Category: Uncategorized

Thoughts on investment evaluation

I recently finished reading Brian Trelstad's article, "Simple Measures for Social Enterprise" in the MIT Innovations Journal, where he discusses how the Acumen Fund evaluates its investments.  I had several thoughts when reading the article, most of which I hope to have the time to write up on the blog.

Are there cases when you do not need to evaluate the impact of your efforts?  

I can imagine several cases.  Say, that you are deciding between ten potential investments.  One has a positive net outcome on your target population, the others have no impact on net.  You have $1000 and can invest only $100 or nothing into each investment.  Ex ante evaluation of your efforts is ridiculously expensive, at $900, but reveals perfectly information about each investment’s outcome.  Would you pay $900 to have that information revealed?  If you do, then you will end up paying $900 for due diligence and $100 into the actual investment, for a total of $1000.  However, you can also choose to pay $100 for each of the 10 interventions, for a total of $1000.  

This example is silly because it assumes a very expensive due diligence process, investments that have either no impact or one positive impact, and the ability of due diligence to reveal perfect investment outcomes.

However, the example does point out the tension between investing and analyzing the impact of your investment.  It also suggests that, sometimes, impact analysis is not necessary to reach the ultimate goal of revealing the impactful investment.  

Research and other jobs at Harvard

I often wish that the Kennedy school offered students a resource for finding part time jobs while they’re at school.  This blog post is an attempt to compile all the different job posting locations for graduate students on campus.

  • JACK.  The Kennedy school job portal occasionally, although not often, features job postings from HKS centers and other locations on campus.
  • Employment@Harvard.  Harvard’s official employment website is the largest resource of job sites around the university.   Most of the postings on this site are for full time positions, but I have seen about a dozen new postings per month for part time positions open to students.  The site is divided into an external and internal component.  Always search the internal site.  It contains all the external positions plus the positions that are open only to internal candidates; HKS students qualify as internal candidates.
  • The HLS Administrative Updates blog constantly advertises research assistant positions with law school professors.  Many of the research positions do not require, although often prefer, law school students.  I know of several students with no legal training who have found jobs there.  
  • The Harvard Economics Department maintains a list of job postings for internal and external economics positions.  Some are for research at think tanks in DC or in Boston; others are for research assistant positions; yet others are for course assistance.

Here is a listing of jobs pages, if they are available, at HKS centers.

  • Ash Center — Does not have an official job page.  Possibly advertises for RA and intern opportunities on Jack or HKS Today.
  • Belfer Center — Does not have an official job page.  Some of its projects and programs, which largely act as independent entities, post job offerings on their parts of the Belfer website.
  • Carr Center for Human Rights — Has a dedicated internship page here.  Also lists positions open to students who are eligible for work/study on a separate page here. 
  • Center for International Development — Does not have an official job page, although it looks like they would advertise for jobs on their Student Programs page.
  • Center for Public Leadership — Advertises for jobs here.
  • Edmond J. Safra Foundations Center for Ethics — Advertises for jobs here.
  • Hauser Center for Nonprofits — Doesn’t have an official job page, but it looks like they would advertise for jobs here.
  • Institute of Politics — Has a great research assistant page here.  Not only does it offer an application for students to apply for RA positions, but it also lists current RAs and the professors with whom they are affiliated.
  • Joint Center for Housing Studies — Is not advertising for positions now, although it looks like if they did it would be on their Student Opportunities page.
  • Malcolm Wiener Center for Social Policy — Doesn’t have a job page, although it looks like they advertise for jobs directly on their home page.
  • Mossavar-Rahmani Center for Business and Government — Has a dedicated page for student jobs here.
  • Rappaport Institute — Couldn’t find a job page, but they offer research funding here. 
  • Shorenstein Center on the Press, Politics, and Public Policy — Has an RA application here, and describes that application here.
  • Taubman Center — Doesn’t have a dedicated job page, but discusses fellowships here, so they might also use that space to advertise for RA and part time positions.
  • Women and Public Policy Program — Has a dedicated internship page here.

Compromises

Obama and Leaders Reach Debt Deal
By CARL HULSE and HELENE COOPER
Published: July 31, 2011
http://www.nytimes.com/2011/08/01/us/politics/01FISCAL.html?hp

The fundamental, albeit ironic, feature of a compromise is that every participant to it looks at the finished document and thinks, “This can be improved to favor me.”  The compromise is, by definition, worse for each participant than the best deal that participant can get if he or she crafted the deal alone.  For each participant, therefore, there is another deal that he or she prefers to the one being signed.

How is it possible, then, for a negotiator who enters the discussion with the desire to get the best outcome to settle for what he clearly believes to be the second best outcome?

I guess during the course of the discussions, the negotiator comes to realize that the first-best outcome is impossible to achieve, because others, facing their competing first-best outcomes, will never concede too much.  The negotiator is then faced with the decision between no deal and the second-best deal.  If the second-best deal is preferable to no deal, then he agrees to the compromise.  If the second-best deal is worse than no deal, yet the first-best deal is better than the no deal, then the negotiator continues fighting; but this case is likely rare.

But how does the negotiator come to know that the others will not “concede too much”?

I think negotiations that end in compromises are, at their base, learning exercises in which each negotiator not only learns the contours of the other negotiators’ first-best agreements (and in rare cases find out their first-best agreements entirely), but also gains an insight into how far they past their first-best scenario they can be pushed before they start resisting.  That resistance, in fact, controls the outcome of the negotiation much more clearly than does some understanding of the others’ first-best outcomes (which, I believe, comes secondary).  If your opposing negotiator is will to give you more of what you want and get you closer to your first-best outcome, you’ll take all that he will give (everything else being equal), since your goal is to attain your first-best outcome.  The less your opposing negotiator gives, the less you can take without a fight.  The settlement, then, is controlled by the give-and-take of each negotiator.  

But how much education really happened here?

Perhaps in typical congressional negotiations, little education takes place because each participant already knows so much about his own team and the opposing negotiators.  Take Biden, who has spent decades in the Senate and can pick up the phone and talk to McConnell like they were old pals, albeit at different sides of the aisle.  But this negotiation different in that respect because Obama was somewhat new, the Tea Party Republicans were somewhat new, and many were new to each other in the context of such a high-stakes issue.  Previous negotiations, especially over the healthcare bill, were high-stakes.  But these hit on spending and taxing much more directly than those.   And the Tea Party candidates saw this as their chance to take a stance much more clearly than they saw the healthcare bill — plus, many were just settling in at the time.

So this was a remarkably educational moment for both sides — and the education that took place was one of understanding not only the contours of the first-best outcomes of the TP candidates, but also how much they were willing to budge with respect to their desired outcomes.

I bet Boehner will not soon forget their reluctance to fall in line.

Corporate form and pay-for-performance contracts

“The Case for For-Profit Charities”
Anup Malani and Eric Posner
http://www.virginialawreview.org/content/pdfs/93/2017.pdf

Eric Posner (the son of Judge Posner) writes about an interesting discrepancy in tax treatment of for-profit and nonprofit firms. Nonprofits that perform social services, meaning services that are not designed primarily for the purpose of generating profit for the firm’s owners, receive a tax break. When nonprofits engage in commercial services, however, that tax break is withdrawn; they are treated, effectively, as for-profit firms. For-profits that engage in social service work, however, do not receive the same tax breaks that nonprofits do when they engage in the same activity. The tax breaks ascribe to the corporate form rather than to its function. Posner argues that the function should control the tax breaks.

However, as Posner points out, some for-profits receive financial incentives from the government for delivering some public goods. For example, the government subsidizes alternative energy producers.
It’s interesting to examine how the agency theory, which Posner describes in the article, pertains to social impact bonds. Agency theory says that in a model with an entrepreneur, a donor, and a beneficiary, if the entrepreneur runs a for-profit firm, then the money she she has the incentive to keep as much as possible of the money she receives from the donor to deliver as a product to the beneficiary. If the donor gives the entrepreneur $100, with the understanding that the entrepreneur will keep $10 for herself, spend $10 on operations, and deliver $80 to the beneficiary, then the entrepreneur wants to keep more than $10, which means either delivering less than $80 or spending less than $10 on overhead, or both. In a nonprofit, however, the entrepreneur does not get to keep the profit, and therefore has less incentive to underdeliver on the services that she provides to the beneficiary.

Posner writes about the problematic agency relationship: “Technically speaking, this means the entrepreneur sells a product (transferring charitable money to a beneficiary) whose quality (getting 80% to the beneficiary) is nonverifiable, that is, cannot be stipulated in a contract that is enforceable by a court” (p. 2032).

Now imagine an entirely different model, one describing ex-post payment for performance.  In this model, the donor announces his intention to pay to any entrepreneur $100 for each $80-worth of service that she can deliver to the beneficiary.  One example can be a donor willing to pay a company $100 for each $80 vaccine that the latter will purchase, deliver, administer, and document in a developing country.  The entrepreneur has not yet incorporated her efforts in either a nonprofit or for-profit corporate form.  The question of which form to choose now seems less relevant to her.  If she thinks that she can carry out the task by spending, say, $17, then she should expect to get a profit of $3.  If this profit is above her expected opportunity cost (presumably she is choosing from among several projects), then she will take the donor up on his offer; the lower she expects her overhead to be, the greater her profit and the more incentive she has to take the offer.  She should become agnostic about corporate form, since she will simply set her salary at the $100 less the expected overhead.  The only concern is that the salary will appear excessive.  But that seems to be the donors concern, since he is the one who would have lost money he could have otherwise saved by mispricing his offer.

While the entrepreneur is now agnostic about corporate form, the donor is not.  The donor can deduct his gift to a nonprofit, but not to a for-profit, and therefore prefers that the entrepreneur’s efforts take the latter form.