INCREASING

PRODUCTIVITY USING
SIMPLE FRAMING
MANIPULATIONS
INTRODUCTION

•Power of framing: the manner in which a choice set is presented has been

found to affect individual decision making and action considerably.

•This study uses natural field experiment to explore how incentive schemes can

be used to increase labor productivity in a real marketplace

•This experiment tests if workers are susceptible to framing manipulation

 METHODOLOGY
•Behavioral Research using method of
framing manipulation
•Natural Field experiment
• Main insights derived from a
comparison of productivity
measures across a baseline and
two treatments
•Positively framed bonus (Reward)
treatment
•Negatively Framed bonus
(Punishment) treatment 
•Hypothesis: if loss aversion strong
among workers then losses loom
larger than gains, Punishment
treatment should outperform
productivity of Reward treatment 
METHODOLOGY 
 METHODOLOGY
•Experiment lasted roughly 4 months for groups and 6 months for
individuals, total of 165 workers participated in experiment. 
•Theory behind Methodology: Loss Aversion 
•Theory behind why there may be differences between gain
( reward) and loss (punishment) treatments is prospect theory
•Predicts that if workers are loss, averse, negative utility derived
fros a loss of X is greater in magnitude than any positive utility
derived from a gain of X.
•Loss averse worker will expend more effort, higher productivity
output rate on average under the Punishment treatment than the
Reward treatment.
MAIN RESULT 1
•The Punishment frame outperforms the
Reward frame in both the individual and
group treatments, with observed
differences slightly above 1%. Meaning that
total productivity increases by 1% when
moving from the Reward to the Punishment
treatment.
•The overall framing effect is stronger
among groups than among individuals due
to the differences for the group treatments
are statistically significant and robust to
various controls, whereas the individual
differences are not statistically significant at
conventional levels.
•Therefore, within this particular labor
market, the overall framing effect is
stronger among groups than among
individuals
 (N) DENOTES THE NUMBER OF WEEKS A GROUP OR INDIVIDUAL WAS SUBJECT TO THAT PARTICULAR INCENTIVE SCHEME.

For a within-team assessment, one needs to compare

Reward treatment had an average of 401 units
produced, whereas the Punishment treatment
had an average of 402 units produced per hour
numbers from each set diagonally. For example, for set G-
1, the Punishment treatment induced 30 more units of
production (430–400) per hour from one team, and for the
other team the Reward treatment outperformed the
Punishment treatment (429–402).
RESULTS
•observed treatment effects persist over the
entire sample period suggesting the power of
simple framing manipulations in enhancing
productivity. This implies that there are
economically significant levers that can be
pulled which can materially influence the
long-term growth of firms.
•conditionality of the reward and worker
reputation are substitutes where if either is
available, the worker is sufficiently
incentivized
PROSPECT THEORY
•The theory underlying why there might be
behavioral differences between the gain and
loss frame is due to prospect theory where it
states that agent’s net utility can be described
by a value function that is :
• (i) measured over deviations from a
reference point that is assessed over some
narrowly bracketed time- frame 
•(ii) convex for losses and concave for gains
•(iii) initially steeper for losses than for gains. 
•
→ Hence, a loss-averse worker will expend
more effort, leading to a higher productivity
rate on average under the Punishment
treatment compared to the Reward treatment.
 This table reveals that productivity
Main Result 2:  increases in the bonus treatments
for both individual and group data
are sizable
•the Reward and Punishment
treatments increase by 3.7%–8.6%
and 4.7%–9.7%, respectively.
•
For individual inspectors, the
Reward, Punishment, and Gift
treatments increase productivity by
5%–11.8%, 3.1%–13.5%, and
3.4%– 12.6%, respectively.
•
The Gift treatment, where workers
receive an unconditional gift as a
one-time bonus of RMB 320,
significantly increases productivity
compared to the Baseline
treatment.
 MAIN RESULTS 2
•For all groups, individual and teams, the monetary
incentives considerably enhanced productivity for, both
teams and individuals, even when the incentives were
provided unconditionally.
•Even though there are significant productivity shocks
across weeks, most of the variance in productivity arises
at a level well below or well above the target level. As a
result, whether the team or individual reaches the target
has less of an impact on productivity. This is important
because it means that our treatments are cost neutral. 
 Discussions
•First discussion: Why are there treatment effects in this experiment?
•Inexperienced workers have similar behaviours to students in lab
experiments when it comes to treatments
•In this experiment, there were lot of treatments such as the punishment
treatment that workers most likely have not experienced before.
•The empirical results are consistent with the first point, therefore the
treatments used have an effect.
•
Second discussion: Why the experiment worked better for groups than
individuals.
•Peer-monitoring
•Peer effects
•Guilt aversion
•
Third discussion: Whether incentives were profitable for the firm and whether
firms should adopt an incentive structure for the firm.
•Short answer: yes
•Long answer: yes shown in Main Result 1 and 2
 • This paper combines the understanding of productivity
enhancements with behavioral economics to determine whether
the knowledge of behavioral economics can explain or assist the
Conclusion understanding of productivity enhancements.
• This was done using the methodology presented earlier in this
presentation. In short, the simple framing manipulation changed
productivity by roughly 1% for teams of workers. This is
important as it comes with no extra cost only from the language
of the contract used in the experiment.
Pros and Cons

Pros: Cons:
•Innovative and reliable •Sample size of
approach that takes experiment, only
advantage of natural field 165 workers were
experiment  part of the
•Incorporation of
experiment  
baseline/control group and •Ethical concerns,
statistical analysis adds rigor manipulating
to study helps establish a worker behavior  
clearer causal relationship
between treatments and
productivity outcomes 
•Study provides insight into
long-term outcomes of
experiment 
THANK YOU!!!