How Does Hope Work?

Does hope need to be based in reality for it to help facilitate movements out of poverty?  Recently my co-authors and I (Travis Lybbert and Irvin Rojas at UC Davis) have been reviewing the 12‑month follow-up results from our study on the impact of hope and aspirations, what we call the Oaxaca Hope Project. What we see in our results have caused me to reflect on the nature of hope and how it functions among the poor.  Much of the new research on hope and aspirations seems to be rooted in the idea that increases in hope alone can be transformative.  A number of questions have been brought to mind, including whether hope by itself can lift people out of poverty, or whether hope must be based in some existential reality—such as a new opportunity or acquisition of a new skill.  If the latter is true, then when poverty organizations refer to “bringing hope to the poor,” it may be the case that hope is merely a by-product from the optimism created from a more tangible intervention like an improvement in health, education, infrastructure, or microfinance.

In our randomized controlled trial in Oaxaca, Mexico, the intervention within our treatment group consisted of women from half the community banks in a group of 732 indigenous women in Oaxaca—all with access to microfinance loans—watching a 30-minute documentary featuring four of the most successful borrowers in the organization.   In their weekly community bank meetings, women in the randomly selected community banks were also recipients of a “hope curriculum” in which they were encouraged to develop aspirations for their personal lives and businesses, better understand and appreciate their gifts and abilities, and learn to conceptualize tangible pathways out of poverty.

Since all of the women in the experiment already had access to microloans, the working assumption underlying the Oaxaca Hope Project has always been that access to credit may not constitute the true constraint to upward economic mobility. One of the reasons that we chose this particular region for our experiment, was that at least anecdotally, it appeared to be that “internal constraints,” such as lack of aspirations and low self-esteem, were likely to be more binding than “external constraints” such as a lack of working capital.  By encouraging women in our treatment group to develop aspirations and confidence in their ability to expand their enterprises, we would be addressing the fundamental issue holding them back.  So our hypothesis was that an intervention addressing internal constraints should be effective if external constraints have already been released and internal constraints were those that were relevant.

There are a number of interesting results that we find from our 12-month follow-up data.  The first ones are not grounded in the results of the experiment per se, but simply from the fact that we were able to obtain specific data over time on different components of hope (aspirations, agency, conceptualization of pathways, future orientation, optimism, willingness to take risks, and happiness) as well as data on economic behavior and business outcomes.

The first thing we find is that end-line levels of hope at follow-up are strongly correlated with end-line sales and profits.  In other words, when things are going great, people tend to be a lot more hopeful.  This isn’t surprising at all, because this psychological relationship is probably familiar to most of us.  We also find that baseline levels of hope that we obtained around July 2015 were a strong predictor of business performance for the following year.  Again, this is not too surprising, but it also makes one wonder if hope needs to be rooted in a tangible expectation of a future reality.  The women were obviously more hopeful when the expected things to go well in the future, and these expectations apparently had at least some basis in reality–things did go pretty well for those who were more hopeful at baseline.

A preliminary analysis of our experimental results, which can’t change much because they were specified in a pre-analysis plan, indicate that our experiment was pretty good at elevating levels of hope among the women in the hope-treated community banks.  These 12-month results can all be seen underneath the 1-month follow-up results in the beautiful color figure as measured by impacts in standard deviations.  As you can see from the beautiful color graphic of our regression results, every psychological metric that we measure increased in the 12-month follow-up survey, especially the women’s sense of agency, conceptualization of pathways, future orientation, and optimism.  Both our Hope-3 Index (Aspirations, Agency, and Pathways) and our Hope-7 Index (all hope-related attributes) increased significantly at the 12-month follow-up.  So our experiment, the documentary, the goal-setting, the hope curriculum, did indeed make the women more hopeful.

But how did this affect their microenterprises? We find some evidence from institutional data that savings within the community bank appear to have increased about 20% among the women in the hope-filled community banks relative to those in the control group.  We also find an increase in our business-performance index that is statistically significant at the 95% confidence level.  But within this index the effects differ greatly between the measures of business performance.  Specifically, we find only small increases in sales and profits that are not close to statistical significance, and we find no changes in the number of hours per week women in the treated group devote to their enterprises.  However, we do find mildly significant increases in employment created within enterprises, and especially plans to hire employees in the future.

The employment findings are interesting, because they seem to reflect what other people have found in studies of microfinance impacts themselves. The number of employees, aside from the female entrepreneur herself, was zero in the vast majority of these enterprises at baseline; these were women typically working alone selling food on the street and making clothes for other women and children.  In the control group, not a single employee was added by any of the 330 women who did not receive our hope treatment.  But among the 402 women who were exposed to the hope treatment, 8 employees were added in these little businesses, generally not more than one for each job-creating business.  Now eight new employees does not seem like a very big number, but it is a lot bigger than zero, relatively speaking.  What appears to be true is that a handful of women were strongly influenced by what they experienced and took proactive measures to expand their enterprises.  But this does appear to be a small group at the upper tail– the vast majority of women in the treatment group didn’t create any new jobs at all.

So we would like to report that as an intervention, hope is overwhelmingly effective.  But instead what we can honestly report at this initial stage of analysis is that 1) hope seems to be based partly on things going well in the present and on a realistic expectation of things going well in the future;  2) it appears possible to increase hope; and 3) experimentally induced increases in hope may have modest, measured, and quite heterogeneous, impacts on the performance of small enterprises.  These initial and somewhat ambiguous results leave us with just enough to whet our appetite to try a similar experiment somewhere else, and to encourage others to do so as well.

Follow AcrossTwoWords on Twitter @BruceWydick.

• ← Using Children’s Drawings in Program and Policy Evaluation
• On Fake News, Narrative, Science, and Truth →