本文将消费者产生的数据作为一种关键要素引入到知识积累过程当中，并由此构建了一个全新的内生经济增长模型，是数据经济的长期效应领域的奠基性研究之一。在这个模型中，消费者的效用由最终品的消费及提供的数据量共同决定，受雇于生产部门或R&D部门，并且需要在提供数据带来的收益与数据被使用导致的潜在隐私侵害之间做出权衡。最终品生产厂商通过雇用劳动力和中间品在完全竞争的环境下生产。中间品生产厂商在进入市场之前首先使用劳动力和数据进行研发活动，取得成功之后即获得相应的专利权，并可在垄断的环境下生产。结果表明：（1）去中心化经济能够达到与社会最优情形相同的平衡增长路径（BGP）的增长率水平，并且在消费、技术水平等的BGP增长率为正的同时，数据提供量的BGP增长率为负；（2）虽然两种情形下的BGP增长率相同，但是与社会最优的分配相比，去中心化经济下的R&D部门仍然存在着劳动力雇佣不足和数据使用过度的问题。 在进一步的扩展中，本文还考虑了数据的动态非竞争性对结果的影响。在允许不同进入时间的厂商之间的数据交易并考虑创造性破坏风险的条件下，我们得到了关于跨期数据交易行为的新结论：除非创造性破坏发生的概率升高到了极端的情形下，去中心化经济中的在位者总是愿意出售多于社会最优水平的数据量给进入者的。除此之外，本文从多个方面进行了深入讨论，例如，考虑当企业拥有数据产权时的情形、考虑消费水平对数据的使用效率有增强效果的情形，以及相关的政策含义。最后，我们还初步展示了一个重要的扩展模型：考虑由生产者直接产生的数据作为另一数据来源。这一新类型的数据不再受到隐私问题的制约，从而能够推动经济以新的模式增长。 除了对BGP情形的研究外，本文还采用数值分析的方法研究基准模型的社会计划者问题条件下在到达BGP之前的过渡态变化趋势，并得到结论：随着数据经济走向成熟，经济中数据的使用量会由快速增加至某个峰值之后逐渐减少，隐私保护问题也将由不断加重逐渐变得不再重要。当然，在当前数据经济发展的初期阶段，我们仍然需要平衡好当下的隐私保护问题与长期经济增长之间的关系。在本文的末尾，我们还通过整理和比较欧洲、美国和中国在数据隐私保护和数据经济发展方面相关的证据，来对我们得到的结论进行初步的验证。
This dissertation introduces the consumer-generated data as a key factor into the process of knowledge accumulation, and proposes a novel endogenous growth model based on this, which is one of the fundamental studies of the field on the long-run effects of data economy. In this model, consumers' utilities are determined jointly by the consumption of final good and the provision of their data. They are employed either in production sector or in R&D sector, and need to balance the revenue from providing personal data and the potential risks of being invaded by privacy linkage due to the usage of data. The final good producer produces in a perfect competitive environment employing labor and intermediate goods. Intermediate good producers use labor and data for innovation activities before entering the market. They can get the patent once they succeed in inventing, and can produce in a monopolistic environment. Our results show that: (1) The decentralized economy can achieve the same level of Balanced Growth Path (BGP) growth rate as the case in the social planner's problem. Moreover, while the BGP growth rates of consumption, technology level are positive, the BGP growth rate of data provision is negative. (2) Although the BGP growth rate in the two cases are the same, there still exist the problem of labor underemployment and data overuse in R&D sector in decentralized economy compared with social optimal allocation. In a further extension, this dissertation considers the influence of dynamic data non-rivalry to the results. Under the condition that we allow the trading of data between producers entering in different periods and take the risk of creative destruction into consideration, we come up with new conclusions on the intertemporal data trading behavior: Unless the probability of creative destruction increases to an extreme level, in the decentralized economy, incumbents are always willing to sell more data to the entrants than the social optimal level. Besides, this dissertation also makes some further discussions in various aspects, for example, considering the case when firms have the property of data, considering the case when consumption level can augment the efficiency of using data, and some relating policy implications. At last, we initially presents an important extension model: considering the case when data are produced directly from producers act as another source. This new type of data are no longer restricted by the privacy problems, and thus can lead to a new mode of economic growth. Besides the studies on BGP, this dissertation also applies numerical analysis to find out the transitional dynamics before reaching BGP, based on the case of social planner's problem in the baseline model, and come to the following conlusion: As the data economy goes mature, data in the economy change from increasing highly to a certain peak, then decrease gradually, and thus the privacy protection problem will change from increasingly serious to trivial. Indeed, as the data economy is still at the initial stage, we still need to balance the relationship between current privacy protection and long-run economic growth. In the last part of this dissertation, we collect and compare the evidences on privacy protection and data economy development from the Europe, the U.S., and China, in order to support our conclusions.
