2016 Boulder Summer Conference on Consumer Financial Decision Making
IMPORTANT Disclaimer: These are notes from an attendee and do not represent the views of the researchers. To fully understand each study, please read the papers. Reference documents with abstracts
Conference Hosts: John Lynch & Donnie Lichtenstein, Conference Co-Chairs, and our fellow members of the Program Committee: Tony Cookson, Shaun Davies, Bart de Langhe, Phil Fernbach, Diego Garcia, Nick Reinholtz, Yanwen Wang, Brian Waters
Notes and Summary thoughts/questions by: Kristen Berman, CommonCents
Consumption and the Credit Channel Over the Business Cycle: Evidence from Bankruptcy Flags
Tal Gross (Columbia, Public Health) Matthew Notowidigdo (Northwestern, Economics) Jialan Wang (CFPB)
*Discussant: Mark Cole, Hope Loan Port
- Bankruptcy flag removal increases debt by $1500. Policy question: Is flag put on for too long?
- There are ‘Long strugglers” who don’t file bankruptcy as quickly – they try many more coping mechanisms (increasing limit) but they end up worse when they finally file (with more debt)
- Top fear of filing is ‘ever buy a house again?’
- Implication: How can we get people to file bankruptcy quicker?
- Who could be the predictive actors that could push people to into filing when they should be doing it? (e.g., credit card companies to flag people when they are on track to be bankrupt in future – Rachel Schneider comment)
- Seems like fears are legit: Can people actually buy a house after bankruptcy? How long after? Could there be a new mortgage type name for people who filed, to imply that it’s possible but difficult.
Frugality is Hard to Afford
Ye im Orhun (University of Michigan, Marketing) Mike Palazzolo (University of Michigan, Marketing)
*Discussant: Janneke Ratcliffe, CFPB
- Low income do not accelerate consumption (toilet paper) due to sales as fast as high income households
- Why? The story could go…inability to buy in bulk–> means that have less toilet on hand–> means less ability to pay attention to sales
- What happens when lower income feel richer? Low income increase package size in first week of the month, getting closer to consumption patterns of high income.
- Impact over the month: 392 sheet deficit from high to low income over the month. Low income buying less sheets. This is driven by the 2nd week of the month and onwards.
- LMI pay 5.9% more toward toliet paper than high income. (Dollar amount?)
- They can say that liquidity = 30% of the forfeited long term saving strategy.
- How can we help lower income lock in sales for bulk items, throughout the month? (group buying, programs to refund money if it goes on sale later, better planning for 1st of month purchases)
- General insight is that when you’re living on very little, you have to make myopic vs planned decisions, which have other costs (like missing sales). What are the most damaging myopic decisions that people can make in the grocery store and can we design product for these? (e..g, toilet paper at 5.9% more has to be less than $5 a month – what are more $$ impactful products to design for)
- Instead of pushing them to new stores, help them navigate stores they already have access to (the researcher comment…)
Growing up Without Finance
James Brown (Iowa State, Finance) Tony Cookson (University of Colorado – Boulder, Finance) Rawley Heimer (Federal Reserve Bank of Cleveland)
*Discussant: Janneke Ratcliffe (CFPB)
- In areas without banks (specifically native american reservations), there was randomized access to financial developments
- In areas with low access to financial development there was an impact on credit. 10% less people had access credit reports, It took 10% longer to get into credit market and it takes them a long time to catch up.
- Look at how the community compounds the effects of decreased access to banks – may not be driven by just visceral existence of a bank (Wendy comment)
- Why are we focused on bank retail locations in 2016? Hello Mobile. What effect will mobile have on lower income adoption of financial services, given most retail locations will disappear?
Rumination and Decision Making Among the Poor
Gita Johar (Columbia University, Marketing) Rachel Meng (Columbia University, Marketing) Keith Wilcox (Columbia University, Marketing)
*Discussant: Gal Zauberman (Yale University, Marketing)
- You start thinking about gas when the gas gauge is low…not when it’s high. This causes rumination.
- The poor chronically ruminate on finances. (poor as defined by 40k)
- The poor more myopic when choosing rewards, but when taking focus away from money (e.g., given social support message) the differences went away
- High ruminators more impulsive and perform worse on CRT, but not always bad: high ruminators say they are more likely to file EITC
- Most fin tech apps are trying to get people to pay more attention to their money by sending their balance or looking at transactions…but if rumination on our finances makes us more short term for the lower income, this is not a good thing
- Are there upsides to rumination that could be captured? What does positively rumination look like?
The Liquid Hand-to-Mouth: Evidence from Personal Finance Management Software
Michaela Pagel (Columbia University, Finance) Arna Vardardottir (Copenhagen Business School, Economics)
*Discussant: Gal Zauberman
- When do people get a payday loan? In theory, this should be when they are rock bottom – no cash left. However, liquidity and cash holdings are at least 3x times larger than predicted.
- Only 10% of people have less than 10 days left when they borrow. Less than 3% have less than one day spending left
- We are not rational about when we need debt and most do it before we actually need it, why? We have a different view of liquidity from an economist, that is above 0.
- Not sure this is a bad thing – this one side / negative effect of mental accounting. However to attempt to override mental accounting and get people to use the very last dollar seems depressing and very difficult.
Tough Times Borrowing: Effects of Fringe Lending Regulation on Credit Standing, Search and Access
Roman Galperin (John Hopkins University, Management) Kaili Mauricio (Federal Reserve Bank of Boston)
*Discussant: Alex Horowitz, Pew Charitable Trusts
- Removed payday access on military base to military people
- Increased Credit card access by 17%-25%
- No long term change in credit standing
- And they go through a more intense search for new credit, after access to fringe loans goes down.
- This implies removing payday lending pushes people to better credit – which is good. This finding *may* imply that people take payday loans when they could find better sources of credit and thus eliminating payday loans doesn’t have a negative impact on families.
- However it was unclear from this research what the actual net impact to the families was without payday loan access. Did they not eat? Did kids suffer? This is the age old payday question: Is allowing small dollar short term loan access permissible, even despite it’s interest rate – if they have no other way to get money and need it?
Does Sustained use of Overdraft and Deposit Advance Lead to Spiraling Fees?
Sergei Koulayev (CFPB) Charles Romero (CFPB)
*Discussant: Alex Horowitz
- Researchers favorable on Deposit Advances – as the product is designed to prevent spiraling fees with limits in usage.
- The researchers don’t see inherent spiraling fees for overdraft and more individual differences (Can’t find actual paper so unclear more specific findings)
- Overdraft: $32 average fee, median duration about 2 weeks
- Unclear consumer takeaway. From a policy POV, it was: DAP is not bad, because of it’s limits. Don’t regulate more.
- Impact question: Overdraft costs people a lot of money over their lifetime as a bank customer. Because of this life time cost, does it really matter if overdraft is “spiraling” or not?
The Influence of Goal-Setting on Credit Card Payment Decisions
Daniel Bartels Marketing) Abigail Sussman (Chicago, Marketing)
- Researchers tried to call out the difference between language used in academia around anchors and reference points by looking at how people deal with min payments on credit card and other suggested amounts
- In studies, they saw that people felt bad when missing a reference point and would take extra effort to meet or exceed the reference point vs be under it.
- Conclusion: People use suggested amount to infer target (which can increase payment).
- Unclear if the suggested target on CC bill or min amount on CC bill was determined to be reference point or anchor
- It does not seem to be novel finding that providing a repayment reference point (or anchor) on a credit card statement will push people towards this re-payment amount, given it is a reasonable amount.
- Implication of this general insight: Can we redesign CC statements and online portals to include not only Min payment but a goal payment that ensures you pay it off in X years. Ideally this payment acts like a ‘reference point’.
The Effects of Interest Rate Changes and Add-On Fee Regulation on Consumer Behavior in the U.S. Credit Card Market
Alexi Alexandrov – CFPB Ozlem Bedre-Defolie – ESMT (Economics) Daniel Grodzicki – Penn State University (Economics)
- They looked at impact of changing APR and late fees on CC usage and balances.
- Late payment fee changes matter more for prime consumers.
- Lowering late fees increases demand for purchases and balances.
- Changes made with the APR matters for usage and balances, more for subprime people
- Hypothetical increase in 100bps rise in rates, reduce CC balances by 9%, without change in usage.
- It is curious that increasing CC late payment fees matters more for prime customers vs subprime customer. Possible explanation: Like crime, increasing the punishment doesn’t actually change the amount of crime committed.
- Implication is that we cannot improve credit scores by merely increase late fees to incite on time payments from subprime borrowers – however on the flip side – we respond to decreased fees. By decreasing late fees people may spend more.
Toward a Framework of Understanding Financial Fragility: Conceptualizing and Documenting the Relationships Between Psychological Factors and Emergency Reserves
Stephen Spiller – UCLA (Marketing) Daniel Bartels – UCLA (Marketing) Jonathan Westfall – Delta State University (Psychology)
- Impulsivity / patience correlated with financial well being (valuing future and plan implementation)
- Work vs play game, once every day over 3 weeks
- Do you want to do a captcha (work) vs. video (play)
- Payday = 7 tokens
- Every day there is a chance they may encounter a large shock or small shock.
- The conditions? Vary information about the shocks.
- Following a big shock, they spend less.
- Knowledge about a shock is not enough. People in foresight condition, don’t reduce consumption prior to shock, even though they know.
- They do take longer to decide prior to the shock but it doesn’t translate to reduced consumption. (e.g., rumination doesn’t equal changed behavior)
- People who are more financially resilient are ruminating more about decisions … less financially stable are spending less time ruminating, but this does not correlate with consumption.
- Rumination does not equal behavior change Interpretation: Just because financially savvy people stay on top of their finances does not mean that telling less financially savvy people their balance will incite change.
- Should practitioners try to target people who just had a financial shock and use this as a trigger to change behavior?
- Just like we struggle to financially plan for car repairs (these expected but surprise shocks), participants didn’t plan for expected but surprise shocks in the game. Just telling people there will be an upcoming shock is not enough.
The ABCs of Financial Education: Experimental Evidence on Attitudes, Behavior, and Cognitive Biases
Bilal Zia – World Bank Shawn Cole – Harvard University (Finance) Fenella Carpena – UC Berkeley (Economics) Jeremy Shapiro – Princeton University (Economics)
- What are key barriers that prevent education from working? Where does it break down?
- Internal (self control, lack of attention, change habits)
- External (lack specifics, skilled educators)
- Research: Try to eliminate the barriers and see how financial education performs.
Many treatments – combining these in different ways:
- Local high profile actors to do financial education videos (saving, budgeting, credit). Delivered in classroom w/ moderator
Videos+ Pay for performance after for everyone. Why did they choose this?
- Low interest and poor attendance (Bruhn, 2013): People not showing up, need to solve for this
- Angrist and Lavy, in israli = it has worked
- Non trival: Wage was a day’s wage
- Come 3 weeks and test them
Videos + Goal setting:
- Goerge, Kube, 2012, Harding Hsiaw, 2014, Health management (shifts 2004)
- Give them the calendar with target date w/ a sticker of achieving
Videos + Counseling:
- Willis, 2011 education with groups not effective
- 1:1 shown to work in personal health (Lerman, 1995)
- Insurance sign up (Dalal Morduch, 2010)
- 1:1 helped on variety of issues (budget, insurance, etc)
- Trained by mircofinance team, so counselors were fairly good
The results: Counseling only sustained impact.
- How did they measure?
- Measure financial literacy – current literacy dependent on numeracy so they added financial awareness and attitude.
- Financial education has effect on awareness and attitude. 10% higher
- Pay for performance does not work at all.
- Re: Budgeting:
- People believe budgeting is helpful for all treatments
- however when asking “have you tried making a budget?” Sig higher in counseling, people say Yes in other treatments.
- Sustained impact is only from counseling.
- Re: Savings:
- Financial education does nothing for informal or formal savings.
- Goal savings has effect on formal/informal savings but small.
- Counseling has no effect on informal savings and large effect on formal savings.
- Financial education improves knowledge…but it is again proven not to have sustained effects by itself.
- Goal setting includes simple follow up action
- Individual counseling allows for sustained action.
- How to scale 1:1 counseling for the masses?
- Why is gov still funding financial education?
Intergenerational Linkages in Household Credit
Andra Ghent (Wisconsin Madison, Economics) Marianna Kudlyak (Federal Reserve Bank of Richmond)
- Children credit outcomes correlated with parents,even controlling for location
- Transitory shocks to parental credit conditions do not differentially affect children.
- Intergenerational credit linkages are stronger in cities with lower mobility
- More inputs into education do not weaken links
- What are potential solutions to get kids out of a credit trap? (e.g., exposure to high credit people in same mobility bracket…early education type interventions?)
- Points to more root causes in inequality – can’t train people on improving credit score when conditions/environment for success are absent. That said, study was very depressing without offering learnings on who has made it out of the cycle
On a Need-to-Know Basis: Divergent Trajectories of Financial Expertise in Couples and Effects on Independent Search and Decision Making.
Adrian Ward (University of Texas-Austin, Marketing) John Lynch (University of Colorado-Boulder, Marketing)
- Who becomes CFO? New couples are responsive to weak signals of advantage to figure this out. (e.g., it appears random)
- CFO has no more financial literacy than other, it may seem it looks arbitrary.
- But when one is assigned, this person gets better and the other person gets worse. Good for present and delegating responsibilities, terrible if someone gets divorced, widowed, etc.
- Fin tech companies have two targets – people who take care of the finances and people who let someone else do it.
- Large opportunity to get the recently divorced/ widowed people into some counseling.
- Open question: What are the key mistakes that people without financial responsibility will make? Or is it just that they will procrastinate financial action because they feel less confident?
Checklists as Selective Choice Architecture
Melissa Knoll – CFPB Kirstin Appelt – UBC (Marketing) Eric Johnson – Columbia University (Marketing); CFPB Jonathan Westfall – Delta State University (Psychology)
- Checklist order affects choice for when people choose to take their social security
- “Pro later” list before “pro early” list delays claiming age by 13 months. (i.e., checklists presenting reasons to claim benefits later before reasons to claim early)
- Reason order checklist matters: query theory, 1st things you think about are more accessible. (processed 1st items at 65 sec vs 77 sec)
- And…Checklist is more impactful than default…by a large 10 mo
- Is this good? Yes: Everyone (94%) is claiming too early… checklist minimizes this error
- Total Impact: Delaying claiming for 13 months = 122 per month or 53k in expected lifetime benefit.
- Choice architecture consideration: We may overlook this preference checklist as a viable option when designing interfaces. Could this also be used to set emergency savings allocation? Or choose min payment amount for CC?
- Application: Given claiming SS benefits is controlled by gov (and these forms are hard to change apparently) where else can the choice be made, such that a decision has already happened at the point of gov forms? (Tbd?)
- Wendy comment: Are you looking at race? Seems effect will drastically change when looking at margin of error for white female vs black males
Who is Easier to Nudge?
John Beshears (Harvard, Management) James Choi (Yale, Finance) David Laibson (Harvard, Economics) Brigitte Madrian (Havard, Public Policy) Sean Yixiang Wang (NBER)
- They looked at default persisting in 401k allocations: Why does the default persist for some and not others? Looked at below median-income and above-median-income employees at ten large companies
Differences could be alignment w/ preference or speed/cost of taking action. Implication for policy: Should we increase default contribution rates? If alignment then increasing may not work because people would opt out, if it’s speed, then increasing default could drive people up.
- Low income more likely to be stuck at default, where there is a default. They also move when there is a default more often.
- Researcher are attributing speed vs alignment as the main cause for sticking with a default.
- General question: If the team has data from all contributions of 10 companies, is this the only question that was asked?
- Consumer’s high likelihood to keep the default, given there is a default, has been shown in other experiments. Given this, it’s unclear that we need to dive into why people stay on the default. In addition, the idea of ‘alignment’ to the target is chicken/egg. If there is a default, and you pick it, you’d be more likely to align this as your target. (preferences beget preferences)
- They did not look at matching rates by the company (which seems to have more of an effect than default rate)
- More interesting question for the dataset: Would have loved to know the ideal default/ matching rate for firms to set that would increase opt in for the longest time.
Struggling to Bankruptcy: Struggling to Bankruptcy Robert Lawless (University of Illinois, Law) Katherine Porter (UC Irvine, Law) Deborah Thorne (Idaho, Sociology)
Scam the Declined or Decline to be Scammed: A Model of Elder Fraud Simon Gervais (Duke, Finance) Terrance Odean (UC Berkeley, Finance)
Out of Sight, Out of Mind: Consumer Reactions to News on Data Breaches and Identity Theft Vyacheslav Mikhed (Federal Reserve Bank of Philadelphia) Michael Vogan (Moody’s Analytics)
Attention Variation and Welfare: Theory and Evidence from a Tax Salience Experiment Dmitry Taubinsky (Univ. of California-Berkeley, Economics) Alex Rees-Jones (Wharton, Operations Research)
Does Salient Financial Information Affect Academic Performance and Borrowing Behavior Among College Students? Maximilian Schmeiser (Federal Reserve Board) Christiana Stoddard (Montana State University, Economics) Carly Urban (Montana State University, Economics)