In 1973, American League and its owners implemented the designated hitter rule.  This would give a team in the American League the ability to have a non-position player hit for the pitcher when it was their turn to hit in the lineup  This paper looks at the offensive statistics from each league before and after the designated hitter rule took effect. With the exception of home runs, both the American League and National League improved significantly in offensive statistics but no differences were found between the two leagues.

Since 1973, the American League and National League have been separated by the designated hitter. Used only in the American League, this player bats for the pitcher when it is their turn to hit in the lineup. The main purpose of the rule was to boost offense in the American League and thus make the game more interesting for fans. While there has been much debate about whether the designated hitter should be used in baseball, the rule remains after 27 years with no major push by the owners to get rid of the designated hitter.

The average baseball fan can tell you that the weakest hitter on any National League team is the pitcher. They are not relied upon to hit home runs, drive in runners in scoring position or in general, to help the team win a game with their offense. The American League has the advantage of getting to replace the pitcher with a more productive batter who can be expected to do the things that help a team win.

The objective of this paper is to find out if the American League improved offensive production with the use of the designated hitter. Also of interest is to see if the American League was significantly more productive than the National League.

The years between 1968 and 1977 were split up into periods of five years. The 'Before' period is from 1968 to 1972. The 'After' period ranged from 1973 (the year the designated hitter started to be used) to 1977.

All teams in both leagues were present in the statistics. Because of expansion (in 1969 and 1977) there was not the same number of teams throughout the ten-year period. In 1968, ten teams were in each league. From 1969 to 1976 there were a total of 12 teams in each league. In 1977, only the American League expanded, raising its number of teams to 14. This did not turn out to be a problem since the statistics were taken in terms of the averages per team within each league and not by league totals each year. Expansion years have been shown to cause a dramatic increase in offensive production. Fortunately, their was expansion in each of the before and after periods. Otherwise it might have been worthwhile to drop the expansions years from our model.

Four offensive statistics were used as the determinants. Batting average, home runs and runs batted in (RBI) are popular offensive statistics used to determine a players' and team's value. Another statistic used, slugging percentage, is used by Major League Baseball but not as widely known as the first three. Slugging percentage is calculated by taking the total number of bases divided by the number of at-bats a player has. Home runs are given four total bases, a triple given three total bases and so on. Slugging percentage is also referred to as a power statistic since it accounts for the extra base hits, unlike batting average where a single and a home run are calculated the same way in the average.

The analysis was done in SAS using a 'Mixed Methods' approach for a general linear model. The three fixed factors were league (American League and National League), the before and after five year periods, and then the league by those before and after periods. The random factors were the teams within the league and then the teams within the league by the before and after period.

Each of the four offensive statistics will be looked at separately.

Looking below at Figure 1 we can see how the American League's average increased from a 0.242 to a 0.259 after the designated hitter. A rise of .017 may not seem much to someone that has never looked at baseball statistics. That increase actually works out to be about 100 more hits per team, per year during the after period. The National League also sees an increase from the before and after periods, going from a 0.250 to a 0.257 average.

Below in Table 1 are the differences in least square means. P-values are also given. As shown by Figure 1, the largest significant difference is between the American League's before and after averages. A p-value of less than 0.001 further shows the significant difference between the before and after periods. The National League also has a significant difference in the before and after periods. Other than the before period, batting average does not differ significantly between leagues. Throughout the ten-year period there was only a .002 difference and a .003 difference during the after period.

Figure 2 shows how home runs actually decreased for both league during the after period. The American League went from about each team averaging 123 home runs down to 121. The National League also decreased slightly, going from 117 to 113.

Table 2 gives the 95% confidence intervals for each league during the before and after periods. Each lower bound crosses with the other upper bounds, suggesting further that there are no significant differences between the leagues and before/after periods.

Figure 3 shows how RBI behaves much in the same way that batting average did for the before and after periods. The National League has a higher team average in the before period but the American League overtakes the National League during the after period. The increase for the American League is about 71 RBI, which breaks down to about a half RBI more per game during the after period.

Table 3 shows the differences in least square means for RBI. We see significant differences between the before and after periods for both leagues. The American League's difference though was almost twice as much as the National League's difference between the two periods. Comparisons between leagues though show there were no significant differences in either the before or after periods.

The interaction plot in Figure 4 shows the same crossover effect that the interaction plots showed for batting average and RBI. We would expect this to be similar to batting average since the statistics are calculated in a similar fashion.

The differences for least squared means and the p-values are shown in table 4 for slugging percentage. Again, there are no league effects for the before and after periods. Unlike batting average and RBI, slugging percentage did not show a significant difference for the National League in the before and after periods. The American League is again proven to have a significant difference with a .021 increase and a p-value of less than .001.

Although we treat the years within the before and after periods as random, slugging percentage showed how trends from year to year are consistent in both leagues. Figure 5 shows slugging percentage year by year. If the American League's slugging percentage increased from one year to the next then so did the National League's. Similarly, if one league decreased then so did the other.

With the exception of home runs, the American League improved significantly in offensive statistics after the designated hitter rule took effect. In home runs, the average per team actually decreased a small amount per team when the designated hitter was used.

The National League, without the designated hitter, also showed a significant increase in batting average and runs batted in during the after period. This increase however, was not as substantial as the American League's increase. Slugging percentage increased, but not significantly. Home runs also decreased by a low number during the after period for the National League.

It was shown that the there were no significant differences between the American and National Leagues in either the before or after period for any the four offensive statistics. A crossover effect though was shown for batting average, RBI and slugging percentage. The National League had higher averages during the five years before the designated hitter. During the after period the American League overtook the National League and had higher averages in each of these three offensive statistics. Trends for slugging percentage from year to year showed how the American League and National League were consistent in increasing or decreasing from one year to the next. Although not shown in the analysis, these trends were also in the other offensive statistics. This shows that even though the leagues are independent, trends in offensive production from one year to the next is still similar for both leagues.

It would be easy to conclude that the designated hitter is responsible for the American League significantly increasing its offensive statistics. The designated hitter is a major factor but a more in-depth study is needed to look closer at the designated hitter's statistics. I would recommend a study that looks at the American League's designated hitter's statistics compared to the National League's pitcher's statistics to see how much more production the designated hitter had over the pitcher. Also of interest would be to see how much the designated hitter contributed to the American League increasing its offensive statistics.