Political interference in ESRI ?

Delancey

Kinda surprised that nobody has started a thread on the ESRI report that claimed 44% of adults with families are better off unemployed than working.
Doubtless a controversial claim but perhaps even more controversial was the unprecedented step by the ESRI to withdraw the report describing it as a '' work in progress '' thus opening up suggestions of political interference.
Richard Tol insists he is standing over his figures - anyone here have any views on whether the ESRI lacks independence ?

http://www.rte.ie/news/2012/0613/author-stands-by-working-paper-findings-on-welfare.html

Sponge Bob

Is Tol not in a new role abroad since last year?

Dotsie~tmp

Sponge Bob wrote: »

Is Tol not in a new role abroad since last year?

Whatever take u have on ESRI interference or Tol credibility the only thing I take is ERSI trustworthiness undermined.

Pandora2

The ESRI are funded directly by the Office of the Taoiseach and I firmly believe someone over there dropped a dime when the media went crazy with this. If only to give someone a b%llocking!

Afaik it is based on 04/05 statistics and I'm not sure how relevant this paper is now, given the economic shift we have endured in the interim.

jimmycrackcorm

I think some doubt has arisen over the report details. Some people were questioning why the figures given, for example detailing employees on work related food spending for example where many would bring in a home made sandwich...

Pandora2

I think what they are getting at is, it was taken for granted those who went to work took the most expensive commute, lunch options and so on and that this is a matter of choice in most cases, not necessity...therefore skewing the figures:o That was my take on it anyway...open to correction

Farmer Pudsey

Pandora2 wrote: »

Afaik it is based on 04/05 statistics and I'm not sure how relevant this paper is now, given the economic shift we have endured in the interim.

If it is it is worse as wages have droped by 20%+ , workers are paying more tax, 2005-2008 were years when welfare got a lot of extra money and it was only cut by 10% post 2009.

So maybe it is 50%+++ family's better of not working now,

Sponge Bob

Probably is 50% + once you also factor in green carbon taxes in 2008/9 and commuting costs. Tol now has a job abroad. This may have been on his desk for years and may not have been fit for review and publication when he was here.

Pandora2

Farmer Pudsey wrote: »

If it is it is worse as wages have droped by 20%+ , workers are paying more tax, 2005-2008 were years when welfare got a lot of extra money and it was only cut by 10% post 2009.

So maybe it is 50%+++ family's better of not working now,

So, depending on clarifaction of the age of the data, it's possibly much talked about recycling fodder:rolleyes:. Look at it this way, if I was sketching out a personal budget, would I use salary/expense figures from 04/05...Nope, it would be highly irrelevant to my life today!!

There are lies, damn lies and then there's statistics:rolleyes:

Farmer Pudsey

Pandora2 wrote: »

So, depending on clarifaction of the age of the data, it's possibly much talked about recycling fodder:rolleyes:. Look at it this way, if I was sketching out a personal budget, would I use salary/expense figures from 04/05...Nope, it would be highly irrelevant to my life today!!

There are lies, damn lies and then there's statistics:rolleyes:

But it will give you a general idea since 2004/5 wages have gone back by 20%, tax and fuel costs have gone up so for the average worker the senario has got signifinitly worse.

Between 2005-2008 social welfare increases it was not until post 2009 that there was any cut in social welfare.It was cut by about 10%.

This report may be inaccurate but it would have widened the gap to make 50%+++ family's better off not working.

If the government could have picked holes in it it would not have been suppressed

antoobrien

The report is linked from RTE

Enjoy

The data used is the 2004/2005 Household Budget Survey (CSO, 2007), which is a survey of a representative random sample of all private households in Ireland. In the 2004/2005 survey 6,884 household participated in the survey which is a return of 47%. The analysis in this paper examines only those in full-time employment and those who are unemployed on either a short-term and long-term basis as well as those undertaking home duties. The self-employed, students, retired and those in temporary employment schemes are omitted for the purposes of this analysis. This yields 4,028 households for inclusion in the analysis.

Is 4,028 out of 1.6m households a big enough statistical sample?

Farmer Pudsey

antoobrien wrote: »

Is 4,028 out of 1.6m households a big enough statistical sample?

Political polls use a smable of 1000 people their health warning is = or- 2% I believe so a sample of 4000 would reduce the possible problems below 1%.

you do not expect them to sample all 1.6 million.

antoobrien

Farmer Pudsey wrote: »

Political polls use a smable of 1000 people their health warning is = or- 2% I believe so a sample of 4000 would reduce the possible problems below 1%.

It's a significantly higher sampling rate that than an opinion poll then.

Farmer Pudsey wrote: »

you do not expect them to sample all 1.6 million.

No, just wondering how it compares to other surveys.

Godge

antoobrien wrote: »

It's a significantly higher sampling rate that than an opinion poll then.



.

And with significantly higher accuracy as a result. Given that opinion polls taken 3-4 days before the election are nearly always very accurate (even though there is time to change minds) I would take it that the sample size is such as to make the survey highly accurate all other things being equal.

You would have to examine the methodology closely to see whether there is any bias. Assuming that regional distribution, household size, household type, income distribution all closely follow the census data, there should be no reason to mistrust the data.

sarumite

Farmer Pudsey wrote: »

Political polls use a smable of 1000 people their health warning is = or- 2% I believe so a sample of 4000 would reduce the possible problems below 1%.

you do not expect them to sample all 1.6 million.

A poll has very few variables. For every new variable introduced, you magnify the associated error. Generally to correct for this, you need a larger pool size, or in the case of many polls widen your statistical significance threshold. Furthermore the polls are usually compiled from simple question, not interpreted from a matrix of different data set, so there is little room for ambiguity. In short, this was study based on statistical interpretation of a two different data sets and not a poll.

Farmer Pudsey

sarumite wrote: »

A poll has very few variables. For every new variable introduced, you magnify the associated error. Generally to correct for this, you need a larger pool size, or in the case of many polls widen your statistical significance threshold. Furthermore the polls are usually compiled from simple question, not interpreted from a matrix of different data set, so there is little room for ambiguity. In short, this was study based on statistical interpretation of a two different data sets and not a poll.

When you talk about variables say the survey has a 10% error rate this means that the number of family's that are better of on welfare could be as little as 38% or as high as 46% that is with a 10% variation.

The reality is some posters want to shoot the messenger when some poster's refer to a political poll having simple questions I imagine that the Professor Tol paper is a reserch paper so that they would have gone throught the finiancal data of the different respondents.

If you go through the Irish welfare system it is very easy to see how some family's can do very well off it. It has not changed a whole pile in six years and I find it a bit disgenous that some posters are trying to rubbish it.

First they tried to use the age of the reserch paper to rubbish, in other threads some are questioning Professor Tol himself and now they use the sample size. It may be inaccurate but it is not that inaccurate as a 10% variation shows.

daltonmd

Pandora2 wrote: »

The ESRI are funded directly by the Office of the Taoiseach and I firmly believe someone over there dropped a dime when the media went crazy with this. If only to give someone a b%llocking!

Afaik it is based on 04/05 statistics and I'm not sure how relevant this paper is now, given the economic shift we have endured in the interim.

Actually Pandora, I'd say it has more relevance now than it did in 04/05, given the paycuts, introduction of the USC, cost of petrol etc.

Sand

Stephen Kinsella and Brian Lucey have made some points on this at IrishEconomy and their own blogs - firstly, its not an ERSI report, its a working paper - which is released with the basic concept of "I'm working on this - take a look and knock it down". Its thrown up to invite criticism to ensure the idea is sound (something I'm sure is entirely alien to Official Ireland - criticism...sound ideas...whut?).

This makes the ERSI decision to remove the paper all the more embarrassingly stupid. It implies that the ERSI will only allow the hosting of working papers where they approve the findings - this is awkward because the ERSI disavow responsibility for the views of the authors in its actual quarterly reports.

Do I believe someone picked up the phone in Enda Kenny's team? Nope - the best sort of censorship is self-censorship. When not only are groups prevented from releasing views or reports unapproved, they themselves refuse to release or report anything they fear might be unapproved. In the history of the ERSI, its disagreement or criticism of government policy has always been more noteworthy due to its rarity.

The ERSI is part of Official Ireland - they panicked when they realised they had offended the group think about "the most vulnerable in society". Hence their over-reaction.

Tol's theory is itself sound - given transport and childcare costs, there are large costs to going to work everyday. Wages have gone down, welfare has remained relatively steady ("most vulnerable in society..") so logically a large proportion of people with kids would be better off on the dole.

sarumite

Farmer Pudsey wrote: »

When you talk about variables say the survey has a 10% error rate this means that the number of family's that are better of on welfare could be as little as 38% or as high as 46% that is with a 10% variation.

Thats actually not what I meant by variables. Asking a person who they would vote for is one variable. Asking a person what are their main political concerns is another variable. Each variable has an error associated with it. If you draw a conclusion based on both variables (say person x votes for party y because of reason z) then you now have a statistic which has its own error and will be affected by both the errors from each of the two variables. The more variables, the greater the possible error.

The reality is some posters want to shoot the messenger when some poster's refer to a political poll having simple questions I imagine that the Professor Tol paper is a reserch paper so that they would have gone throught the finiancal data of the different respondents.

I don't think anyone has taken any umbrage with the messenger. Professor Tol is not the messenger, he is the author of the message. I am unaware of whether Prof. Tol paper was peer reviewed or not. That would clearly put some weight behind his paper. However, that wouldn't detract from the fact that a simple poll is not the same as a statistical study.

If you go through the Irish welfare system it is very easy to see how some family's can do very well off it. It has not changed a whole pile in six years and I find it a bit disgenous that some posters are trying to rubbish it.

non sequitor.

First they tried to use the age of the reserch paper to rubbish, in other threads some are questioning Professor Tol himself and now they use the sample size. It may be inaccurate but it is not that inaccurate as a 10% variation shows.

You don't think the period of the data or the sample size is important? You would find research papers fail peer review for a lot less.

Farmer Pudsey

Professor Tol produce a reserch paper it is that a reserch paper. The Ersi usually produce these publish them ane wait for other reserchers/acdemic's to produce papers to agree/disagree with the findings.

The main finding is that for a family unit on average it costs them 10,000 euro to go to work in food, transport and childcare costs and because of this 42% of families would be better off on welfare.

Now in 2004/5 wages were higher, tax lower, fuel and car costs lower, childcare and food costs neutral. However welfare and welfare benfits increased sharply from 2005-2008 before about 10% cut post 2009.

So the gap has widened if anything now below is a costing that I have done in another thread.

Young couple with two childern one 2 years of age one school age.

They may need two cars one can be a old car and one fairly new
if unemployed only one car needed. I will put extra car cost to cover tax, insurance, extra maintenance all comes to 1500 euro's. Extra fuel 50 euro's/week so 2500 euro.

Clothes to go to work lets put it at 350 euro's for the two. We will allow them 2 euro's/day each for lunch so 1000 euro's a year. Because of being home late in the evening have to buy semi-prepared food 4 euro's/day so another 1000 euro's

Child care costs for childern grannie helps out but it is still 75 euro's/week
comes to 3750 euro's.

So the cost of going to work is arounf 10K for these family's

creedp

Farmer Pudsey wrote: »

But it will give you a general idea since 2004/5 wages have gone back by 20%, tax and fuel costs have gone up so for the average worker the senario has got signifinitly worse.

The report is full of sweeping generalisations (just like the above quote re: wage have gone back 20%) and worse still out of date generalisations and therefore should be updated before it is taken seriously. Having said that I have not doubt that there are many cases where families, especially those with young children, would be better off on the dole. Its a pity that the report wasn't revised prior to being released as ita about time the plight of people in this position was highlighted.

KyussBishop

I did a small amount of research on the ESRI in the last couple of days, and from what I can tell, they are an almost 100% discredited organization whose views/reports should probably be totally ignored/dismissed, and not given any credibility/attention.

I know that may lump a number of potentially worthwhile reports in with the bad, but I think anyone working with them with a clean reputation and who can do worthwhile research, should probably look to work elsewhere and not tarnish their name through association with the ESRI.

Joe1919

The problem of 'work incentive' is not new and every labour economics textbook will have a chapter or two on the problem. (They generally show an inverted U shape type graph to show that it is also wrong to put too much pressure on people to work, as it leads to people taking up employment that does not match their skills and interests and this lowers productivity and standards.)

Interestingly, I can recall an ERSI working paper by Nolan and Callan as far back as 1996 that suggests that there may be no financial work incentive for up to two-thirds of married families with young children .
http://www.google.ie/search?ie=ISO-8859-1&hl=en&source=hp&q=improving+work+incentives+callan+nolan&meta=&gbv=1&btnG=Google+Search

Farmer Pudsey

Then why not let the paper out there and let other reserchers/acdamics tear shreads out of it. The reality is that the paper hit a bone with official Ireland. If it was left in place the Government would not be able to increase taxes workers and have had to reduce welfare amd welfare benfirs rapidly this will not happen now.

Alot of poster are rubbishing the report by discrediting the ERSI and or the report it self. It would be intresting if they would give what they think what it costs the parents of a young family to go to work or how much it cost a single worker to go to work

Head The Wall

Joe1919 wrote: »

The problem of 'work incentive' is not new and every labour economics textbook will have a chapter or two on the problem. (They generally show an inverted U shape type graph to show that it is also wrong to put too much pressure on people to work, as it leads to people taking up employment that does not match their skills and interests and this lowers productivity and standards.)

I fail to see how that same person is more productive while on the dole doing nothing and getting looked after by the state for doing so.

Joe1919

Head The Wall wrote: »

I fail to see how that same person is more productive while on the dole doing nothing and getting looked after by the state for doing so.

The potential is always there for someone to do more harm than good. It is better from the employers point of view to have a smaller but more commited and better matched workforce. Also a person who is unhappy in their job has a greater potential to do harm. (For example, it took only one incident of food contamination to cause massive disruption of Irelands pork and bacon industry a few years ago.)

This may have been a huge problem under the old communist systems. They may have had zero unemployment but the workforce may have been badly motivated and many of their products were inferior. (Think of all the jokes about Eastern European cars in the 80s)

Quality counts as well as quantity.