2026-09-22T16:09:40Zhttps://dspace.mit.edu/server/oai/requestoai:dspace.mit.edu:1721.1/1270312026-06-17T14:44:40Zcom_1721.1_7582com_1721.1_7581col_1721.1_131022
Sara Ellison and Nikhil Agarwal.
Anghel, Anca-Patricia.
Massachusetts Institute of Technology. Department of Economics.
Massachusetts Institute of Technology. Department of Economics
2020-09-03T17:43:04Z
2020-09-03T17:43:04Z
2020
2020
https://hdl.handle.net/1721.1/127031
1191625536
Thesis: Ph. D., Massachusetts Institute of Technology, Department of Economics, May, 2020
Cataloged from the official PDF of thesis.
Includes bibliographical references.
This thesis consists of three chapters. The αrst chapter and second chapter contribute to the literature on consumer search, while the third chapter analyzes the e
The optimal consumer behavior for a model with normal priors is characterized and the likelihood function is constructed. The learning model generates predictions that can be used to test whether consumers are learning. These predictions are conαrmed by the data. Using clickstream data on search and purchase behavior of consumers looking to book a hotel on Expedia, the theoretical model is leveraged to estimate consumer preferences, a clicking cost distribution, and prior beliefs parameters. Consumers have overoptimistic prior beliefs about the distribution of the unobservable characteristics compared to the true distribution. Thus, consumers click on more hotels compared to when they know the true distribution. However, the estimated clicking costs are too high for consumers to click enough to αnd the optimal hotels for them. The results suggest that the consumer experience could be improved by personalized recommendations and an improved selection of hotels.
distribution in demand models with sequential search. I show that if consumers need to pay a search cost to αgure out some product characteristics, aggregate data is not enough to fully non-parametrically identify the demand and the search cost distribution. I show that the distribution of the search cost can be identiαed in the mixed logit models using only aggregate data. Moreover, if micro data is available, the demand functions and search cost distribution cannot be identiαed fully non parametrically. However, if data on the search behavior of consumers is available (clickstream data), I provide a proof of non-parametric identiαcation of the demand function, distribution of utility, and search cost distribution. In the third chapter, I use data from the CPS March survey from 1990 to 2014 to analyze the e
It has been suggested that introducing mandatory paid parental leave will have a negative impact on women's wages since women are more likely to take advantage of this policy and since this will result in increased costs to employers. The data is analyzed using the synthetic control method proposed by Abadie and Gardeazabal (2003). There are no signiαcant short nor long term effects on women's wages or employment rates following the introduction of this program.
by Anca-Patricia Anghel.
Ph.D.
Ph.D. Massachusetts Institute of Technology, Department of Economics
169 pages
eng
Massachusetts Institute of Technology
MIT theses may be protected by copyright. Please reuse MIT thesis content according to the MIT Libraries Permissions Policy, which is available through the URL provided.
http://dspace.mit.edu/handle/1721.1/7582
Economics.
Three essays in economics/
Thesis
application/pdf
2020-09-03T17:43:04Z
Publication
Doctoral
Econ
unknown
unknown
<Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="94c0e612-a5af-432a-b95a-91cb520152b6">
<Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type>
<Language>eng</Language>
<Title>Three essays in economics/</Title>
<PublishedIn>
<Publication>
</Publication>
</PublishedIn>
<PublicationDate>2020</PublicationDate>
<Authors>
<Author>
<DisplayName>Anghel, Anca-Patricia.</DisplayName>
<Affiliation>
<OrgUnit>
</OrgUnit>
</Affiliation>
</Author>
</Authors>
<Editors>
</Editors>
<Publishers>
<Publisher>
<DisplayName>Massachusetts Institute of Technology</DisplayName>
<OrgUnit />
</Publisher>
</Publishers>
<License>http://dspace.mit.edu/handle/1721.1/7582</License>
<Keyword>Economics.</Keyword>
<Abstract>This thesis consists of three chapters. The αrst chapter and second chapter contribute to the literature on consumer search, while the third chapter analyzes the e</Abstract>
<Abstract>The optimal consumer behavior for a model with normal priors is characterized and the likelihood function is constructed. The learning model generates predictions that can be used to test whether consumers are learning. These predictions are conαrmed by the data. Using clickstream data on search and purchase behavior of consumers looking to book a hotel on Expedia, the theoretical model is leveraged to estimate consumer preferences, a clicking cost distribution, and prior beliefs parameters. Consumers have overoptimistic prior beliefs about the distribution of the unobservable characteristics compared to the true distribution. Thus, consumers click on more hotels compared to when they know the true distribution. However, the estimated clicking costs are too high for consumers to click enough to αnd the optimal hotels for them. The results suggest that the consumer experience could be improved by personalized recommendations and an improved selection of hotels.</Abstract>
<Abstract>distribution in demand models with sequential search. I show that if consumers need to pay a search cost to αgure out some product characteristics, aggregate data is not enough to fully non-parametrically identify the demand and the search cost distribution. I show that the distribution of the search cost can be identiαed in the mixed logit models using only aggregate data. Moreover, if micro data is available, the demand functions and search cost distribution cannot be identiαed fully non parametrically. However, if data on the search behavior of consumers is available (clickstream data), I provide a proof of non-parametric identiαcation of the demand function, distribution of utility, and search cost distribution. In the third chapter, I use data from the CPS March survey from 1990 to 2014 to analyze the e</Abstract>
<Abstract>It has been suggested that introducing mandatory paid parental leave will have a negative impact on women's wages since women are more likely to take advantage of this policy and since this will result in increased costs to employers. The data is analyzed using the synthetic control method proposed by Abadie and Gardeazabal (2003). There are no signiαcant short nor long term effects on women's wages or employment rates following the introduction of this program.</Abstract>
<Access xmlns="http://purl.org/coar/access_right"
>
</Access>
</Publication>