To test the supposition that the lost votes recovered by California through the modernization of voting technologies may be being undone by the trend toward more voting by mail, Table A1 attempts to quantify changes in the residual vote rate in each county from the presidential elections of 1992 to 2008 due to changes in voting technologies, and due to the growth in voting by mail. The table starts by reporting turnout in the 1992 and 2008 presidential elections; the percentage of ballots cast by mail in 1992 and 2008, along with the change across the two years; the type of voting equipment used in 1992 and 2008; and the estimated change in the residual vote rate from 1992 and 2008 based on changes in voting technology, using the coefficients from Table 4 in the text. It then calculates the estimated number of votes represented by this coefficient by multiplying it by turnout for 2008. Negative values indicate an estimated reduction in the residual vote in 2008 due to technology changes since 1992. For instance, for Alameda County, we estimate that the change from punch cards in 1992 to precinct-count optical scanning in 2008 resulted in a reduction in the number of residual votes in 2008 by 5,343 (≈ - 0.85% × 628,545).