IN CHAPTER 5, it was noted that a method frequently used to illustrate the effects of the poll tax upon voting is to compare voting in poll tax states with voting in non-poll tax states. The inadequacy of this method when applied to poll tax states as a group in comparison with all non-poll tax states is obvious. A refinement of this procedure is to compare voting in two states that are similar in most respects, except that poll tax payment is required in one and not in the other. Sole reliance on this manner of ascertaining the effects of the tax upon voting is open to serious question. No two states are alike in every respect apart from the fact that one has a poll tax and the other does not. If such a condition existed, the effects of the tax could be tested with scientific precision. Since many differences exist between states and since many factors influence voting in them, the results of interstate comparisons need to be taken with several grains of salt. This warning must be remembered in considering the material which follows.
Table 6 contrasts voting participation rates in all southern states, poll tax and non-poll tax, for the Democratic gubernatorial primaries in 1950 or in the year nearest to 1950. The average participation rate for southern non-poll tax states was just over four percentage points more than that for southern poll tax states. If the turnout in the North Carolina general election had been used, the differential between these two groups of states would have been about three percentage points higher. Due to Republican strength, more citizens vote in the