compensation of employees ", of which wages and salaries , i.e. what in Italian we call "gross wages" (because they are gross of taxes even if they are net of social contributions) are one component, the other component being the social contributions paid by the employer .
Alles clear?
In this graph:
I was using the Wage and salaries of Italy.
Before going any further, I would like to highlight a limitation of the Eurostat database: in the vast majority of cases it does not provide data prior to the mid-1990s. Now, we know that the stasis in real wages began in the early 1980s:
in obvious synchrony with the Treasury-Bank of Italy divorce, but on such long samples we do not have quarterly data. What is needed is the annual data from AMECO, the use of which has been illustrated here (so anyone who wants can go and see it there).
Let's get back to us.
While the definition of GDP on the demand side (consumption, investments, etc.) is provided in both nominal (current prices) and real (constant prices) terms, the definition on the income side is provided only in current prices, because it is not obvious to understand which price index to use to "deflate" wages and profits. As a rule, therefore, you will not find real wages in the sources:
but you will have to choose which price index to deflate wages with to calculate the purchasing power of workers. This, as we will immediately see, involves an additional level of complexity, but one thing at a time.
In the meantime, let's see how to download the Wage and Salaries. If you click on this link:
you get this screen:
that is, by default the database provides you with raw GDP (not seasonally adjusted) at current prices for the last 10 quarters out of 43 countries (reading the information counterclockwise starting from the highlighted one.
Instead, you need seasonally adjusted wages and salaries at current prices for all available quarters and for a single country ( Italy ). You must therefore "customize" the request and you can do this by clicking on the blue button at the top right of the last screen. You are brought here:
where, for example, by clicking on "Chack all" you select all available periods. You can then select the various elements you wish to modify from the list on the left, and use the right pane to modify them. For example, in the box that allows you to select the variable of interest you should do this:
by selecting "Wages and Salaries". At this point the most obvious thing to do would be to select the constant price measurement, like this:
but the result would be this:
given that for the reasons I specified above, wages in real terms are not calculated. Then you have to select the size at current prices, etc.
At the end of this long story you should end up with this result:
and by dragging "Time" on the lines you can reorganize the information like this:
so you can see more data:
after that, using the "download" button:
you will find everything in a convenient Excel sheet:
All what?
The total wage bill in millions of euros at current prices. So not the average wages at current prices (to obtain them we must divide by the number of employees ), not the wage bill at constant prices (to obtain it we must divide by a price index), not the average wages at constant prices or "in real terms" (to obtain them we must divide by the number of employees and a price index).
Let's start with price indices.
We can use the consumption deflator, the GDP deflator (to stay in the same database), or the HICP (harmonized index of consumer prices), and whatever you want. In the graph that intrigued you so much, I used the HICP, but in fact the most obvious thing is to use a deflator, simply because you have them in the same database. To obtain them you must select one (or more) demand components instead of Wage and salaries , for example family consumption expenditure:
and instead of the measurement at current prices, the deflator. You get this:
i.e. this:
The two indices differ for obvious reasons (the GDP deflator takes into account not only the prices of consumer goods, but also of capital goods, imports, etc.):
and therefore the inflation rates calculated using the two series differ. It is normally believed that to evaluate the purchasing power of families the consumption deflator is more appropriate, so we will use that and the calculations are quickly made:
We divide the nominal wage bill by the consumption deflator of the corresponding year, which is in turn divided by 100). We thus obtain the series freed from the effect of prices. Here I represent them both:
For today we'll stop here: I prefer to break the lesson into four for reasons of my health and your understanding. If there are things that are unclear, you can report them in the comments. Perhaps, before asking, interested parties could try to reproduce this last result: if something doesn't add up, then we need to think about it. I could always have been wrong…
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/2025/05/come-si-calcolano-isalari-parte-prima.html on Fri, 02 May 2025 10:48:00 +0000. Some rights reserved under CC BY-NC-ND 3.0 license.