Tuesday, January 12, 2016

Why suppress evidence of welfare abuse?

Bear with me while I go back in time. The relevance will become apparent.

In 2011 work was undertaken on behalf of the Department of Labour (now the Ministry of Business, Innovation and Employment or MBIE) which involved matching data from the Household Labour Force Survey or HLFS (the official source of our unemployment rate) and Ministry of Social Development benefit data.

If individuals on the HLFS also receiving a benefit were providing accurate information to both departments - Statistics New Zealand and MSD - the data should provide no surprises, right?


The major finding of the paper was that, "About 40% of people on work-tested benefits may not be meeting their labour market obligations, as they appear to be either working too much or searching too little."

Despite being expected to maintain an "unrelenting focus" on getting work - as prescribed by law -  1 in 3 unemployment beneficiaries reported to the HLFS no job search activity in the past 4 weeks; 1 in 5 reported no job search activity and no intention to seek work in the next year.

At the other end of the spectrum, "1 in 10 people being paid an unemployment benefit report to the HLFS that they are working more than 30 hours per week."

"The inter-temporal pattern suggests that as the unemployment benefit numbers decline in good times, the overall stock of unemployment beneficiaries contains a larger proportion of people who are not really seeking work. When unemployment numbers rise in bad times, there are a higher proportion of genuine job seekers among the stock of unemployment beneficiaries."

So 10 percent of people claiming an unemployment benefit were working full-time (a further check against PAYE records showed 4% earned in excess of $2,000 in the reference month); a third hadn't fulfilled their obligations to look for a job in the past month; and a fifth had no intention of looking for a job. That's what the beneficiaries told the interviewer from Statistics New Zealand anyway. What motivation would they have to lie? None that occurs to me. Though they could be motivated to lie to MSD in order to continue to receive benefit money.

Now here's the odd thing. The findings were never released.

The existence of the research only became known to me when the author referenced* it in a later publication. He did so because he believed the paper had been published. He left the Department of Labour after completion understanding that the paper merely required a ministerial briefing then would be made public. For that reason (and because the taxpayer had funded the work, he said)  the author supported my application to the Ombudsman to compel MBIE to release the paper after their initial refusal to do so. That process took over one year.

So why weren't the findings made public, and why the resistance to release the paper to me? I can only speculate as to why.

But here goes.

The findings were politically sensitive. They would be a bad look for a National government claiming to have tightened up on welfare, and powerful ammunition in the hands of the opposition. Possibly. But the data matching exercise extended back into the Labour years and the abuse levels then were, proportionately, even higher!

It is more likely that the government didn't want media broadcasting their data matching exercises far and wide. No, law-abiding Joe Public wouldn't particularly object. The 'nothing to hide, nothing to fear' attitude is rife in this relatively corruption-free country (rightly or wrongly).

But imagine a beneficiary reads or hears about how a survey they are being forced to participate in is being checked against their Work and Income records. For the welfare abuser, that would merely tip them off to lie more consistently to government departments.

Data-matching is being used increasingly but its effectiveness lies in keeping the public in the dark. There's an irony at work. Non-transparency is required to improve integrity of systems.

So ultimately that's where I find the most convincing rationale. But that leaves me with a dilemma.

As a long-time critic of the welfare system, the findings vindicate or illustrate my concerns about the rampant misuse of the system (which hurts  genuine beneficiaries and the taxpayers funding it). Do I want to make a song and dance about these findings though, if the information acts to assist those with the worst motivations?

Because of the time lapse is the issue even newsworthy? I ran it past a couple of wiser heads than mine and they thought it was, but the two parliamentary journalists I ran it past didn't.

You will form your own opinion.

There is no doubt that the National government has been more vigorous about detecting and stopping fraud but against the above revelations the numbers still look inadequate. From parliamentary oral questions in March last year ( note Jo Goodhew's description of the numbers as "a small minority"):

Benefits—Savings 6. Hon JUDITH COLLINS (National—Papakura) to the Associate Minister for Social Development: How much has the Government saved as a result of its benefit fraud initiative?
Hon JO GOODHEW (Associate Minister for Social Development): Since benefit fraud reform initiatives began 2 years ago we have saved the taxpayer over $60 million in future benefit payments. Only a small minority of beneficiaries take money they are not entitled to, but those who do cost tens of millions of dollars each year. These changes make it difficult to defraud the welfare system and hold people accountable for their actions.
Hon Judith Collins: How is the Government encouraging beneficiaries to comply with the welfare system?
Hon JO GOODHEW: Over the past 2½ years around 9,500 benefits have been cancelled after fraud was discovered. We expect to see fewer cases of benefit fraud as our case officers continue working closely with clients to ensure they declare their income and any changes to their relationship status. We have also identified 3,000 clients who have previously committed fraud. By managing these clients more closely, we can help to ensure that they do not reoffend. 

If the 2011 findings held true for current Jobseeker beneficiaries, as many as 48,000 people may be either failing to meet work-search obligations, or already working full-time and continuing to claim a benefit.

*The reference also stated, "... about 10 per cent of people whose welfare records showed that they were receiving a DPB reported being partnered or living as married." That data, however, was not included in the paper released to me.


Eric Crampton said...

So long as beneficiaries never are actually punished for having answered honestly on the surveys, then that rationale evaporates. And Stats would be good at ensuring that IDI linkages can't be used for such purposes.

Maybe you could get a mechanism where MSD is able to do predictive modelling on who'll cheat based on the linking of HES/HLFS data with MSD data, then run some checks on those at high likelihood of cheating as predicted by those models, while dropping the cases that allowed the predictions to be made but applying the model again to the full sample in a few years' time - when the two-years-ago cheats' background administrative data would have changed enough that they wouldn't perfectly predict cheating.

Jim Rose said...

Well done. I worked with the author of the study for a while. He is very smart.

JC said...

So basically you need to take the figures, work out a rough cost to the taxpayer and extrapolate them to the present day and see what the damage is..

If its (say) more that the Govt's trumpeted savings then (to me) its worth publishing. Also, the accumulated cost would be useful too because there may be hundreds of millions involved..


Anonymous said...

I thought the government had always matched welfare recipients with IRD records. Is that not the case?

Anonymous said...

This - is why I always regularly check your blog Lindsay. You always do classy analysis.

Thank you.

This is what happens after the first few years of the first term of a government. Officials learn how to 'manage' their ministers and ministers become less willing to roll out damaging data.

It's only ever 'clean' for the first two years of every new government when everyone is feeling their way and generally playing by the rules - after that the gaming starts. This is why we so desperately need a change of government - the problem is the current opposition is chaotically broken and showing no signs of getting their act together and so is unelectable.

What happens next is the government thinks they can't put a foot wrong and keep on straying further and further of the path of good intentions and all of a sudden the electorate snap and are so repulsed they will vote in even the most demented of oppositions.

In turn that opposition will mistake revulsion of the outgoing government for a mandate and they'll irrevocably break the country (as we all know a Labour/Green/Mana/Internet would have done. Look at Australia post Howard to see how it's going to look for us.

Anonymous said...

All benefits are fraud - fraud on hardworking taxpaying citizens. Frankly for these kind of surveys a 5-10% error rate is quite likely. When will even ACT campaign on ending all benefits?