This paper has two major parts. First, it reviews U.S. and Japanese broadband universal service policies, with special attention to the process of and discussions on (re)defining scope of the network services to be dealt with in the policy. Second, it explores the shape of universal service policy for Internet of Things. IoT is still in its formative years, but there is a potential for IoT-based devices and services to be popular and important for the social life in the near future. This paper takes an approach to first review how the universal service policies determine target service for the policy. More specifically, recent policy discussions and formal processes in the U.S. and Japan regarding inclusion of broadband network for universal service are reviewed. In the U.S. formal inclusion process of broadband network for universal service has happened in the early 2010's. In Japan, there is an ongoing policy discussion, which may or may not lead to inclusion of broadband in the near future. U.S. values advanced services more in the context of universal service, while Japan has a more welfare-oriented approach, the kind which emphasize guarantee of essential services for individuals. Broadband, when compared to traditional telephone network, supports a widerrange of uses. This is the case almost by definition - telephone is primarily a simultaneous, two-way, two-person, voice communication, whereas broadband simply means a large amount of bandwidth, without specifying the type of communication or services such network supports. Similarly, IoT may be greater in scope of uses it supports. This is again a consequence of the meaning of the term: greater scope of physical objects are (going to be) connected to the Internet. It is not difficult to imagine such services as health and safety monitoring services for the elderly and in the near future to be considered essential for the society. In case of IoT, however, there is still a degree of uncertainty regarding the diversity of network infrastructure. That is, it seems that there is a fair amount of chance that specific network is tied (such as by technological standard and contract terms) to specific set of devices and services. Home monitoring service, e-reader, and other devices and services may have different sets of network requirements to provide optimal services. In this sense, IoT may present a set of relatively new challenges to universal service policy. As we see in some existing products, network cost may be internally subsidized in case of bundled offers, and difficult to single out. We may face such questions as how meaningful is it to subsidize only network-related cost, ignoring service and device charges; how we can define supported services when speed may or may not be all that matters. Helpful lessons could be drawn from handling of difference of mobile and fixed broadband. The two are rather different from fixed broadband in terms of connectable devices and services offered.
