Doctoral Dissertation Research in economics: Examining the Impact of Hybrids on the U.S. Automobile Market

建议编号：SE-1327008机构：耶鲁大学NSF计划：主要研究员：贝瑞，史蒂文·T标题：经济学博士论文研究：考察混合动力车对美国汽车市场的影响该项目的目标是通过回答两个问题来确定混合动力车对美国汽车市场的影响：消费者从混合动力车的创造中获得了多大的效用，因此，美国整个车队的构成随着时间的推移发生了怎样的变化？混合动力车目前在汽车车队中占据了相当大的且不断增长的部分，由于它们在替代能源和环境保护政策中的相关性，近年来受到了极大的关注。然而，尽管混合动力车已经纳入了汽油税或对汽车制造商实施的燃油效率标准的影响的讨论中，但对混合动力车的整体好处的分析仍然不完整。通过使用跟踪家庭十年汽车保有量的数据，该项目将估计混合动力车的创造如何使整个国家受益，以及市场上进一步的技术进步可以带来多少福利。具体地说，该项目着眼于1999-2009年，并检查了2004年丰田普锐斯的发布和混合动力车的总体发布，这两者都发生在观察到的时间段。2004年的普锐斯特别有趣，因为它是美国第一款在商业上成功的混合动力车，也是每年最畅销的混合动力车。通过估计如果没有普锐斯和其他混合动力车，消费者可能会有什么不同的购买方式，研究可以确定混合动力车给市场带来了多少价值。与大多数汽车市场研究中使用的数据集相比，这个项目使用的数据集有两个明显的优势。首先，这些数据跟踪了同一个家庭一段时间的情况。大多数关于汽车市场的研究都基于多年的数据，但从每年的家庭中随机抽样。由于每一年的数据都包含一组新的家庭，每个家庭？S都有独特的难以察觉的品味(例如，热爱快速、强大的汽车；决心？将汽车开进地下？)无法在分析中控制。因此，随着时间的推移，这些品味的影响与购买汽车的影响合并在一起。然而，通过对每个家庭多年的识别和控制，该项目可以控制口味效应，根据需要隔离车辆效果。第二，尽管大多数数据集在年份、制造商和型号的水平上区分车辆，但用于本研究的数据也区分了汽油和混合动力内饰。2000年代末市场上的许多混合动力车是汽油动力汽车的替代内饰，如本田思域(Honda Civic)和丰田凯美瑞(Toyota Camry)。有了数据中的这一特征，混合动力车可以与其他车辆区分开来，这使得拟议的分析成为可能。此外，该项目使用的新模型允许过去的购买直接影响当前的购买，这一特征不被最近的计量经济学文献使用的模型所共享。以前，对车队的影响估计只有在影响十年或更长时间后才适用，因为所有消费者都需要有机会对变化做出反应。通过检查过去的购买情况，此项目中的模型控制家庭吗？无论对市场的影响如何，都倾向于将汽车持有数年，这使得对冲击后两到十年的预测成为可能。新模式的用处并不局限于汽车市场；它可以应用于许多耐用品市场(例如，家用电器、计算机)。

AbstractProposal No: SES - 1327008 Institution: Yale UniversityNSF Program: ECONOMICSPrincipal Investigator: Berry, Steven T.Title: Doctoral Dissertation Research in economics: Examining the Impact of Hybrids on the U.S. Automobile MarketThe goal of the project is to determine the impact of hybrids on the United States automobile market by answering two questions: how much utility have consumers gained from the creation of hybrids, and as a result, how has the composition of the overall U.S. vehicle fleet changed over time?Hybrids, which currently make up a significant and growing portion of the vehicle fleet, have received a great deal of attention in recent years due to their relevance in alternative energy and environmental protection policies. However, while hybrids have factored into discussions on the impacts of gasoline taxes or fuel-efficiency standards imposed on auto manufacturers, the analysis of the overall benefits of hybrids remains incomplete.By using data which track the vehicle holdings of households across a ten-year period, this project will estimate how the creation of hybrids have benefited the nation as a whole, and how much welfare can be expected from further technological advancements in the market. Specifically, the project looks at the period of 1999-2009 and examines the release of the 2004 Toyota Prius and the release of hybrids in general, both of which occur during the observed time. The 2004 Prius is particularly interesting, as it is the first commercially successful hybrid in the United States, and has been the best-selling hybrid in every year. By estimating how consumers might have purchased differently if the Prius and other hybrids had not been available, the research can determine how much value hybrids bring to the market.The datasets used for this project have two distinct advantages over those used in most studies of the automobile market. First, the data track the same households over time. Most of the research done on automobile markets draw on multiple years of data, but take a random sample of households from each year. Since each year of the data contains a new set of households, each family?s unique unobservable tastes (e.g., a love of fast, powerful cars; the resolve to ?drive a car into the ground?) cannot be controlled for in the analysis. As a result, the effect of these tastes merges with the effect of vehicle purchases over time. However, by identifying and controlling for each household over multiple years, this project can control for the tastes effect, isolating the vehicle effect as desired.Second, while most datasets differentiate between vehicles at the level of year, make, and model, the data used for this research also differentiates between gasoline and hybrid trims. Many of the hybrids on the market in the late 2000s are alternative trims of gasoline-powered cars, such as the Honda Civic and Toyota Camry. With this feature in the data, hybrids can be distinguished from other vehicles, making the proposed analysis possible.Additionally, the new model used by the project allows past purchases to directly influence current purchases, a feature not shared by models used by the recent econometrics literature. Previously, estimates of impacts to the vehicle fleet were only applicable ten or more years after the impact, since all consumers needed to be given a chance to respond to the change. By examining past purchases, the model in this project controls for households? tendencies to hold onto cars for several years regardless of impacts to the market, making forecasts two to ten years after the impact possible. The usefulness of the new model is not restricted to the automobile market; it can be applied to many durable-goods markets (e.g., appliances, computers).