When you look at an introductory textbook for microeconomics, it seems as though economists live in a world that barely resembles the real one. Their models are built on a series of assumptions, including that individual actors have perfect information about their choices or that subjective human values can be measured quantitatively. Some models even assume away competition, substitute goods and marketing.
Economists create these elaborate fantasies in order to apply testable theories onto an otherwise deductive social science. Without assumptions, quantitative economic models would not be able to produce any meaningful conclusions.
Science, Positivism and Deduction
In his 1953 essay titled "The Methodology of Positive Economics," Milton Friedman explained why economists need to make assumptions to provide useful predictions. Friedman understood economics couldn't use the scientific method as neatly as chemistry or physics, but he still saw the scientific method as the basis. Friedman stated economists would have to rely on "uncontrolled experience rather than on controlled experiment." He was promoting a method of logical positivism, which is a subset of epistemology.
The scientific method requires isolated variables and testing to prove causality. Economists can't possibly isolate individual variables in the real world, so they assume most of them away to create a model with some constancy. Errors are bound to happen, but most economists are OK with that as long as those errors are small enough.
Not all economists are methodological positivists. Some argue that standard models are too unrealistic to ever be relied on, and applying positivism to social sciences implies human choice is an illusion.
Economist Robert Murphy of the Institute for Energy Research, a not-for-profit organization that conducts research and analysis on the functions, operations, and government regulation of global energy markets, argues economics should be focused on logical deduction, not logical positivism. Logical deduction is the science used in philosophy, mathematics and computer science. Murphy, along with some other Austrian-school economists, contend certain economic tendencies and laws are only knowable through human reasoning. (For related reading, see: Is economics a science?)