The source I am using is The Fall of Rome and the End of Civilization by Bryan Ward-Perkins. link

The Roman Empire was a common market for goods, characterized by relative peace and stability for several centuries. Producers were able to ramp up production, as they had a massive empire-wide market to sell to.

Ward-Perkins, an archaeologist, focuses on pottery. Roman pottery, according to him, is extremely plentiful and of consistently high quality. The volume of Roman pottery is far higher than that of the pre-Roman period and the post-Roman period. Testaccio, a neighborhood in Rome, is on a hill formed by 53 million broken amphorae, a staggering number. Roman pottery is of good quality, being lightweight, smooth, and remarkably sturdy. The consistency of the pottery is evidence that it was mass-produced by workshops at various sites throughout the empire. Pottery can be traced to specific workshops, which shows an empire-wide distribution network. Pottery from a workshop in southern Gaul has been found in Britain, Spain, Africa, Italy, and along the Rhine-Danube frontier.

The mass-production and distribution of pottery indicates that people were purchasing all this pottery. Ward-Perkins gives the example of a small Roman farm in Luna, a site in Italy, that contains remains of large storage jars, kitchenware, a set of tableware, and remains of pots imported from Southern Italy and Africa. Ward-Perkins argues that the presence of large amounts of pottery, a representative consumer good, in simple peasant farmhouses, shows that the Roman economy generally provided affordable, good quality goods not just to the elite, but to a large segment of the population.

This changes with the fall of the Roman Empire. In Britain, for example, post-Roman pottery is locally made and of very low quality. Roman pottery was wheel-formed and fired in massive kilns, while post-Roman British pottery is hand-formed and fired at relatively low temperatures. Compared to Roman pottery (the author provides sketches) post-Roman British pottery looks rough and misshapen. The quantity sharply drops off, as well. Britain is particularly remarkable, as pottery production in post-Roman Britain is even lower than in the pre-Roman Iron Age. In Italy, Roman-style pottery is still present, but at a much lower quantity and restricted to richer households. The large distribution networks of the Empire collapse, and most pottery in post-Western Roman Italy is local, but there are some imports from southern Italy to northern Italy.

Ward-Perkins does not only examine pottery; he also looks at coin distribution, roof tiles, the prevalence of writing, and the size and complexity of building construction. His conclusion is that the Roman Empire had farms and workshops which produced goods on a large, though still pre-industrial, scale, and distributed their goods throughout the empire. This was due to the economic specialization that centuries of peace and an empire-wide market provided. When the empire collapsed, the distribution networks collapsed, so production had to be more local. Local production for local consumption is less efficient than centralized production for wide distribution. Furthermore, due to the centuries of economic specialization, many areas did not have people with the expertise to make certain items. Each village had to learn how to be self-sufficient again. The two problems of lower efficiency (due to the necessity of local production) and lower item quality (due to the loss of technical knowledge) caused an economic crash that coincided with the fall of the Roman Empire in the west.