MSD using data

February 20th, 2014 at 12:00 pm by David Farrar

CIO reports:

Paula Bennett, Minister for Social Development, reveals the critical role data analytics is playing in reducing the number of people on welfare, and deciding which programs to target vulnerable citizens, especially children.

Analytics has helped in investing and using this data for positive action and make a massive influence in the lives of New Zealanders.

“Without using data analytics, we will be throwing a lot of money,” she says in her keynote the SAS User Conference in Wellington. “We will not be getting into the heart of the problem, we will not be putting resources and investment where these are needed,” she says. 

“Since 2011, analytics has been at the heart of our welfare reforms,” says Bennett. While cognisant of the criticisms against her when the welfare reforms were announced, she lists the compelling figures on how the government applied analytics to determine funding for programs and beneficiaries.

A key feature used by the government is analytical technical segmentation, she says. This looked at the welfare system at macro level, an overall evaluation, and how much it is costing over a lifetime of the current population.

We spend an average of $22 million dollars a day on welfare, $8 billion a year, she says. The projected lifetime cost for people on benefit was $78 billion.

This baseline valuation of the welfare system is at a level that has not been attempted anywhere in the world, she says. “It is a pretty compelling story on why we want to look at this population to see what we can do,” she says.

The way MSD is using data analysis is gaining attention around the world. The lifetime cost and investment approach is leading edge stuff, and there is a lot of interest from other countries about how NZ is doing it.

This doesn’t necessarily mean spending less in the short term. To the contrary it has often meant spending more to help people into work, as the short-term cost of that assistance is less than the longer term cost of them remaining on welfare.

Tags:

7 Responses to “MSD using data”

  1. tvb (4,422 comments) says:

    This is swamped in jargon and explains nothing. An illustration would have been helpful.

    Vote: Thumb up 1 Thumb down 7 You need to be logged in to vote
  2. Mobile Michael (451 comments) says:

    I understand from a former Minister of Employment that Treasury would argue that the cheapest option was to do nothing. Beneficiaries would either have the skills to move off welfare when the circumstances that got them on were changed again, or that it was cheaper to keep paying benefits than develop skills to make them employable.

    I am glad that the current Government rejected that view and worry about Labour snd the Greens returning to power and dismantling the very good work of National and its other support parties.

    It seems that analysis like this will make reforms harder to undo, so good work MSD and Minister Paula Bennett.

    Vote: Thumb up 5 Thumb down 0 You need to be logged in to vote
  3. Harriet (4,972 comments) says:

    “……This doesn’t necessarily mean spending less in the short term. To the contrary it has often meant spending more to help people into work, as the short-term cost of that assistance is less than the longer term cost of them remaining on welfare…The lifetime cost and investment approach is leading edge stuff……”

    It’s not leading edge. Cost analysis of the unemployed has been around for decades. Or have Treasury just been telling us lies over the years?

    The government has also been investing in employment/training programmes too, in the full knowledge that it is profitable to do so for all those concerned. In the medium term and the long term.

    This is just the case of using a computer with complex mathmatical equations, and applying some flash words to give the impression that a different answer is popping out. It isn’t. It’s just worded differently.

    Vote: Thumb up 1 Thumb down 0 You need to be logged in to vote
  4. peterwn (3,273 comments) says:

    I have wondered about this in the health sector. Primary health, hospitals and Pharmac seem to run as isolated fifedoms, each with its own budgeting, key performance indicators, etc. For example a new drug becomes available which is moderately expensive but can avoid surgery. If a data mining exercise shows the drug is cost effective overall, then Pharmac could be given an allocation to reflect this.

    Similarly with allegedly substandard homes and medical costs – which is more economic?

    Vote: Thumb up 2 Thumb down 0 You need to be logged in to vote
  5. DRF (19 comments) says:

    My impression from attending the presentation is that the answer is actually quite different. That is to say, it actually downplays the value in unemployment intervention as the real damage is coming from other benefits that people get onto and stay on much longer than the unemployment benefit, which actually is more transient and operating as you might hope.

    So the focus shifts very much onto those starting on the benefit before the age of 20, particularly female.

    Again not sure why they needed a computer model to tell them that but I have certainly been abused mightily over the years for pointing out that that’s where I would focus my efforts. Not just for the monetary benefit but also to stop the abuse.

    Vote: Thumb up 2 Thumb down 0 You need to be logged in to vote
  6. SPC (5,619 comments) says:

    To put it simply they are focusing on work capability. Not being work capable leads to dependence/poverty.

    I may be a leftie, but I did support this approach in my submission on welfare reform. For those on the DPB this includes access to affordable child care and after school programmes. It also means we invest in drug rehabilitation (for inmates and those incapacitated from work).

    Similarly healthier homes and ECE participation will improve educational outcomes etc.

    This should not be a right left issue.

    Treasury is aware that work for the dole cost money and there was no improved employment outcome over those not on the programme. And that training programmes for the long term unemployed only work when there are jobs for them to then go to (so subsidising placements in on the job training is more effective – apprenticeships work) as they will still be the last preference from the labour pool.

    Thus some contesting of the concept of investing in work capability will come from them, as it should.

    Vote: Thumb up 0 Thumb down 0 You need to be logged in to vote
  7. Psycho Milt (2,412 comments) says:

    The projected lifetime cost for people on benefit was $78 billion.

    It strikes me as utterly pointless to work out the “lifetime cost” of beneficiaries when relatively few are on one for their lifetime. Is this “lifetime cost” the cost for the projected averge amount of time a person is on a benefit during their lifetime, or is it one of those bullshit accounting things in which you calculate how much you’d be liable for if each current beneficiary remained one for the rest of their lives?

    Vote: Thumb up 0 Thumb down 1 You need to be logged in to vote