Desktop version

Home arrow Sociology arrow The future of mobility

Creating the Scenarios

Our methodological approach represents a state-of-the-art scenario process while recognizing that scenarios can be developed using several different approaches. Our approach combined expert opinion, gathered via in-person workshops, with cross-impact analysis, consistency analysis, and cluster analysis using specialized computer tool support. Even though it relies more on substantive expertise than on formal research and modeling, the approach was highly empirical. Because some of the terminology might be unfamiliar, we provide definitions in Table 1.1.

Table 1.1. Key Definitions




Influencing area

A broad topic area that is thought to affect mobility

This study uses four:

• demographics

• economics


transportation supply and constraints


A metric that represents one specific element within the influencing area

Demographics contains five descriptors:

total population

geographic distribution of population


commute distance

household type


A prediction of the future value of a descriptor

For total population, there are two projections: ^ 1.39 billion ^ 1.44 billion

Select Influencing Areas

The scenario process begins by defining three key study parameters: (1) topic (the future of mobility), (2) geographic scope (China), and (3) time horizon (2030). The research team identified four influencing areas and specific descriptors to fit the study parameters. An influencing area is a broad topic area that is thought to affect mobility. The four in this study are demographics, economics, energy, and transportation supply and constraints.1 These were selected based on the German and U.S. work, as well as additional background research on China. For each area, we then identified descriptors, which are metrics that represent one specific element within the influencing area. For example, among the descriptors identified for energy were the price of a barrel of oil and the percentage of vehicles in 2030 that would be hybrid or electric.

Although we developed a long list of potential descriptors, we narrowed these down based on two criteria: uncertainty and impact. In a scenario analysis, because the differentiation between certain and uncertain descriptors is based on a range of predictions, including uncertain descriptors is more important than including those that are more certain. Impact is important because we want to use factors that are more active than passive. An active descriptor influences many other descriptors; many other descriptors influence a passive descriptor. Appendix A provides a full list of the 24 descriptors. The research team produced a paper on past and current trends for all descriptors in each influencing area; Chapter Two summarizes these papers.

Elicit Projections on Descriptors

The research team held one workshop for each influencing area to gather expert opinion on projections for each descriptor. Two workshops (those on demographics and on economics) were held in the United States with American experts on China, and two workshops (those on energy and on transportation supply and constraints) were held in Beijing with Chinese experts. We defined a projection as an estimate of future possibilities informed by past and current trends. We invited six to eight prominent outside experts (RAND and ifmo staff identified American experts, whom we list in Appendix B; our partner, Tsinghua University, identified Chinese experts2) to attend each workshop. Prior to each workshop, each expert received the paper on trends in his or her influencing area.

At each workshop, using facilitated discussion, we asked experts for projections for each descriptor in 2030, clarifying that we were not asking them to extrapolate from past trends but rather to consider a variety of factors that they thought might influence the descriptor. For each descriptor, the experts provided between one and three qualitative or quantitative projections. For example, we asked the energy experts to project growth in electric two-wheeled vehicles (E-2Ws), which led to estimates of high, medium, and low levels of adoption, each of which was considered plausible depending on circumstances. The number of projections depended on the degree of consensus on likely futures among the experts. We also asked the experts to provide their reasoning (or assumptions) for each projection. For example, those who projected a relatively low level of adoption cited market saturation and a continued decline in the price of other competing vehicle types, while those who projected higher adoption levels noted the low price and convenience.

Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >

Related topics