2018 Annual Report

common cents lab

Download our full Common Cents Lab 2018 Annual Report Download Report

About Us

Common Cents Lab

Common Cents, supported by MetLife Foundation, is a financial research lab at Duke University that creates and tests interventions to help low-to-moderate income households increase their financial well-being. Common Cents leverages research gleaned from behavioral economics to create interventions that lead to positive financial behaviors.The Common Cents Lab is part of the Center for Advanced Hindsight at Duke University. Common Cents is comprised of researchers and experts in product design, economics, psychology, public policy, advertising, business administration, and more. The lab is led by Behavioral Economics Professor Dan Ariely. Ariely has written three New York Times bestsellers, including Predictably Irrational.

To fulfill its mission, Common Cents partners with organizations, including fin-tech companies, credit unions, banks and nonprofits, that believe their work could be improved through insights gained from behavioral economics. To learn more about Common Cents Lab visit advanced-hindsight.com/commoncents-lab/.

MetLife Foundation

At MetLife Foundation, we believe financial health belongs to everyone. We bring together bold solutions, deep financial expertise and meaningful grants to build financial health for people and communities that are underserved and aspire for more. We partner with organizations around the world to create financial health solutions and build stronger communities, engaging MetLife employees to help drive impact. To date, our financial health work has reached more than 6 million low-income individuals in 42 countries.To learn more about MetLife Foundation, visit www.metlife.org

met-life logo icon-cah logo

Dear Reader,

Now that Common Cents Lab has concluded its third year, we are officially past the startup (or lift-off) phase and are, so to speak, ready to add fuel.

This doesn’t mean that the journey has been without turbulence. Designing products and processes to transform the financial lives of people with low incomes isn’t simple, even with all we’ve learned to date. What it does mean is that the Common Cents Lab team has become ever more adept at responding to and learning from our changing environment – the economy, politics, research insights, and the increasing availability of data – in our shared quest to build financial health for all. Recognizing that change is the only constant, the team remains focused on ensuring that the research they rely on and experiments they conduct have potential to translate into a positive impact on the people we care about. Along with the successes, the team is also dedicated to sharing what’s not working – or, what’s working, but not having a significant impact. Sure, it means that the work ahead remains difficult – but, you weren’t really expecting a silver bullet (right?).

As I am a couple of weeks late in writing this opening letter, I also am relieved to discover that CCL and its partners also suffer from optimism bias – the justified assumption that all important work has a way of getting done in the hours optimistically allotted. Tardiness does not, however, minimize our real optimism about and commitment to our collective ability to serve low-income folks with the right products and services.

MetLife Foundation’s optimism is expressed through our continued commitment to this work in the US, and also through our support to CCL in expanding its approach to partners outside the US. We look forward to bringing the practical and rigorous approach of behavioral science and experimentation that CCL champions to new markets and contexts. MetLife Foundation continues to support financial health projects and programs in more than 40 countries. We’ve reached our goal of committing $200 million to the cause of financial inclusion, and we’re continuing our work to ensure that financial health belongs to everyone.

Our energy and optimism around financial health through product and service design and delivery continues to grow as we work with partners like CCL and the universe of nonprofits, tech firms, credit unions, academics, and others who have joined together in identifying new solutions to strengthen families and communities. We trust that this (very readable) annual update from CCL will inspire you and your organizations to join in our mission, too.

Evelyn Stark MetLife Foundation

About this Report

The following report summarizes all of our 2018 projects in case studies, each of which falls into one of the following categories:

We encourage you to read through each case study and ask yourself how you can apply the insights to your organization. We hope the work we document here provides inspiration and guidance both for organizations already using behavioral economics as well as those just beginning to think about how to better support the communities they serve.

We are proud of, and excited by, the impact we have had with our partners. Even still, we recognize how much more there is to learn about how to effectively apply behavioral science to help improve the lives of those we serve. In this report, we document our work over last year and show how we have worked with our partners to apply behavioral insights in their work. In some cases, these efforts successfully changed how their clients behaved. In some cases, they did not. Across all of them, we learned more about how we behave as human beings and how we, as organizations, can translate those learnings into practice.

Looking forward to 2019, we are excited to expand our impact even further. By developing products and services that reflect the real experiences of people and by leveraging the ways that people actually make decisions, we will be able to continue to foster financially stable and healthy communities with our partners.

Executive Summary

A healthy economy is not enough

Just over the past year alone, GDP growth reached 4.2%, the U.S. labor market is widely considered to be at full-employment, and inflation is just above 2%. By all metrics, 2018 was representative of the past few years when we have experienced a healthy, even booming economy. And yet, over half of all households in the U.S. regularly worry about their finances.

This feeling is not unfounded – by and large, U.S. households are not financially stable:

They struggle to keep up - This year, 6 in 10 households will experience at least one financial shock, averaging $2,000, and yet almost half are struggling to make ends meet from one paycheck to the next.

They struggle to keep up - Households are relying on high-cost debt to cover necessary and discretionary expenses: 48% of households have revolving credit card debt and 1 in 3 households have debt that is in collections.

Over and over again, we see that a strong economy is not enough. A good job is not enough. Financial education is not enough.

Behavioral interventions can help

At the Common Cents Lab, we partner with mission-aligned organizations across the country to create and test interventions that use behavioral science to overcome the challenges their clients face.




Unique projects and experiments

We currently have 17 projects that have either launched and are still collecting data, or plan launch early next year.

Our partners are a diverse group that includes credit unions, banks, governmental agencies, non-profit organizations, and technology companies. Some of our partners serve hundreds of people while others serve hundreds of thousands or even more. Although diverse, at the heart of our partnerships is a common desire to meaningfully improve the financial lives of people and communities across the country, with a special emphasis on low- to moderate-income (LMI) households.

In our partnerships, we have learned together how and when to apply findings from behavioral research to products and services in order to achieve this goal



5.7 million


Our Approach

Across all the various partnerships and experiments that we have launched, we have learned a great deal about how human beings behave but also about how organizations can translate those learnings into practice. Through our experience over the years, we’ve learned that our environment and surrounding context significantly shapes the decisions we make.

Designing that environment is a little like trying to build a rocket to send into space

– if you are going to be successful, you have to calculate exactly where you are going (by getting very specific with your target behavior and desired outcome), find ways to reduce friction points on the path to the desired behavior, and add more fuel and motivation to spur them on.

Please read more about our approach in the following sections. In each, we outline several strategies that we have seen to be successful in either reducing friction or to adding fuel to motivate behavior and improve decision-making.

1Reducing friction
to follow through on good decisions

Friction costs are real. Each additional click, field, step, choice, form, or call is a toll on the decision-maker and can deter even the best of intentions. For example, research suggests that reducing a form from 9 fields to just 5 fields can increase conversion rates by 3.4 percentage points. Simply removing or reducing the number of steps required of people can increase how many of those people end up following through.

We are currently testing both light-touch and heavier re-design strategies for friction reduction in several 2018 projects. Beyond simply cutting steps out of the process, there are several strategies that we are using to reduce friction.

Strategy 1: Make the right behavior automatic

Evidence suggests that changing intentions is unlikely to lead to big changes in behavior. Even with strong intentions, people have lapses in self-control or they might procrastinate. In reality, there are many barriers and biases that make it difficult for even the best among us to follow through on our intentions. When designing products and services, it is often better to recognize this and make the right behaviors automatic.

For long-term savings, automatic enrollment and default contributions have been the biggest factor in increasing savings rates, increasing participation from 34% to 90%. Setting up automatic transfers is often heralded as a best practice for building short-term savings as well. However, current automatic savings programs require the saver to pick a day each month, like the 25th, and a fixed amount to transfer, like $100. These are not designed well for most low-wage workers, who are more likely to be paid bi-weekly and income that varies paycheck to paycheck. We are building better automatic systems for LMI households. This means programs that align with when and how much people are paid and programs that de-risk signing up for an automatic savings transfer.

making saving automatic can help overcome the many barriers and biases that make it difficult, but many automated savings options are not well designed for low-wage workers.

With Narmi, a tech platform for Credit Unions, we are testing ways in which we can connect automating savings directly with when someone gets paid. Specifically, we are exploring whether a fixed percent-based transfer amount (i.e. 6% of every deposit) is more or less attractive than an escalating percent-based transfer amount (i.e. 2% to start, increasing by 1% until 6%). With Earn, a non-profit, financial technology company that incentivizes savings, we are testing whether offering overdraft protection helps make automatic savings feel less risky and more attractive to potential savers.

We are also exploring how automation at the employer level can help employees’ financial wellness. For example, in our work with Homebase, a scheduling platform for small- to mid-sized employers, we are encouraging managers to publish work schedules earlier. This gives employees more notice to better plan for their work schedules or pick up additional work outside of their primary job. To do this, we are creating default schedules that allow employers to easily and quickly copy schedules from prior weeks. We believe that automatically generating schedules will make it easier for employers to publish their schedules sooner, giving employers significantly more time to plan ahead.

The tests with Earn and Homebase are currently in the field. The test with Narmi is poised to launch in summer 2019. We look forward to sharing our results.

Strategy 2: Make the right choice the default

In addition to automation, it’s important to set the right default. Defaults are incredibly powerful, and yet, many financial institutions and companies set defaults that are not ideal for the consumer. For example, the most common default contribution rate for auto-enrollment retirement plans is 3%, and 40-70% of employees stick to the default rate. Yet, experts recommend that people save closer to 15% of their income to be adequately prepared for retirement. Defaults are powerful because they make that choice the easy choice, they imply that is the “right choice”, and they create an anchor around the default number.

Another area where carefully choosing the right defaults is important is debt: taking on debt and paying it back. For example, the default for taking out a car loan or a student loan is often the maximum available. In partnership with Duke’s Office of Personal Finance, the department that manages student loans for the University, we re-designed how graduate students were offered and accepted loans. Instead of providing the total amount available to borrow, we split the total loan package into two bundles – one covering known, fixed costs like tuition and the other covering unknown, variable expenses like the cost of living. By breaking up the loan into two separate numbers, we decreased the percent of students borrowing the maximum loan amount from 29% to 13%, a more than 50% reduction.

figuring out what is affordable to pay for a car is complicated and most people anchor on the monthly payment, neglecting uncertain or future expenses

With Beneficial State Bank, a community development bank based in California, we changed the default payment due date of car loans from 30 days from the day they buy the car to the day they get paid. By defaulting the due date to their payday, we are reducing the mismatch between income and expenses, which may improve rates of repayment. This project is currently in the field and results will be shared in our 2019 Annual Report.

Strategy 3: Simplify complexity

Complexity creates its own kind of friction. A study from 2015 found that complexity around the Earned Income Tax Credit explains why not everyone takes advantage of the tax benefit. As a decision becomes more complex, we should take more time and effort figuring out the right decision for ourselves – but instead, we rely on mental shortcuts that may be error-prone and imprecise

One complex decision is figuring out how much car you can buy. With 7 million Americans more than 90 days behind in their car payments, it appears that many people are mis-calculating how much they can afford. Most car buyers only consider the monthly payment of their auto loan, neglecting to consider uncertain or future expenses that are difficult to determine. These expenses include repairs, maintenance, insurance, or how much they will spend on gas. As a result, many people end up paying more than they can comfortably afford. We designed an auto loan calculator that breaks down these complex costs and conveniently adds them to the monthly payment, therefore adjusting the total amount one should borrow. In a lab experiment, we found that going through the auto loan calculator significantly reduced the size of an auto loan participants felt they could afford, moving from $15,000 to $12,500 on average, a 20% reduction in the size of a potential loan. We plan on testing the calculator in the field to see if it actually decreases the amount people borrow.

making saving automatic can help overcome the many barriers and biases that make it difficult, but many automated savings options are not well designed for low-wage workers.

With the St. Louis Housing Authority, the government entity that manages St. Louis’ section-8 housing voucher program, we are testing a text-message program that breaks down the complex steps of moving to better neighborhoods with a voucher. Voucher recipients receive weekly text messages that provide clear action steps for that week to get them closer to moving. We think that simplifying the complexity of moving with a voucher will encourage more voucher recipients to move to higher opportunity areas, which has been shown to improve life outcomes. This project is currently in the field; we will share results in our 2019 Annual Report.

Making a process completely frictionless is often impossible due to technological, regulatory, financial or logistical constraints. This means that we also have to find ways to reward and motivate people to overcome the remaining barriers. Once friction has been reduced as much as possible, we look to boost motivation either by making the existing benefits more appealing or by adding benefits, often closer in time.

2Adding fuel
to increase motivation

Strategy 1: Emphasize benefits that are immediate and tangible.

We are naturally more motivated by wants and desires in the present moment than by our possible needs in the future. This makes good financial decision-making very difficult, where we are regularly asking people to give up something concrete and appealing right now for a long-term, intangible pay-off.

In some cases, simply reframing the way benefits are described can change how people perceive them. In a project with NetSpend, a pre-paid card provider, we sent an email to encourage users to use the savings feature. However, instead of talking about it as an opportunity to save, which has connotations of future benefits, we reframed saving as an opportunity to “earn” money, which people think of as a more immediate benefit. Consistent with previous research, reframing saving as “earning” increased response to the email by 4 percentage points as compared to the usual response rate.

creating a goal-setting exercise that coaching participants connect with their long-term aspiration increased the program retention rate from 9% to 13%

With LISC’s Financial Opportunity Centers, a network of non-profits committed to improving financial wellbeing through financial coaching, we designed postcards to facilitate a visual goal-setting exercise in a financial coaching session. This activity helped frame the benefits of coaching as more tangible and immediately rewarding. Coaching participants who participated in the goal-setting exercise were more likely to attend subsequent coaching sessions, leading to an increase in the retention rate from 9% to 13%.

Strategy 2: Add new benefits that are immediate and tangible

Sometimes the benefits a product or services provides are neither immediate nor tangible. In these cases, you may need to create new benefits within the process. These benefits do not need to be financial incentives – they can include emotional, reputational, social, or other psychological benefits.

We are currently testing the effect of giving parents a sense of progress toward their savings goals with San Francisco’s Kindergarten to College program, a college savings program run by the city of San Francisco. Together, we are currently testing a tool that helps people visualize and monitor their progress toward long-term savings goals. We have designed a simple card that helps people visualize and monitor progress. We will share the results in our 2019 Annual Report.

not all incentives need to be (or should be monetary) - adding in emotional, reputational, social, or other psyhological benefits can be equally effective.

With OregonSaves, a state-run retirement savings program for all employees in Oregon, we are testing adding a social benefit. We designed a simple decision-aid that communicates the benefits of the program through profiles of Oregonians. We are testing whether we can increase retirement savings rates by providing an implicit recommendation using social proof (i.e. most people like you have decided this is the right thing for them).

With the City of St. Louis’ College Kids program, we are redesigning the existing incentive structure for their college savings accounts. Currently, parents receive matched savings that get deposited into their child’s account at the end of the year and will not be touched or used for 12 years. We are testing a new model where parents can receive part of the match immediately if they sign up for automatic savings transfers. We look forward to reporting out on those learnings next year.

Strategy 3: Provide simple, relevant information at the right time

Life is busy and it is easy to procrastinate, delay, and forget. A simple reminder can be effective at getting people to open an account, make a payment, or complete a task. What we’ve learned, though, is that when you remind someone can be as important as what you are asking them to do.

With Homebase, a scheduling platform, we reminded hourly workers to enter in their desired work hours in the moment that they were in the platform checking their schedule. These simple and timely reminders increased the number of entries from 1.4% of employees to between 2.2%-4.6% of employees, depending on the message text. We found this was especially effective when we framed the reminder as an ask to employees to provide “missing information”.

Similarly, we increased short-term savings for employees at IH Mississippi Valley Credit Union by sending weekly savings reminders for a month as part of a credit union-wide “Autumn Savings Fest.” Employees are paid weekly, so we tested sending the reminders on their payday versus on Monday, when they were past the weekend and the next payday was in sight. We found that Monday reminders were more effective – savers saved an average of $120 more in the month than when they were reminded on payday.

severed saved an average of 120$ more in the month when they were regulary reminded a few days before their weekly paycheck

Unfortunately, sometimes a reminder isn’t enough. Reminders work best when the task is easy and people may simply forget. Similarly, reminders that are not sent through a proper channel or not timed correctly end up just providing information, which in and of itself does not significantly change behavior.

With the Pennsylvania Treasury and the City of St. Louis’s College Kids, we tried sending the message from a more familiar source and address message to the child rather than the parent to see if either made people notice the message more. Neither were successful. Similarly, with United Way of Tucson and Southern Arizona, we tried door-hangers using behaviorally-informed messages to remind and encourage people to use VITA sites rather than paid preparers for tax preparation. Ultimately none of the conditions had a significant impact.

We are continuing to try and better understand when and how to effectively to deliver timely information. With UpTrust, we are helping people avoid additional fines by providing reminders and timely information to encourage attendance in court-ordered classes. Relatedly, with the Vera Institute for Justice, we are designing a calculator that quickly calculates a defendant’s ability to pay before their arraignment hearing. This information can be presented to judges as they are making decisions about bail – a decision that often has significant financial implications for the defendants and their families.

3Other lessons

In addition to what we’ve learned about the various behavioral nudges we have been testing across the years, we also have general learnings about working in this field.Strategy 1: Emphasize benefits that are immediate and tangible

Light-touch and superficial changes often mean light (or no) effect.

Over the last three years, we’ve completed 56 experiments and about half have had a statistically significant and positive effect. When we look at what differentiates what works from what does not, our biggest wins have also been those requiring the biggest changes. While small tweaks and different frames in emails have sometimes generated small lifts, most of these projects have had negligent or null effects. This suggests there is a real trade-off between what is easy and what is meaningful. We are taking this learning to heart and our mantra for 2019 will be “Be Bold.” We will focus on more in-product flow and feature tests and fewer simple email tests.

our mantra from 2019 will be "be Bold" as we focus on more in-product flow and feature tests

Some savings decisions are harder to move than others.

We have worked on emergency savings, short-term goal savings, spending reductions, and retirement savings with a relatively high success rate. However, we have been unable to make meaningful movement on saving for college. This year alone, we have run six experiments, ranging from simple email tests to encourage participation to quick-enroll forms for employees to create auto-transfers to their child’s account, without any of them creating a measurable lift. While we are not ready to give up (we have another three studies in the field or ready to launch), this does suggest that not all savings decisions are equal – that the same combination of reducing friction and adding fuel does not work as well for parents saving for college, which is likely a lower priority for struggling families. As we move into the next year, we will examine how various savings decisions interact with each other and explore ways to reduce the complexity around which account to save in, when, how much, and develop a deeper understanding of how people and parents prioritize savings decisions.

We are all subject to planning falacy.

We completed 19 studies this year. We have another 8 that are still in the field collecting data and another 9 that we are actively preparing to launch. Even after three years, we and our partners continue to be overly optimistic about timelines from ideation to build to launch to analysis. While this is a summary of our learnings this year, keep in mind that we still only have just over half of all studies reporting. We encourage you to keep in touch, subscribe to our newsletter, and stay tuned for more results.

Interested in learning more?

Download our full Common Cents Lab 2018 Annual Report Download Report