In most all cases, benefits and costs used to evaluate the efficiency of a policy are measured in dollar terms. Dollar prices are often used in this estimation because they reveal how people value scarce resources. This is what may be referred to as ‘use value.’ Economists also recognize ‘non-use value,’ or the value or benefit that people conceive from simply knowing that resources exist. ( ex: you may get non-use value from simply knowing that bald eagles have not become extinct etc. regardless if you ever benefit from their existence in some direct financial way.) Of course, these values are much more difficult to quantify.
Without market prices, economists often rely on survey data or data provided by the physical and natural sciences. Both revealed ‘use-value’ and approximated ‘non-use’ value are converted into monetary terms to provide a common unit of measure. This is not to say that we can put a price on everything, and this does not give greater weight to ‘markets vs. nature,’ it just provides a common unit of measure. This is no different than converting from meters to feet etc.
In the end, if total benefits translated into monetary terms exceed costs, ( the policy produces positive net benefits) then the policy is said to have a favorable cost/benefit analysis, at least from a pareto-potential perspective.
This does not mean that the policy should be undertaken, just that it may be favored over an alternative with negative net benefits. There are plenty of other criteria that must be considered such as constitutionality, distribution, rent seeking, etc. ( see ‘Public Choice’ under selected topics)