Abstract: This dissertation deals with energy demand and consists of two parts. Part one proposes a unified econometric framework for modeling energy demand and examples illustrate the benefits of the technique by estimating the elasticity of substitution between energy and capital. Part two assesses the energy conservation policy of Daylight Saving Time and empirically tests the performance of electricity simulation. In particular, the chapter 'Imposing Monotonicity and Curvature on Flexible Functional Forms' proposes an estimator for inference using structural models derived from economic theory. This is motivated by the fact that in many areas of economic analysis theory restricts the shape as well as other characteristics of functions used to represent economic constructs. Specific contributions are (a) to increase the computational speed and tractability of imposing regularity conditions, (b) to provide regularity preserving point estimates, (c) to avoid biases existent in previous applications, and (d) to illustrate the benefits of our approach via numerical simulation results. The chapter 'Can We Close the Gap between the Empirical Model and Economic Theory' discusses the more fundamental question of whether the imposition of a particular theory to a dataset is justified. I propose a hypothesis test to examine whether the estimated empirical model is consistent with the assumed economic theory. Although the proposed methodology could be applied to a wide set of economic models, this is particularly relevant for estimating policy parameters that affect energy markets. This is demonstrated by estimating the Slutsky matrix and the elasticity of substitution between energy and capital, which are crucial parameters used in computable general equilibrium models analyzing energy demand and the impacts of environmental regulations. Using the Berndt and Wood dataset, I find that capital and energy are complements and that the data are significantly consistent with duality theory. Both results would not necessarily be achieved using standard econometric methods. The final chapter 'Daylight Time and Energy' uses a quasi-experiment to evaluate a popular energy conservation policy: we challenge the conventional wisdom that extending Daylight Saving Time (DST) reduces energy demand. Using detailed panel data on half-hourly electricity consumption, prices, and weather conditions from four Australian states we employ a novel 'triple-difference' technique to test the electricity-saving hypothesis. We show that the extension failed to reduce electricity demand and instead increased electricity prices. We also apply the most sophisticated electricity simulation model available in the literature to the Australian data. We find that prior simulation models significantly overstate electricity savings. Our results suggest that extending DST will fail as an instrument to save energy resources.