ERC FUNDED PROJECTS

This project studies information aggregation in multiple-linked auction markets with large numbers of goods and bidders. Past work assumes bidders trade in a single, centralized, frictionless auction market. Instead, I study bidders with unit demand who decide to purchase one of many possible goods which are on auction in distinct markets. The goods traded in each market are identical, common-value objects and the price is determined by a uniform-price auction. Bidders receive imperfect signals about the state of the world and select to bid in one of the auction markets. The markets differ in institutional structure and therefore frictions. Market frictions result from imperfect competition, government interventions, informational frictions, and preference heterogeneity. All such frictions render the gains from trade uncertain. I address the following questions: How do market frictions affect information aggregation if bidders can strategically choose between markets? What are the mechanisms through which market imperfections disrupt information aggregation? Which market’s price is a better statistic for market participants’ information? Which market attracts better-informed bidders? Do prices aggregate beliefs more accurately in good times or in bad? Initial findings suggest that the proposed framework can prove particularly fruitful in addressing these questions. Specifically, I show if the gains from trade are uncertain in even one market, then prices do not aggregate information in any of the markets. In contrast, if all markets are frictionless, then the price in each market aggregates information. These findings are driven by how bidders self-select across markets: Better-informed bidders select frictional markets while uninformed, pessimistic bidders select the safety of frictionless markets. These findings suggest a novel mechanism through which market imperfections in one market can have widespread effects across all linked markets.

1 089 432 €

Start date: 2016-05-01, End date: 2021-04-30

"Control of matter via light has always fascinated humankind; not surprisingly, laser patterning of materials is as old as the history of the laser. However, this approach has suffered to date from a stubborn lack of long-range order. We have recently discovered a method for regulating self-organised formation of metal-oxide nanostructures at high speed via non-local feedback, thereby achieving unprecedented levels of uniformity over indefinitely large areas by simply scanning the laser beam over the surface. Here, we propose to develop hitherto unimaginable levels of control over matter through laser light. The total optical field at any point is determined by the incident laser field and scattered light from the surrounding surface, in a mathematical form similar to that of a hologram. Thus, it is only logical to control the self-organised pattern through the laser field using, e.g., a spatial light modulator. A simple wavefront tilt should change the periodicity of the nanostructures, but much more exciting possibilities include creation of patterns without translational symmetry, i.e., quasicrystals, or patterns evolving non-trivially under scanning, akin to cellular automata. Our initial results were obtained in ambient atmosphere, where oxygen is the dominant reactant, forming oxides. We further propose to control the chemistry by using a plasma jet to sputter a chosen reactive species onto the surface, which is activated by the laser. While we will focus on the basic mechanisms with atomic nitrogen as test reactant to generate compounds such as TiN and SiN, in principle, this approach paves the way to synthesis of an endless list of materials. By bringing these ideas together, the foundations of revolutionary advances, straddling the boundaries of science fiction, can be laid: laser-controlled self-assembly of plethora of 2D patterns, crystals, and quasicrystals alike, eventually assembled layer by layer into the third dimension -- a 3D material synthesiser."

1 999 920 €

Start date: 2014-06-01, End date: 2019-05-31

This project studies the causal role of peer interactions and the nature of the classroom environment in the formation of individual beliefs, attitudes and non-cognitive skills in childhood, with a view towards (1) mitigating gender gaps in choices and outcomes, (2) improving attitudes towards immigrants and promoting integration, (3) improving achievement outcomes. To this end, the project will design randomized interventions on social interactions in the classroom, that exogenously manipulate (i) exposure to peers (randomizing gender and ethnicity combinations), (ii) the setting in which the interaction takes place (cooperative vs. competitive). The conjecture is that particular types of peer interaction (e.g. cooperative work) can permanently shift both individual preferences and attitudes toward others, and lead to contagion of non-cognitive skills. The interventions will be embedded into a novel extracurricular program to be designed around teaching “coding”, and” and will be implemented across a large set of elementary schools in Istanbul, for the duration of a semester. The project will utilize an interdisciplinary methodology that involves collecting a large set of outcome measures through novel incentivized experiments, questionnaires, and neural measurements. Impact of different types of peer interactions and a cooperative vs. competitive classroom culture on attitudes towards gender and immigrants, individual preferences and non-cognitive skills, creativity, aspirations and achievement will be measured both in the short- and the longer-run, in an entirely new social context. Comprehensive baseline data on individual, family and classroom characteristics will allow us to study heterogeneous effects and potential mechanisms. The results will provide inputs for how the educational environment can be designed early on, to build a culture that promotes gender equality, cultivates positive attitudes towards immigrants, and improves achievement outcomes.

2 000 000 €

Start date: 2020-11-01, End date: 2025-10-31