of a policy that provides nonmonetary bene- 

 fits. An example will clarify the distinction. 3 



Let us assume that the United States de- 

 sires to achieve a growing level of personal in- 

 come and general economic prosperity 

 throughout the nation. It also desires to im- 

 prove the lot of the lowest quintile (20%) in 

 the income distribution (i.e., persons with 

 poverty-level incomes). Alternative fiscal poli- 

 cies may be presented, all of which would re- 

 sult in an increase in personal income, but in 

 different magnitudes and distributed differ- 

 ently in a relative sense. The first alternative 

 might achieve an increase in personal income 

 of $1 billion in such a way that it is spread 

 evenly throughout the nation. The second al- 

 ternative might be able to achieve a total in- 

 crease in personal income of only $0.8 billion, 

 but would increase income to the lowest quin- 

 tile by $0.5 billion, with $0.3 billion going to 

 the remaining 80 percent of the distribution. 



'The example in the text is a simplification of the 

 concept. For a more precise but readily understand- 

 able treatment see Major (1969). See also Marglin 

 (1962). 



The third alternative promises to increase per 

 capita income by $1.01 billion, but would re- 

 sult in only $0.1 billion going to the lowest 

 quintile. The required decision is which policy 

 alternative to pursue. 



The values of personal income that are at- 

 tainable do not in themselves indicate the 

 proper avenue to take. If no special "weight" 

 were attached to the goal of redistribution to 

 the lowest quintile in the income distribution 

 (situation A), then the third policy, which 

 yields the greatest total increase, would ap- 

 pear appropriate. If, on the other hand, a pos- 

 itive weight or priority were placed on the 

 redistribution (situation B or C), either the 

 first or the second policy would be preferable, 

 even though both would yield lower total in- 

 creases than the third alternative. The deter- 

 mining factor will be the degree of impor- 

 tance attached to the goal of income redistri- 

 bution relative to the goal of increasing total 

 personal income. The following discussion 

 shows how decisions will be affected by plac- 

 ing alternative weights on the achievable val- 

 ues of the three fiscal policies. 



Goals 



Increase in personal income in 

 the lowest quintile (Pip) 



Increase in personal income, 

 total (PI t ) 



Policy 1 



$ 200,000,000 

 $1,000,000,000 



Policy 2 



Policy 3 



$500,000,000 $ 100,000,000 



$800,000,000 $1,010,000,000 



Various alternative weights or priorities may 

 be placed on the two components of the goal, 

 but whatever the weights, the decision rule is 



Maximize the weighted aggregate increase in 

 personal income. 



This can be expressed quantitatively as 

 Maximize (PI^ + aPI p ) 



where a is the premium placed on PI 



In situation A, with no distinction with 

 respect to income distribution, no special 

 weight or premium is given to income gained 

 in the lowest quintile of income distribution. 



The optimizing behavior is 

 Maximize (PI t + PI p ) 



Policy 1 yields 

 $1,000,000,000 + $ = $1,000,000,000 



Policy 2 yields 

 w $ 800,000,000 + $ 0= $ 800,000,000 

 jfe Policy 3 yields 



T $1,010,000,000 + $ = $1,010,000,000 



By choosing policy 3 we maximize the weight- 

 ed value of increase in personal income. 



In situation B, a special premium or weight 

 of 1.0 is given to income gained in the lowest 

 quintile of the income distribution. In effect, 

 any income going to the poorest segment of 



6 



