The real world is a far less tidy place than described in the welfare optimising model. An alternative to this messy world eschews its main features and embodies those that define the alternative model.
Firstly, the alternative model does not use the social welfare function as the core concept. It abandons the idea that governments have clearly defined and ordered objectives. Instead it strives to accommodate the view of government actions as being only partly controlled, the outcome of a bargaining process among conflicting interest groups in society.
Secondly, in making policies governments are required to admit in this model the possibilities of large uncertainties ('known uncertain' in modern parlance) about problems. Similar scepticism and doubt prevail over expected policy outcome.
Thirdly, government recognise under this model that institutions in polity have limited capacities for considering and choosing among a large number of alternatives. This is aggravated by the fact that informational and other analyses entail recurrent costs.
As a consequence of these factors the alternative model requires that governments should incorporate flexibility providing room for manoeuvre. In other words, a margin of error should be built-in in policy making.
The welfare maximising model is developed by borrowing from the theory of consumer behaviour. The alternative model, on the other hand, borrows from the theory of the firm. This theory meets most of the conditions mentioned above. The alternative model may be called the satisficing model of decision making or policy making. It assumes that governments don't search for optimum outcome in policy making but settles for solutions that are good enough to command the minimum support of groups, organisations and individuals behind the policy without provoking violent resistance from any of them.
The alternative model can be illustrated in a manner similar to the one used in social welfare optimising model although the model can be applied to many situations. In this context, examples of choice between public and private consumption goods and services can be mentioned. But in contrast to the first model, no use is made of social welfare function. It is simply assumed that in its attempt to strike a compromise among the preferences of competing groups the government lays down minimum levels of public and private consumption. Any policy decision that meets these minimum requirements is acceptable as being satisfactory.
Following the above paradigm the government may specify that consumption of publicly provided goods and services should not be less than the bare minimum. If less than the bare minimum is provided under a policy it will result in drastic reductions in social and economic services (public goods) provoking dissatisfactions from the current beneficiaries. The government may similarly, specify a minimum level of private consumption because less than that may affect peoples' ability to satisfy their basic needs. Moreover, this may necessitate a very unpopular level of taxation. The minimum levels of consumption of both public and private goods and services should, therefore, be considered as targets or, negatively speaking, as constraints within which policy makers have to operate. In this case, therefore, any set of policies that brings the economy to a point where both the minimum levels of public and private consumption meet or are above it should be acceptable to a government.
The satisficing model presents the choice of government policies as being the outcome of a set of targets arrived at as a result of conflicting pressures and adjustment in the light of feedback regarding success or failure in meeting the targets. Any set of policies which meet the targets will be acceptable to a government. The search for better policies may even cease once the minimum targets are met.
The relative merits of the welfare optimising model and the satisficing model should now be obvious. The welfare optimising model has been criticised for being unrealistic. By this crucible, the satisficing model scores very high. It assumes far less precision and consistency in government objectives, requiring only that minimum targets be set. It dispenses with the fine calculations of marginal rates of substitution and trade-offs and eliminates the necessity of a controversial core concept like social welfare function. What is more, it can accommodate the views of various interest groups as being based on compromise. It also copes better with the existence of uncertainty and the need for flexibility by allowing a wide range of possible outcomes. By the same token, it also simplifies the decision process by allowing the search for solutions to stop as soon as a satisfactory one is found, instead of continuing until the optimum is reached. The satisficing model is more dynamic in that it has more to say about the policy making process and in explicitly allowing for the adjustment of targets and policies over time. Finally, there are enough evidences that the satisficing model conforms more to how governments actually go about making policy decisions. In contrast, it would be difficult to find a real-world application of the welfare optimising model.
If a descriptive model is wanted the sacrificing one is clearly preferable. However, this model has less to say if the normative question of how governments ought to make policy decisions is asked. If that is the question to be answered then the optimising model should be taken more seriously, because by definition it is a normative model. But at the same time it should be recognised that by incorporating the notion of adjustable target levels the satisficing model does provide a mechanism which ensures that grossly sub-optimal solutions in policy decisions will not persist for long.
