Loading...
Please wait, while we are loading the content...
Similar Documents
Supplemental Online Appendix for “ Trading Across Borders in Online Auctions ”
| Content Provider | Semantic Scholar |
|---|---|
| Author | Krasnokutskaya, Elena A. Terwiesch, Christian Tiererova, Lucia |
| Copyright Year | 2017 |
| Abstract | Recovering demand-side primitives presents several methodological challenges. To see this let us first consider an environment without transitory sellers (or where transitory sellers differ only in their observable attributes, x). In this setting we only need to focus on recovering the quality levels (fixed effects) of permanent sellers. Recall that in a traditional discrete choice setting fixed effects associated with different alternatives are identified from the observed probabilities that a given alternative is chosen, conditional on the choice set. In our setting, choice sets are buyer-specific since the sellers’ participation varies across auctions. Due to the large numbers of sellers and buyers conditional choice probabilities cannot be precisely estimated. To get a sense of the magnitudes consider that the number of permanent sellers present in the market for a given type of work is around 300 to 500 whereas only 2 or 3 permanent sellers participate in any given auction. This means that the number of possible choice sets is at least C 300 = 300! 3!297! = 300∗299∗298 3 = 8, 910, 200 which exceeds the number of projects we have in our dataset. In fact, the highest number of projects sharing the same set of participating permanent sellers in our data is five. One way to deal with this issue would be to consider probabilities that aggregate |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | https://assets.aeaweb.org/assets/production/files/8204.pdf |
| Alternate Webpage(s) | http://www.econ2.jhu.edu/people/Krasnokutskaya/OnlineApp_0309.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
