Opinion: Taking the next step after New York City’s Intensive Mobile Treatment initiative

After reading last week’s New York Times article about the Intensive Mobile Treatment Initiative to treat severely mentally ill people on the streets and New York City Comptroller Brad Lander’s associated audit and e-blast I was left wondering why the city doesn’t just drop IMT and adopt a program I’ve read about that seems effective. A program where a majority of those who start homeless are housed; only 2% of the housed lose their housing; two-thirds receive treatment most of the time; three-quarters take their meds, and only 20% are re-incarcerated.

This other program I’ve described sounds like a success versus IMT, right? But I’ve tricked you because the program I just described is IMT and the results I listed above come straight from the appendix in the comptroller’s audit.

So, is IMT a success as my reading of the appendix suggests? Is it a “mixed success” (per the audit), “increasingly failing” (per the later e-blast), or a failure (implicit in the New York Times reporting)? Or is IMT actually “a uniquely effective program” as it was characterized to me by the leader of a nonprofit that delivers the program?

The simple answer is that we have no idea. Nothing in the audit gives us even the slightest clue. Worse yet, we would probably still be clueless even if the audit recommendations were adopted since they are so focused on reporting, reviewing, and submitting information about activities. On its own, reporting on activities does nothing to understand results let alone make them better. (A better speedometer does not help a car go faster and it says nothing about how fast the car could go.) In fact, there is a possibility that more reporting could make outcomes worse if it consumes time and money that would otherwise have been spent on the program.

None of this is news to nonprofit leaders. None of this is specific to the IMT program. But it is still a depressing reminder of how little difference the entire lather-rinse-repeat cycle – journalists investigate a high-profile issue and generalize (maybe fairly, maybe not) from a small number of incidents; public officials get defensive and issue (or reissue) reports that blame some agency (in this case DOHMH); the agency adds incremental reporting burdens (without incremental financial support) on its nonprofit partners – makes to the lives of real New Yorkers. Everybody involved knows this, yet they each have their assigned role to play.

Instead of backward-looking, accusatory, reporting-obsessed audits, I dream of future-looking, constructive, outcome-focused reviews that could lead to better results. My dreamed-of-reviews would cover five elements: reporting, baselining, systems, incentives, and context.

1. Reporting: Government agencies must monitor how tax dollars are being spent so a review needs to ensure that the reporting is useful, timely, and adhered to. But the analysis of the reporting should not focus only on adherence, but also on relevance and practicality which raises the possibility that a review could recommend reduced reporting in certain instances. The reporting should clearly distinguish activities from outcomes.
2. Baselining: Nonprofits work with government agencies on very difficult problems. If good outcomes were easy, these problems would already have been solved, for-profit companies would be involved, or the government would provide the services directly to take credit for the results. A review should be clear how definitions of program success – below expectations, meets expectations, exceeds expectations – have been baselined in comparison to background research/evidence base, the relative performance of nonprofits within New York (i.e. grading on a curve), or results achieved in other similar geographies. A review should be honest when none of these are available (and often they are not) but suggest ways that a reasonable baseline might be developed over time.
3. Systems: Human Services are delivered by nonprofits working in partnership with the city so their results depend in part on the dynamics of this partnership. A review should examine the important elements of this partnership including the timeliness of contract registrations, the timeliness of payment, the staffing levels in the city agency, the staffing levels in the nonprofits, and the sufficiency of the rates.
4. Incentives: Implicit in current audits is the assumption that more reporting will lead to better results through some wonderous, unexplained mechanism. Which raises the question: why aren’t the nonprofits already getting the better results?  After all, the primary reason nonprofits are considered ideal partners to government in delivering human services is that they have strong, mission-driven reasons to perform and no profit-driven reasons to shirk. Moreover, the particular organizations delivering a program have been selected through a competitive RFP process. Given all this, a review should provide an explicit hypothesis about why the nonprofits are not already delivering the goods. There are four - laziness, ignorance, insufficient capacity, and lack of flexibility – each of which suggests a different corrective “protocol”. For laziness, the threat of contract termination or nonrenewal. For ignorance, more information. For insufficient capacity, more money. For a lack of flexibility, more latitude to adjust the program. A review should be explicit about which of these explanations seems most plausible. 
5. Context: A review should put the program in historical context and the context of the city’s current needs.  Is it new and novel or long-standing and well-understood? How are the needs it addresses changing? How large is the unmet need? 

None of this will happen anytime soon but we can’t just accept at face value or dismiss out-of-hand articles and audits criticizing nonprofits and their government partners (most are critical, or they would not make the news). We need to take the time to study the appendices, addendums, and footnotes. And we can’t bemoan the current system without reflecting on what would constitute a better one.