The study of macroeconomics and finance has evolved tremendously over the last few decades---significant advancements have taken place in both gaining access to an increasing scale and scope of observational, policy and private data, as well as empirical methods to derive novel economic insights from such data. In this dissertation, I attempt to shed light on three problems broadly across macroeconomics and asset pricing, taking a data-driven approach to answer them.
For the first essay, we construct a novel dataset which captures the geographic incidence of government revenues and expenditures. Government revenues and expenditures revenues and expenditures occur on three different levels in the United States: local, state, and federal. At all levels, government revenues and expenditures add and subtract resources from the private sector. The dataset encompasses all revenues and expenditures at the county-level and thus allows to see the net resource allocation through the government. We use this dataset to document several new facts about the relationship between economic activity and the resource allocation by the government. The governments' resource allocation is generally redistributive. That is, levels and changes of median income are associated with the level and changes of net resources. Second, we evaluate response of governmental revenues and spending in response to the China shock. We find a decline in total governmental receipts in counties that are hardest hit, while a muted response of total governmental spending. The aggregate response conceals a lot of heterogeneity: a decomposition at the governmental level shows an increase in expenditures and lower receipts at the federal level; at the local and state level we find a simultaneous reduction of receipts and spending. The latter is a consequence of the balanced budget constraint. Overall, total government spending is approximately constant while total receipts are falling. As a result, the insurance function of the federal government is offset by a reduction at the state and local level which renders total government spending neutral to the China shock. This stands in contrast to prior research which has focused on the federal response.
In our second essay, we attempt to answer the question---how should an investor value financial data? The answer is complicated because it depends on the characteristics of all investors. We develop a sufficient statistics approach that uses equilibrium asset return moments to summarize all relevant information about others' characteristics. It can value data that is public or private, about one or many assets, relevant for dividends or for sentiment. While different data types, of course, have different valuations, heterogeneous investors also value the same data very differently, which suggests a low price elasticity for data demand. Heterogeneous investors' data valuations are also affected very differentially by market illiquidity.
Lastly, in the third essay, we examine the economic impact of droughts on asset markets, specifically on land valuation. Specifically, we focus on farmland valuations in California---one of the most productive farmlands in the world. The semi-arid climate makes its valuation particularly sensitive to the amount of surface and groundwater water available for irrigation. The detailed administrative transaction data from the counties' assessor offices allows us to estimate repeat sales indices as opposed to a hedonic model which make our results less likely to be affected by omitted variables. We find that parcels with better access to freshwater see a larger appreciation in land values from 2011 to 2020; whereas we find no statistical significant differential price change between 2000-2011. The differential change in land values points towards large economic effects of water scarcity with beliefs about future climatic conditions being updated due to two severe episodes of drought and signals of legislative willingness to curb groundwater overdraft.
| Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/2fhr-fv05 |
| Date | January 2024 |
| Creators | Singal, Dhruv |
| Source Sets | Columbia University |
| Language | English |
| Detected Language | English |
| Type | Theses |
Page generated in 0.0017 seconds