In a recent PNAS article, Charles Stanish and Abigail Levine explore this issue through the archaeology of the Lake Titicaca region in South America. As they say, states do not arise in isolation:
Current research point to the interaction of many nonstate polities across a regional landscape over many generations as the context in which first-generation states develop.By 500 BC there were many sizable towns in this region that must have had some level of central authority, but each seems to be independent and there is nothing that looks like a state -- no grand architecture, no one town much larger than the others. Then after 500 BC the town of Pukara rose to be much larger than the others, with monumental architecture, royal tombs, and the like (top picture and below). Recent excavations at the site of Taraco, 50 km (30 miles) away from Pukara, show that the elite residential district of this town was destroyed by fire around 500 BC, and that afterward it became a much smaller settlement. Stanish and Levine think the fire and the town's decline are evidence of war with Pukara, and therefore that war with neighboring towns was a key reason for the rise of the Pukaran state.
This makes sense. One reason people submit to state rule is to gain protection against neighboring states. In an environment of increasing competition, each community will need to become better and better organized to defend itself, and those communities able to build defensive walls, create disciplined armies, and so on will dominate the others. Archaeologists often find signs of intensive warfare in early states, for example in iconography (Mississippian kings are usually depicted with a war club in one hand and a severed head in the other -- below) and in fortifications. These early states seem to be highly unstable, probably because it is difficult in Neolithic or Bronze Age societies to mobilize the resources necessary to support the apparatus of a state.
War has played a major part in the rise of civilization, down to recent times. From the rise of states to the harnessing of atomic power, many of our biggest technical and organizational advances have been made to fight wars. Representative government was created in the late Middle Ages to raise money for fighting wars; the bond market was created to raise money for fighting wars; the modern industrial system with its interchangeable parts was pioneered by gun makers. The realization of this is one reason many leftist anthropologists have turned against the whole notion of civilization.
War is not enough, though. The money has to come from somewhere, and this explains why early states are often associated with major irrigation projects -- taming the Nile and the Euphrates provided the surplus to support early states in Egypt and Sumer -- and with long-distance trade. Long-distance trade can provide the profit needed to fund states, and therefore we often find early states heavily involved in promoting trade, and basing themselves in merchant cities. Even the nomad empires of the steppes, which seem so different from settled states, depended ultimately on their control of trade routes for their power. Where neither of these factors operated (Mississippian North America) states cannot establish themselves, and quickly fade.