Mobilizing Savings through

Products and Persuasion

Nava Ashraf

Harvard Business School

Oct. 17, 2005

Based on evidence and experience from:

Ashraf, Karlan and Yin • “Tying Odysseus to the Mast:

Evidence from a Commitment Savings Product in the Philippines” (Quarterly Journal of Economics, May 2006 volume) and

• “Deposit Collectors” (2005 working paper)

Available at:

http://www.people.hbs.edu/nashraf/papers.html

Why Actions Don’t Accord with Intentions: Psychological Barriers

to Savings

• Self-Control Problems & “Time Inconsistency”– Commitment Devices

• Bounded Cognition (ie, numbers/info overload): – Mental Accounting

• Inability to follow through on plans: the role of memory and habits– Reminders and Collectors

Overcoming Barriers to Savings: Designing the SEED Commitment

Savings Product

• Withdrawal restriction– Either time-based (140) or amount-based (62)– Decided entirely by client

• Deposit incentive– Ganansiya box – 167 out of 202 tookup – Automatic transfer – 2 tookup

• Same interest rate as regular savings account– Hence, loss of liquidity is uncompensated

Experimental Design

• Sample frame: 1777 existing (or former) bank clients of a Philippine small, semi-rural, for-profit bank

• Participants randomized individually into:– Treatment (Offered SEED), 50%– Marketing(Encouraged to Save), 25%– Control (Nothing), 25%

• Marketing team from Bank visited one-on-one with T & M groups

• Marketing & Control groups not allowed to take-up product

Findings

• 28% of those offered signed up for the SEED product– Time Inconsistent individuals, particularly women, were much

more likely to take up

• Significant Savings increase for those offered SEED compared to the Control group:

• Average bank account savings increase for those assigned to treatment (ITT): after 6 months=46%; after 12 months=80% increase

• Scaling up estimate by those who actually opened the account: increase in average savings (TOT): after 6 months =192%; after 12 months= 337% increase

• Marketing Group shows some increases, but not statistically different from Control group

Deposit Collectors

• Offered Deposit Collection service for fee• Much smaller scale experiment: 5 treatment, 5

control villages• Find significant impact on increasing saving and

lowering borrowing• Takeup variation we’re able to explain suggests

that married women much more likely to take this up, but large part of take up is unexplained

• Anecdotal evidence suggests deposit collection helps with planning follow-through

Conclusion

• Products designed with psychology in mind increase savings

• Products need persuasion to sell them

• Persuasion alone may not be as effective as combining persuasion with innovative products -- ways to act on the momentum that comes from being persuaded.

Additional Tables

Measuring Impact

Figure 1: Changes in Overall Savings Balances (12 months)

-1000

-800

-600

-400

-200

0

200

400

600

800

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Deciles of Change in Savings Balances

Ch

ang

e in

Ph

ilip

pin

e P

eso

s

Treatment Group

Marketing Group

Control Group

Measuring Impact

Figure 2: Changes in Overall Savings Balances(12 months)

-1000

-500

0

500

1000

1500

2000

2500

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Decies of Change in Savings Balances

Ch

ang

e in

Ph

ilip

pin

e P

eso

s

Treatment: SEEDTakeup

Treatment:No SEEDTakeup

Marketing Group

Control Group

*significant at 10%, **significant at 5%, *** significant at 1%

PANEL A. INTENT TO TREAT EFFECT: OLSLength

Dependent Variable:Change in

Total Balance

Change in Total Balance

Change in Total

Balance

Change in Total Balance

Binary Outcome = 1 if Change in Balance > 0%

Binary Outcome = 1 if Change in Balance > 20%

Sample All

Commitment & Marketing

Only All

Commitment & Marketing

Only All All(1) (2) (3) (4) (5) (7)

Commitment Treatment 234.678* 49.828 411.466* 287.575 0.100*** 0.096***-101.748 (156.027) (244.021) (228.523) (0.025) (0.020)

Marketing Treatment 184.851 123.891 0.044 0.033(146.982) (153.440) (0.028) (0.021)

Constant 40.626 225.476* 65.183 189.074** 0.232*** 0.107***-61.676 (133.405) (124.215) (90.072) (0.020) (0.014)

Observations 1777 1308 1777 1308 1777 1777R-squared 0.00 0.00 0.00 0.00 0.01 0.01

Table 4: Impact on Change in Savings Held at BankOLS, IV

6 months 12 months 12 months

Length

Sample All

Commitment & Marketing

Groups Only All

Commitment & Marketing

Groups Only(1) (2) (3) (4)

10th Percentile Commitment Treatment 146.450* 118.040 317.490** -9.660(80.1) (128.5) (116.3) (54.2)

20th Percentile Commitment Treatment 0.000 0.000 20.000 0.000(6.5) (7.4) (30.3) (8.9)

30th Percentile Commitment Treatment 59.820*** 50.300*** 107.030*** 6.130(16.9) (19.5) (23.9) (21.0)

40th Percentile Commitment Treatment 60.000*** 56.330*** 42.5099** 12.900(14.5) (12.0) (18.4) (9.6)

50th Percentile Commitment Treatment 0.000 0.000 62.000*** 40.42*(7.1) (5.8) (18.3) (20.0)

60th Percentile Commitment Treatment 4.140*** 4.140*** 37.620** 15.030(0.7) (1.4) (8.9) (8.0)

70th Percentile Commitment Treatment 8.690*** 8.740*** 6.550*** 6.55*(1.7) (0.9) (3.7) (3.2)

80th Percentile Commitment Treatment 87.770*** 87.510*** 65.790*** 61.77**(16.2) (7.1) (16.6) (19.6)

90th Percentile Commitment Treatment 403.730*** 367.210*** 437.230** 172.170(113.200) (116.400) (144.3) (195.7)

Observation 1777 1308 1777 1308

Dependent Variable: Change in Total Savings Held at Bank

6 months 12 months

Table 5. Impact on Financial SavingsDecile Regressions

Dependent Variable: BorrowingProbit

Length: 10 months 15 months 10 months 15 months 15 months(5) (6) (7) (8) (10)

Treatment 227.340*** 197.964* 343.413*** 530.415** -0.034**(61.238) (94.092) (100.282) (192.420) (0.016)

SEED Treatment 346.160 702.233***(361.628) (175.185)

Female -268.246 68.702(265.534) (346.441)

Treatment * SEED Treatment -524.033 -1,037.369**(322.871) (442.314)

Treatment * Female 9.399 -629.212(380.653) (364.985)

Treatment * SEED Treatment * Female 204.750 1,088.405(535.365) (844.982)

Constant -476.948** -742.663** -513.974** -904.079***(200.704) (233.947) (198.386) (261.957)

Observations 640 640 640 640 640R-squared 0.10 0.06 0.10 0.06

Fixed effect for block in randomization Yes Yes Yes Yes YesControl for pre-intervention balance Yes Yes Yes Yes YesControl for individual characteristics Yes Yes Yes Yes Yes

Total savings balanceOLS

Table 4: Impact of Deposit Collection Service