1. The government can sell off an asset and have nothing to do with the activities involved with managing that asset. The British government once owned British Airways, but they sold it off and it became completely private. There are no further ties with the government that differ from any other private business.
2. Government-income-generating privatization: The government can rent out the asset. For example, the government can own a tollway and hire a private company to operate it on behalf of the government.
3. Government-expenditure-generating privatization: The government can sell off an asset, but remain as the sole funder for the activity without any private-spending. The private prison system is an example. Military contractors are often in this category.
Yglesias notes that only the first category is really a true privatization. The other categories are hardly reducing the government involvement in the economy at all. But there is a big political difference:
...nothing has been privatized in "prison privatization." The government's just handed out a contract. What you do with the contracting is that instead of handing money over to unionized public sector workers who hand some of the money back to Democratic Party politicians, you hand the money over to a contracting firm that hands some of the money back to Republican Party politicians. From the perspective of partisan politics, these are very different scenarios. From the perspective of prison management, both contract prison operators and prison guard unions lobby for mass incarceration policies. Unions—as membership organizations of middle class people—will also do some lobbying for Social Security while contract prison operators, run by rich businessmen, will also do some lobbying for income tax cuts. But there's no private marketplace in prison management and there never will be.Private prison management is a more concentrated interest group than prison worker unions and that should make them a more effective lobby for additional government spending on incarceration. It is also harder to track the activities of more concentrated interest groups because there are fewer people to potentially leak information.
In true privatization (#1), the government stops making key decisions about how the new business is run. In the other two forms of partial 'privatization', the government continues to make key decisions because the activity continues to be a source of revenue or expenditure just as before. In theory, partial 'privatization' could increase efficiency, but this theory is based upon the presence of market competition and there is rarely significant competition in the primary activities of government like jails and when cities privatize their parking meter operations, they usually sell their contract to a monopoly provider.
One of the problems with government provision of limited resources like parking is that the government often charges too little for them which creates shortages. However, privatized parking meter companies will charge too much for them which also creates inefficiency. For example, when street parking is mostly empty on a Sunday afternoon, it is inefficient to charge anything for parking because that will cause some people to forgo parking and waste that resource. In the case of parking in Chicago, privatized parking probably has a much smaller inefficiency loss than government parking, but this isn't always clear and the Chicago city government still calls the shots and it did not set efficient parking prices.
All three forms of privatization make politicians look good now because they are all ways of boosting current government revenues at the cost of future revenues without raising present taxes. The revenue from the first category is obvious: it is a form of spent savings that will be unavailable to future citizens. The other two categories are less obvious, but both involve long-term contracts and in the case of income-generating 'privitizations', the private company can pay for the lease up front as Morgan Stanley investors did to buy the 75-year lease of the parking meters in Chicago or plans to 'privatize' the Ohio tollway. But then the government loses all future revenue from the meters in exchange for the up-front payment.
In the case of jail privatization, the government must give the private jail operator an incentive to make the fixed investment in the jail by offering a profitable long-run contract. That incentive allows the government can either sell a jail to the private jail corporation or save money by avoiding present expenditures by getting the private company to build a new jail rather than the government. The corporation gets a long-term stream of profits baked into its contract in exchange for the initial outlay for the prison.
Every kind of privatization is a tempting accounting gimmick that temporarily boosts the government budget balance at the expense of the future.
