I had the same reaction, andrewjw. An organization whose key goal is to expand the use of markets starts out looking like it won't be on board with subsidized anything. To the Pioneer Institute'fas credit, it does suggest that the T shouldn't always cut low-ridership bus routes through poor neighborhoods, but instead should find ways to run those routes more cheaply (smaller vehicles, for example).
They have their whole report up on their website. I haven't had time to dig into it, but when I do I'll be looking to see whether their calculations and predictions show that the new fare-collection methods could result in a net loss of $34 million (a year, I assume), or whether they only show that the value of evaded fares could go up by that amount. If ridership increases, and most of the old and new riders pay their fares, the total value of fares evaded could go up by $34 million, but the total value of fares collected (the total farebox revenue) could stay the same. Or go up. Or go down.
Simple example:
100 people currently ride a given system for $1 per ride. 5% of them evade the fares, for fare evasion of $5. Total revenue $95.
Scenario 1:
A new fare collection system has no effect on ridership and costs exactly the same as the current system. 7% of riders evade the fares, for fare evasion of $7. Total revenue $93.
Scenario 2:
A new fare collection system makes boarding a little quicker, so the schedules can be a little faster and the same number of vehicles and operators can offer more trips. Boarding is also more pleasant and wait times are less, making service more attractive. 110 people now ride. 7% of riders evade the fares, for a total of $7.7 in fares evaded; round it up to $8. But 102 people pay the fares, so total fare revenue is $102. The system comes out ahead.
And a whole bunch of other possible outcomes. Just making a big deal about the dollar amount of fares that might be evaded, without considering how many people are riding and paying before and after, could be misleading.
On the other hand, a POP system probably requires a lot of enforcement and fairly high fines, and creating more opportunities for people to commit and then be caught in a very petty offense (when there's already a steady flow of poor people who can't pay things like traffic tickets into jail all over the country) might not be the best plan.