Golden voodoo economics

Ian Harrison of Tailrisk economics looks at the economics of the Wellington Golden Mile revitalisation proposal. He notes:

This paper reviews the cost benefit analysis of Lets Get Wellington Moving’s (LGWM)
Golden Mile ‘revitalisation’ proposal. It is presented in the paper ‘Economic
Assessment of the Preferred Option’ that was prepared by MRCagney (NZ) Ltd, an
Auckland transportation consultancy.
Their conclusion was that the costs (capital and increased maintenance) were $86
million and the present value of the benefits was $399 million. The net benefit is
$313 and the benefit to cost ratio was 4.6.1

This looks too good to be true, and it is.
The benefits are substantially generated by a ‘pedestrian realm’ benefit. Pedestrians
were assumed to be prepared to pay $ 247 million just so they did not have to walk
alongside cars on the Golden Mile. This was an absurd result. When we inspected
the Waka Kotahi modelling that MRCagney used to generate this number we found
that it was not fit for purpose and and should not have been used.
On the other hand the street closures were estimated to cause almost no
congestion. Each vehicle journey would take an average of only about three seconds
longer.

It’s ridiculous. This smells like just manufacturing as big a number as possible to please the decision makers.

Then there were a string of omissions, errors and optimistic assumptions that all
generated positive outcomes for the project.
Once these and other issues are addressed a very different picture of the benefit
cost ratio energes. Our estimate is 0.38. The costs well exceed the benefits and the
economic loss is $ 121 million.

And Harrison pulls apart the details:

The per journey time impact and the reduction in the number of journeys do not
seem to jell. Why would 10 percent of car journeys be abandoned when the
journey time has gone up by less than three seconds?. If the fall in demand
estimate of 10 percent is correct then this would imply a travel time cost of $500 –
600 million.
There is no explaination in the paper of why there is 10 percent reduction in
journeys when the underlying model reports a 2 percent reduction. We recently
attempted to get an explanation from MRCagney. They did not provide one.

Garbage in, garbage out.