Reader mailbag: Caveats on policy costings

This morning's inbox came with excellent comment from an informed reader about the difficulties in getting reasonable independent policy costings. The email reads (and is shared with permission):

• Do not underestimate how effective agencies can be at blocking information they do not want others to know. One reason the costings will be poor is because all kinds of information will not be unavailable because the agency, not the minister, do not want to be challenged. You will find "privacy" and "computer systems" becomes big issues...

• The main advantage of a single separate agency is consistency in the assumptions. At the moment they change depending on the whim of the needs of a budget bid. If agencies have to write down and be long term accountable for the assumptions, that will be a big step forward in like for like comparison. Even if the costings themselves are no more accurate, they comparison between policy costs (bigger, smaller, etc) will be more robust

• My experience of independent agencies is not good: stats NZ, Office of the Auditor General, Ministry for Women, SuPERU, etc They are all small agencies with some kind of "advocacy" role that they use to justify pursuing their own agenda. Ministerial accountability has its limits, but it is at least real accountability.

I agree with most of this. Privacy seems to be the default "Computer says no" reason for not doing things. Also agree that on the advantages of having a single consistent, if biased, ruler over having multiple different rules with different biases. 

My prior piece had looked at the ways a party might avoid having rigourous costings applied. Accountability for the costing agency would be a fun one - perhaps reinstate betting markets on whether policies wind up costing what they're advertised to cost? Ex-post assessments?