Average V Median wage Ireland?

Podge_irl

There is insufficient data available to any agency to give an accurate and complete figure for the median. Therefore they have to estimate it based on combining different datasets and extrapolating a result. It is not a "calculated" figure, it is an mathematically modelled figure and as anyone will tell you, models can give very different results depending on how they are set up.

You have been told this multiple times. It is impossible to accurately calculate the median wage based on the information available. All anyone does is estimate it.

crossman47

https://www.boards.ie/discussion/comment/118494834#Comment_118494834

The CSO do publish two different figures for median earnings and explain the basis for both of them. First of all, they are derived using administrative sources (mainly Revenue data) and so cannot be used to produce any estimate for full time workers since I am fairly sure full/part time is not a record on anyones revenue file. In any event, how do you define partime (under 20 hours, under 40 hours?) If you decide on 30, what about someone who works three ten hour days.

That aside, the median weekly earnings figure (629 in 2020) relates to all employments in October so I presume it excludes the likes of summer jobs. it is a weekly figure since, I presume, not all those included would have been employed for the whole year.

On the other hand, the annual figure of 40,579. For this one CSO say

Employees who worked for less than 50 weeks in the reference year are excluded from the calculations for annual earnings. This is done to improve comparability of the data over time. Typically the proportion of employments retained for annual earnings analysis is relatively consistent at both the total economy and economic sector level. However due to the impact of the COVID-19 pandemic on the labour market, the proportion of employments retained for annual earnings analysis in 2020 is lower than in previous years, specifically for those sectors most impacted by the COVID-19 pandemic.

So this figure includes just those we can say had stable jobs through the year. This, I think, goes a long way to explain the differences.

CrookedJack

https://www.boards.ie/discussion/comment/118494834#Comment_118494834

You asked me a question and I answered, the same answer has been given by multiple posters in at least five different ways on this page alone. If something is getting us nowhere it's hardly my civil response.

Benedict

Perhaps matters are bouncing around more than is needed. Have a look at two statements made by Peregrines 610 above (in the same comment)

"They haven't used "two different methods of calculation" — they haven't used any method of calculation."

Some lines later:

"The methods used may be giving the best estimate they can, given the data available."

I admit to being unable to unravel this sort of entry.

The fact is that if the CSO are using the best data available (not perfect, but the best they have) and using the only method of calculating the median (because there is only one method), then there should not be two different answers.

crossman47

https://www.boards.ie/discussion/comment/118496000#Comment_118496000

See my post (613 above) which shows how there are two different answers because the population in question is defined differently.

Podge_irl

The CSO does not have the data to calculate the median. There is no dataset where this calculation is possible.

The CSO estimates the median using different datasets. This requires "filling in the gaps" in the data - there are different ways of doing this. This gives different answers.

It is impossible to "calculate" the median wage for full time workers

Benedict

"The nominal median household disposable income in 2019 was €43,552, compared with the 2018 figure of €42,865"

The above is an example, a quote is taken directly from CSO website. It doesn't mean they go to every home in the country and check the disposable income. They gather data and extrapolate from it. "Filling the gaps" is normal. Estimating is normal. They take the best data available and ESTIMATE. Statisticians estimate - that's largely what they do for a living.

Dutch people are the world's tallest. Do you think they need to measure everyone to be able to say that? It doesn't mean if you're Dutch, you're tall!

The notion that arriving at a median ftw figure is "impossible" is not true. They don't need to know what every ftw is earning - just enough to arrive at a fair assessment.

So what have I just said which is not true?

moon2

I think it's your repeated rejection of the estimated numbers as "not being accurate" which is the problem.

(Edit: it's a terse answer because I really don't understand why you're still trying to discover 'the true median' , especially after that post)

Sam Russell

Take the possibility that the median is calculated exactly from precise figures. Let us say it comes out as €40,000 for a ft equivalent of all the workers in Ireland. Good, we have a figure.

Then Mr and Mrs Super-rich arrive in Ireland and for tax reasons change their tax residence (and domicile) to Ireland. They both each earn a several zillion euros per year and count themselves as ft employees.

Now, the median calculation is not thrown out of kilter, because the number of ft employees has risen by two, and the total income has risen by two times several zillion, which does not mean the median must now rise to take account of the two times several zillion earned by those two individuals, because - surprise - only one income goes from above the median to below the median, and because the incomes around the median are much the same, the median remains the same.

The average (mean), however is a different story - now that really does rise.

I think the median as a measure of income is meaningless - 50% of the given population earn above it and 50% earn below it, what does that tell us?

crossman47

https://www.boards.ie/discussion/comment/118500353#Comment_118500353

Nothing you say is untrue but, to achieve what you want, you must have a file of full time workers to sample from. CSO medians mentioned above are based on revenue files and these do not separately identify full time workers. So, how do they start.

What they do is estimate a median for all workers. As explained above, "all workers" is defined in two different ways and so they get two different estimates. Anyway, defining full time is not easy. Do you base it on hours worked, days worked? What is your cut off point? If someone who works 30 hours a week over five days full time? if yes, what if those hours are worked over four, or even three days?

Benedict

A quote from the CSO Statistical Yearbook of Ireland for 2017 goes as follows:

"Average annual earnings were €45,611 for full-time employees in 2016 while part-time employees earned €16,597 on average in the year. "

I appreciate that these figures are "average" and it's "median" we're after. But it does demonstrate that the CSO are in a position to discriminate between ft and pt workers.

So the notion that they don't have data to show who's ft and who's pt can't be right.

crossman47

https://www.boards.ie/discussion/comment/118512399#Comment_118512399

You again misunderstand. The figures you quote are from a survey of enterprises who are asked to provide aggregate information for all employees (distinguishing full time). This allows a mean to be calculated but not a median because it does not mean they have a file of all (or a sample) of full time employees. The file they do have does not distinguish full timers.

Benedict

https://www.boards.ie/discussion/comment/118512945#Comment_118512945

I'm not sure what you mean by "enterprises" - is it companies? Anyway, are you saying that the figures I have quoted relate to aspecific group of "enterprises" and not to earners in general?

crossman47

https://www.boards.ie/discussion/comment/118516571#Comment_118516571

All I'm saying is that they are aggregate figures for enterprises/companies so it is not possible to use them to calculate a median.

Benedict

https://www.boards.ie/discussion/comment/118518770#Comment_118518770

I quote from CSO "Median annual earnings were €40,579 in 2020".

Do you have a theory as to how they arrived at that figure?

The CSO mention elsewhere that for the purpose of calculation, they regard ft as 50 or more weeks p/a. Clearly the above median figure is too high to have include pt workers. So presumably it refers to ftw.

moon2

No need for theories. They explain exactly how it was calculated.

https://www.cso.ie/en/releasesandpublications/ep/p-eaads/earningsanalysisusingadministrativedatasources2020/backgroundnotes/

crossman47

https://www.boards.ie/discussion/comment/118520065#Comment_118520065

You're wrong again. The CSO state

For the purposes of this analysis the CSO excluded employees earning less than €500 per annum and employments where the duration was less than two weeks in the year. Also excluded were secondary employments earning less than €4,000 per annum, extremely high earnings values and missing employer and employee reference numbers. Employment activity in NACE sectors A, T and U has also been excluded from the analysis. 

So, it includes part time workers if they worked 50 weeks in the year. As you have been told numerous times, the relevant Revenue file does not distinguish full/part time.

Benedict

"As you've been told numerous times". Dear oh dear - puts me in mind of a weary teacher scolding Johnny in the back row for getting his sums wrong "Yet again". ("What'll we do with you at all Johnny?")

Anyway, I said "presumably"...."presumably" it refers to ftw because "presumably" if ptw were included, the figure would not be in excess of 40k.

It's not my fault that the CSO don't go to the bother of specifying who they're talking about.

There seems to be more than enough data sluicing around to make it possible, with cross-referencing, to distinguish pt and ft. There are references to hourly rates and part time averages (€16,597 on average in the year in 2017).

Peregrinus

They did specify who they're talking about — you just didn't bother to read it. And then crossman47 quoted it for you, and you apparently didn't bother to read it again.

Benedict

https://www.boards.ie/discussion/comment/118524200#Comment_118524200

So can you quote where they specified "part time" of "full time" in their statement "Median annual earnings were €40,579 in 2020".? You seem to have seen it - so quote it please.

moon2

https://www.boards.ie/discussion/comment/118524396#Comment_118524396

Hold on a minute. Let's pause this particular redirection and go back to your underlying assumption:

Anyway, I said "presumably"...."presumably" it refers to ftw because "presumably" if ptw were included, the figure would not be in excess of 40k.

You're presuming that it refers to FTW? Why? Why are you presuming this? The statement I assume you're referring to only says this:

Employees who worked for less than 50 weeks in the reference year are excluded from the calculations for annual earnings.

Based on the detailed explanation for how the figure was calculated do you now agree that this assumption is not accurate?

Benedict

https://www.boards.ie/discussion/comment/118524200#Comment_118524200

I repeat the question. Where did they specify whether who they were talking about was ft or pt when they said "Median annual earnings were €40,579 in 2020"?

If indeed they did state it and I missed it, I don't mind admitting I made a mistake - but I need an answer to that question.

crossman47

https://www.boards.ie/discussion/comment/118524396#Comment_118524396

They didn't. They made it clear what the exclusions were:

For the purposes of this analysis the CSO excluded employees earning less than €500 per annum and employments where the duration was less than two weeks in the year. Also excluded were secondary employments earning less than €4,000 per annum, extremely high earnings values and missing employer and employee reference numbers. Employment activity in NACE sectors A, T and U has also been excluded from the analysis. 

So, the median covers all workers (full and part time) unless excluded as above. As has been explained here a number of times, it cannot be anything else since the relevant Revenue file does not distinguish full timers from part timers.

And in case you come back again, CSO have published mean values for full timers based on aggregate data for enterprises/companies but aggregate data of that type does not allow aanyone to calculate a median.

Benedict

Peregrinius: "They did specify who they're talking about — you just didn't bother to read it. And then crossman47 quoted it for you, and you apparently didn't bother to read it again."

Now Crossman:"They didn't".

So, once again, where was it specified whether "Median annual earnings were €40,579 in 2020" referred to pt of ft?

I have made assumptions - as have others. I had to because the did not specify.

We have my assumptions (which are sensible) and other assumption too. But did they specify? As CM said "They didn't"

moon2

https://www.boards.ie/discussion/comment/118525673#Comment_118525673

"don't feed the troll" comes to mind... But...

There's still no need for guesses or assumptions. The answer to your question has been provided by direct quotation from the publication multiple times. The most recent copy/paste is in the post prior to the one I quoted.

Chips Lovell

Mod Note

I'm closing this thread I don't think the discussion is being advanced any further. At this stage I think the OP's most recent question has been answered at least four times, to little avail.