The ratchet theory of government growth, also known as the crisis theory, is an idea that tries to explain why governments have tended to get bigger during the twentieth century. The model has been postulated by many different observors with varying levels of sophistication, and with the greatest degree that I have come across by Robert Higgs in his book Crisis and Leviathan. The explanation I lay out below owes to his book.
Governments during the twentieth century have tended to grow during periods of crises, especially war and ecomonic depression. This has been the case especially in the United Kingdom and United States. This is the case in terms of secular growth in the percentage of GNP spent by the government, the quantity of government employees relative to the rest of the workforce, and the proliferation of government departments. But it is also true in a more fundamental and important way.
During times of war and crisis, the government tends to expand the sphere in which it exercises economic authority. This is crucial. Imagine the government has just one role, that of providing national defence. It can do this with one soldier and a rifle it bought him for $30 in taxes, or it can do it with a vast military-industrial complex with $370.7 billion. This either entails higher taxes or large inflationary borrowing, but the government is still only exercising authority within a limited sphere. A government could theoretically fight a war without increasing the sphere within it exercised economic coercion, but this is not what tends to happen. Experience shows that what in fact happens is that it tends to dive in and begin regulating huge amounts of economic activity.
Many analysts uncritically accept what is undoubtedly a truism, namely that total war requires the total mobilisation of an economy for war. Few seem to have given thought to why this occurs in the way it does. For instance, the government could tax the economy to its maximum (i.e. to the point where people have enough left for sustenance, but not much else) and use this money to purchase the munitions and manpower it needs. However, this would impose harsh strictures on the civilian population. Although the fact the government is considering drastic action at all means there is a public demand for action, this demand could evaporate and be replaced by opposition if the cost is too high. Even countries facing destruction do not always rally behind their governments - the Bolshevik call for complete surrender in early 1917 being an example.
The government thus has to hide the exact cost and nature of what it is doing from the public, making them not feel the coercive power of the state so directly in their lives. The easiest way is to move into the economy directly, either by direct control or huge volumes of regulation which amount to direct control. This is what the National Defence Act did in the United States just before World War I. It authorised the President to demand goods from a factory at a price determined 'fair' by the Secretary of War, and if the owner refused the President could commandeer the plant and manufacture what he wanted therein. Both paths amount to the same thing, as the right to property includes the right to dispose of this property as you choose and both result in this right being withdrawn.
This approach works because it looks a lot better than liquidation and appropriation. If the government takes control of a factory or regulates its activity to the point of control, then it will do so in a way that means this factory provides it with goods below the cost the market would provide. The loss does not appear as a direct cost, but as an opportunity cost. An opportunity cost is the burden you bear of not disposing of your property in a different way to the one you did. If I buy a copy of Mein Kampf for $10, my opportunity cost is the copy of The Republic I could have got for the same cash. If the government compels me to sell M1A1 Abrams tanks to them for $3,000,000, my opportunity cost is the $1,300,000 extra I could have made if I sold it on the open market. This pisses me off, but it's invisible to the average citizen.
Hence, the government expands the sphere of its economic activity by regulating or controlling numerous industries and removing them from the market economy, which would guarantee their efficiency and operation at normal prices. This is why the U.S. government took control of merchant shipping just before and during the Great War - a shortage of shipping was pushing prices high, so it stepped right in and took prices away from the influence of the market. This is why government tends to step into wholly new areas of economic activity during wars and crises. It has to respond to the demand of citizens to 'do something', but in a way that is not too onerous for citizens. It banishes the market, but with it banishes all the benefits of the market.
The other part of the model that requires explanation is why governments do not retreat from the spheres of activity they have intruded into once the crisis is over. One explanation is the ability of bureaucracy to perpetuate itself once it is created and has established a constituency that benefits from its existence. Cries of "All the experts agree we are necessary!" or "But the people need us!" can discourage the faint-hearted politician from assaulting their privilege; after all, it is easier to rally political support for constructive action than for destruction.
But most crucially, the experience of a mixed economy, or of the government controlling X, Y or Z and seemingly doing an okay job, changes the people who live through it. The conservative domino theory that any assault on property will result in the obliteration of civilization seems to be discredited by events to people who live through it. As the government spends more and more time engaged in a comamnd economy, it learns how to defray the negative parts to a greater degree and becomes generally better at running the economy itself. The fact that it is impossible for a planned economy to be as efficient as a free one does not matter politically at the time; as the government learns how to deal with the situation, it makes a command economy seem like a better prospect in the future.
Added to this there must of course be those who directly gain from the command economy and hence have an interest in its perpetuation. In this way, any social or economic experiment embarked upon can never entirely be reversed unless the interests it creates are entirely liquidated. Lastly must be considered the effect of government propaganda which is pumped out to justify the changes - for it invariably convinces some, even if just the extremely patriotic.
Hence, once the crisis is over, there will be a period of retrenchment, but government will not shrink to the level it was at before the crisis. The need to 'do something' and the government following what is apparently to it the most rational course of action in the situation creates new attitudes, interests and expectations that must be fulfilled once the crisis is over. The ratchet moves up a notch.