Ideas Looking for Co-Authors
“Who Does Data Really Belong To?”
This paper explores a pair of complementary rights that have been proposed as ways of formally codifying the idea that “our data belongs to us,”: i) the right to retrieve personal information and data collected about ourselves by organizations operating in certain jurisdictions (e.g. the EU), and ii) to transfer information about ourselves from one organization to another in a format that is easy to utilize. There are many things to be said in favor of these rights. However, I argue that the idea that they are (or should be) grounded in the fact that our data belongs to us is problematic for a number of reasons. For one thing adjudicating claims about who data belongs to are complicated by the fact that much of the data that the proposed rights would pertain to would not exist if platforms did not collect it. Among other things this is because much of this data is not so much about us, as about how we interact with various platforms. Second, when the nature of big data is more closely examined, we see that entrenching the idea that individuals have (or should have) extensive rights to data portability may threaten to exacerbate many of the very market dynamics that proponents of this idea have worried about. That is, prioritizing accessibility and portability may come at the cost of concentrating the market power of the entities that create, provide, and maintain the systems architecture that allow data to be collected, stored, shared, and analyzed.
*Looking for someone familiar with the legal or economic literatures related to data, privacy, and/or database architecture.
“Evaluating the Impact of the Alaska Native Regional Corporations on Native Welfare”
This paper explores the impact that the Alaska Native Regional Corporations have had on the welfare of Alaskan Natives. The ANRCs were created in 1971 following the passage of the Alaska Native Claims Settlement Act which collectively awarded the existing Alaskan Native population $1 billion and the surface and subsurface rights to approximately 44 million acres in land. In order to manage this endowment, 12 regional corporations (and 200 village corporations) were created, with shares in the corporations assigned to Alaskan Natives currently living in the state according to where in the state they lived. The ANRCs are an interesting subject for two reasons. First, the size of the settlement awarded to the Alaskan Natives and the corporate governance structure they subsequently adopted distinguish Alaskan Natives from other groups of American Indians. Second, the ANRCs provide an interesting test case for the idea of "property-owning democracy" which has been proposed by several economists, philosophers, and political scientists as a potentially attractive alternative (or supplement) to existing welfare policies. One of the ANRCs, in particular, Cook Inlet Regional Inc (CIRI), provides an especially compelling case study as a result of a one-time $65,000 dividend that it paid out to each of its approximately 8,000 shareholders in 2000 following its sale of a large stake in a major telecommunications company.
*Looking for an economist or quantitative political scientist interested in collecting and analyzing the relevant data. I have some leads on how to get access to that data.
"Institutional P-Hacking"
As the frontiers of scientific research have gotten more complex, research has increasingly come to embrace specialization and the division of labor. One way in which this manifests itself is for researchers to farm out data analysis to statisticians. This is especially common in medical research where a significant amount of research is done by clinicians whose primary responsibility is patient care and who often lack the requisite background in mathematics needed to analyze their data (although, of course, this isn’t always true and this phenomenon isn’t unique to medical research). This sort of division of labor in research has greatly expanded the research possibility frontier. However, there is a cost to expanding the production possibility frontier in this way that hasn’t been sufficiently appreciated. That cost is that when research teams are designed in such a way that data analysis is farmed out to specialists, those specialists aren’t put in a position to do good analysis. The reason for that is that good data analysis typically requires understanding enough of the field specific background science to know how to account for underlying relationships in the data being analyzed. But this is precisely what specialists in data analysis don’t have (nor should we expect them too!). Nor can this problem be solved by relying on the researchers responsible for collecting data (in our case clinicians) to fill in the gaps in the background knowledge the statistician needs, because filling in those gaps requires the clinician to know enough about the requisite statistical techniques to know what kind of underlying relationships need to be controlled for. To put a slightly hyperbolic (but only slightly) gloss on things we might say that in order to avoid the problem of unintentional p-hacking by naïve researchers, we’ve designed teams in such a way that the team dynamics reproduce the very problem we were trying to solve. And this problem is made all the worse by the fact that the move to team research brings with it the cover of scientific authority. After all, the data analysis is being farmed out to experts! Of course, this doesn’t mean that research shouldn’t be conducted in teams. Instead the takeaway is that when it comes to designing research paradigms we ought to be thinking as hard about how we analyze data as about how we collect that data. And while institutions have made progress on this front - for instance many research hospitals now have science review boards in addition to their ethics reviews – those solutions carry their own costs and can only be counted on to catch so much.
*Looking for a philosopher of science or someone else with experience in medical research and/or biostatistics.
“Who Does Data Really Belong To?”
This paper explores a pair of complementary rights that have been proposed as ways of formally codifying the idea that “our data belongs to us,”: i) the right to retrieve personal information and data collected about ourselves by organizations operating in certain jurisdictions (e.g. the EU), and ii) to transfer information about ourselves from one organization to another in a format that is easy to utilize. There are many things to be said in favor of these rights. However, I argue that the idea that they are (or should be) grounded in the fact that our data belongs to us is problematic for a number of reasons. For one thing adjudicating claims about who data belongs to are complicated by the fact that much of the data that the proposed rights would pertain to would not exist if platforms did not collect it. Among other things this is because much of this data is not so much about us, as about how we interact with various platforms. Second, when the nature of big data is more closely examined, we see that entrenching the idea that individuals have (or should have) extensive rights to data portability may threaten to exacerbate many of the very market dynamics that proponents of this idea have worried about. That is, prioritizing accessibility and portability may come at the cost of concentrating the market power of the entities that create, provide, and maintain the systems architecture that allow data to be collected, stored, shared, and analyzed.
*Looking for someone familiar with the legal or economic literatures related to data, privacy, and/or database architecture.
“Evaluating the Impact of the Alaska Native Regional Corporations on Native Welfare”
This paper explores the impact that the Alaska Native Regional Corporations have had on the welfare of Alaskan Natives. The ANRCs were created in 1971 following the passage of the Alaska Native Claims Settlement Act which collectively awarded the existing Alaskan Native population $1 billion and the surface and subsurface rights to approximately 44 million acres in land. In order to manage this endowment, 12 regional corporations (and 200 village corporations) were created, with shares in the corporations assigned to Alaskan Natives currently living in the state according to where in the state they lived. The ANRCs are an interesting subject for two reasons. First, the size of the settlement awarded to the Alaskan Natives and the corporate governance structure they subsequently adopted distinguish Alaskan Natives from other groups of American Indians. Second, the ANRCs provide an interesting test case for the idea of "property-owning democracy" which has been proposed by several economists, philosophers, and political scientists as a potentially attractive alternative (or supplement) to existing welfare policies. One of the ANRCs, in particular, Cook Inlet Regional Inc (CIRI), provides an especially compelling case study as a result of a one-time $65,000 dividend that it paid out to each of its approximately 8,000 shareholders in 2000 following its sale of a large stake in a major telecommunications company.
*Looking for an economist or quantitative political scientist interested in collecting and analyzing the relevant data. I have some leads on how to get access to that data.
"Institutional P-Hacking"
As the frontiers of scientific research have gotten more complex, research has increasingly come to embrace specialization and the division of labor. One way in which this manifests itself is for researchers to farm out data analysis to statisticians. This is especially common in medical research where a significant amount of research is done by clinicians whose primary responsibility is patient care and who often lack the requisite background in mathematics needed to analyze their data (although, of course, this isn’t always true and this phenomenon isn’t unique to medical research). This sort of division of labor in research has greatly expanded the research possibility frontier. However, there is a cost to expanding the production possibility frontier in this way that hasn’t been sufficiently appreciated. That cost is that when research teams are designed in such a way that data analysis is farmed out to specialists, those specialists aren’t put in a position to do good analysis. The reason for that is that good data analysis typically requires understanding enough of the field specific background science to know how to account for underlying relationships in the data being analyzed. But this is precisely what specialists in data analysis don’t have (nor should we expect them too!). Nor can this problem be solved by relying on the researchers responsible for collecting data (in our case clinicians) to fill in the gaps in the background knowledge the statistician needs, because filling in those gaps requires the clinician to know enough about the requisite statistical techniques to know what kind of underlying relationships need to be controlled for. To put a slightly hyperbolic (but only slightly) gloss on things we might say that in order to avoid the problem of unintentional p-hacking by naïve researchers, we’ve designed teams in such a way that the team dynamics reproduce the very problem we were trying to solve. And this problem is made all the worse by the fact that the move to team research brings with it the cover of scientific authority. After all, the data analysis is being farmed out to experts! Of course, this doesn’t mean that research shouldn’t be conducted in teams. Instead the takeaway is that when it comes to designing research paradigms we ought to be thinking as hard about how we analyze data as about how we collect that data. And while institutions have made progress on this front - for instance many research hospitals now have science review boards in addition to their ethics reviews – those solutions carry their own costs and can only be counted on to catch so much.
*Looking for a philosopher of science or someone else with experience in medical research and/or biostatistics.
