Andrew Harkins Half Flashcards
Coase (1972)
Dynamic Monopoly, monopolist will set competitive prices in all periods as t tends to infinite…consumers foresee future prices decreases
May try to circumvent this via contractual agreements to not sell any excess land, or agree to buy back any land…but these are costly and may not be legally enforceable or implementable.
Klein, Lambertz & Stahl (2016)
Reputation: Exploit the Change in Reputation System on eBay - Classic feedback system changing in 2008 to leave only + reviews and reduce retaliation, Detailed Seller ratings as a control.
DSRs show the average seller behaviour, while the classid ratings show the seller’s marginal behaviour…on the most recent transactions
Results:
- Seller scores increased, addressing Moral Hazard
- No sudden jump in quality suggesting Adverse Selection still exists
- Exogenous increase in market transparency led to a significant increase in buyer satisfaction, but no change in the seller exit rate
Extras:
15,000 sellers, focusing on larger sellers grouped into 5 categories, July 2006-2009, Feedback pages give a score and a “DSR”…exiting the market is interpreted as no longer getting feedback
Mayzlin, Dover & Chevalier (2014)
Exploit differences between Open TripAdvisor and Verified Expedia to investigate reviews.
Null Hypotheses:
- Local Competition raises fake reviews,
- More costly for chains due to reputational damage,
- Hotels close to indepedent will be targeted
Results:
Insignificant for 1 time reviews - hard to distinguish between paid fakers with credible profiles….doesn’t control for reviewer characteristics only hotel.
Promotional reviews not only may mislead consumers into making suboptimal purchasing decisions but also cause a self-fulfilling breakdown of buyer and seller trust and a lack of faith in the review system
Analyse only the 25th-75th percentiles of largest US cities to avoid over densely populated areas which may muddy the impact of local competition increasing fake reviews
Aghion & Tirole (1997)
When is it best for a principal to delegate?
1: Agent cares more for the project
2: Agent is more trustworthy - … prinicipal likes the project the agent would choose better
3: Less congruence of preference between Agent and Principal…efforts are strategic substitutes, so if there is a low degree of preference alignment the agent exerts high effort meaning the principal can shirk
Why Delegate?
- Initiative and incentive benefits from the agent
- Facilitates agent participation in contractual relationships
Anderson & Coate (2005)
Welfare analysis of Hotelling application to 2 sided platform networks
Carpenter, Schirm and Matthews (2010)
Piece Rate vs. Tournament vs. PR + Sabotage vs. T + Sabotage.
Results:
- Bonus (Tournament) increases effort,
- Sabotage (with T only) decreases effort.
- More sabotage on quality than quantity,
- Sabotage prevlaent when they have been beated
- Between worker competition should be avoided, so called “Office politics” cause negative peer reviews when any ambiguity is involved
- Time gets spent on unproductive tasks, morale and atmosphere deteriorate which greatly disincentivise effort levels
Repetitive mail organisation task, paid accoridng to quality weighted output, 224 subjects, 30 minute task, 8 groups, 60-90 seconds per task, independent mailman as control
Gentzkow & Shapiro (2006)
Media bias, reports and publications consistent with our prior beliefs are more likely to be interpreted as coming from an expert or credible source.
Results:
- Perfectly Accurate Signal: normal firms also report truthfully
- Not perfect: normal firms report truthfully on the likely event, mis-report on the unlikely to try and gain a reputational benefit
- Higher prior beliefs lead to greater mis-reporting, more reputational benefits
- If priors vary; we can get some market segmentation of bias being reported to the types with that prior, competition between reporters can reduce bias
- Bias in reporting occurs even though it can make all participants worse off
- Misreporting incentives weaken as the ex-post verification of events increases eg. Sports outcomes, weahter and stock returns see less bias but… war, tax policy alternatives or global warming sees more
eg. American tropps in Samarra during 2003, Fox News vs. New York Times vs. Al Jazeera
Blake, Nosko & Tadelis (2015)
Online vs. offline ads, easier to track RoI online. Brand keywords such as eBay compared with non-brand like Shoes.
Results:
- Paid ads are ineffective: only work for those how haven’t purchased recently or are not frequent purchasers on non-brand keywords
- Those who buy would have bought anyway, switching off paid ads is almost entirely offset by usual ads
- Brand keyword ads have no impact at all
- DiD 0.529% fall in visits, hugely insignificant
- Are long term benefits the compensation for the vast sums spent? Or is it just a case of the Prisoner’s Dilemma?
Issue:
Endogeneity of eBay and key word searches
Chevalier et. al (2003)
Why prices don’t rise during peak demand? Chicago scanner data, unique data on margins to rule out supply shifts in cohesion with data on advertising showing tuna is pushed as a substitute during lent, cross sell with higher margins to cover
3 Potential Reasons:
- Search Costs: more elastic during peak periods as we search more
2 - Collusion: harder to sustain during peak time as deviation more profitable,
- Loss Leading: specific to multi product sellers, bait and switch tactics….most consistent explanation! Prices tied to product seasonality not AD
(4. ) - Increasing Returns to Scale: Not discussed largely as not tied concretely to the seasonal variation, plausible magnitudes and indistinghuisble differences in mechanisms
Extras:
Chicago supermarket chains, 100+ stores, 400 weeks, 100 product groups, 1989-1997, have retail and wholesale prices, grouping items into seasonal demands…thanksgiving, weather and lent
Becker & Millbourn (2011)
Increased competition in the ratings agencies after Fitch market share growth.
H0: Competition decreases ratings as a reputation is less valuable when rent get competed away.
Ratings Quality Evaluated on:
Accuracy - More defaults among investment grade so precision was lost
Stability - Ratings became unstable, qualities dramatically increased
Significant magnitude of effect: a one standard deviation increase in Fitch’s market share is precicted to increase the average firm and bond rating by 1/10 and 1/2 step…although this is significantly more for highly levered firms
Issues?
- Ratings shopping: not an issue here as S&P list all info, no selection bias
- Omitted variables linked to higher competition, lower accuracy and instability; eg. rapid growth industry issuing lots of bonds to finance investment increasing risk but ratings agencies want to win the contract to rate them so overbid each other in terms of rating
Reversed:
Dynamic Monopoly, monopolist will set competitive prices in all periods as t tends to infinite…consumers foresee future prices decreases
May try to circumvent this via contractual agreements to not sell any excess land, or agree to buy back any land…but these are costly and may not be legally enforceable or implementable.
Coase (1972)
Reversed:
Reputation: Exploit the Change in Reputation System on eBay - Classic feedback system changing in 2008 to leave only + reviews and reduce retaliation, Detailed Seller ratings as a control.
DSRs show the average seller behaviour, while the classid ratings show the seller’s marginal behaviour…on the most recent transactions
Results:
- Seller scores increased, addressing Moral Hazard
- No sudden jump in quality suggesting Adverse Selection still exists
- Exogenous increase in market transparency led to a significant increase in buyer satisfaction, but no change in the seller exit rate
Extras:
15,000 sellers, focusing on larger sellers grouped into 5 categories, July 2006-2009, Feedback pages give a score and a “DSR”…exiting the market is interpreted as no longer getting feedback
Klein, Lambertz & Stahl (2016)
Reversed:
Exploit differences between Open TripAdvisor and Verified Expedia to investigate reviews.
Null Hypotheses:
- Local Competition raises fake reviews,
- More costly for chains due to reputational damage,
- Hotels close to indepedent will be targeted
Results:
Insignificant for 1 time reviews - hard to distinguish between paid fakers with credible profiles….doesn’t control for reviewer characteristics only hotel.
Promotional reviews not only may mislead consumers into making suboptimal purchasing decisions but also cause a self-fulfilling breakdown of buyer and seller trust and a lack of faith in the review system
Analyse only the 25th-75th percentiles of largest US cities to avoid over densely populated areas which may muddy the impact of local competition increasing fake reviews
Mayzlin, Dover & Chevalier (2014)
Reversed:
When is it best for a principal to delegate?
1: Agent cares more for the project
2: Agent is more trustworthy - … prinicipal likes the project the agent would choose better
3: Less congruence of preference between Agent and Principal…efforts are strategic substitutes, so if there is a low degree of preference alignment the agent exerts high effort meaning the principal can shirk
Why Delegate?
- Initiative and incentive benefits from the agent
- Facilitates agent participation in contractual relationships
Aghion & Tirole (1997)
Reversed:
Welfare analysis of Hotelling application to 2 sided platform networks
Anderson & Coate (2005)
