STATE TAXATION OF MULTIJURISDICTIONAL CORPORATE INCOME: AN EMPIRICAL ANALYSIS OF FORMULARY APPORTIONMENT

The primary objectives of the study are to present a comprehensive summary of the historical, legal, and political aspects of state corporate taxation and to conduct an empirical analysis of formulary apportionment. The empirical section of the study addresses the issue of apportionment formula fairness. Five research questions are investigated--using a variety of multiple regression and descriptive statistics techniques--by observing the reported values of income, receipts, property, and payroll for 252 large corporations, operating in widely diverse industries, for the period 1977 through 1981. The study reports the following results: (1) The standard UDITPA factors of property, payroll, and receipts, taken as a set, significantly reflect income-generating processes. The explanatory power of each individual factor is highly significant, but the multicollinearity among the factors results in considerable redundancy when the factors are viewed simultaneously. (2) These factors are very stable over time. (3) The relative importance of these factors is disparate. An optimal weighting of the factors, based on the multiple regression results, would assign approximately equal weight to the property and receipts factors and a negative weighting to the payroll factor. (4) The factors, although stable over time, are not stable across industries. The income functions of different industries vary considerably, both in terms of factor dependency and factor weighting. (5) Of the eleven practical alternative apportionment formulas tested, the most appropriate, in terms of the least degree of misapportionment, varied from industry to industry. (6) Three-factor apportionment formulas generally outperformed single-factor and two-factor formulas. (7) Apportionment formulas which double-weighted the receipts factor consistently resulted in a higher degree of misapportionment than those formulas which equally weighted this factor. (8) Formulas which measured the property factor as undepreciated historical cost consistently outperformed those formulas which measured this factor in terms of depreciated value. (9) Formulas which measured the payroll factor in number of employees consistently outperformed alternate formulas which measured this factor in dollars of compensation