There's an assumption that expansion teams dramatically help league-wide attendance in their first year. While it's true that these team usually draw above the league average at home, their overall impact is actually a lot less than I expected. The reason is that these teams don't draw especially well on the road.
In 2007 Toronto came into the league, and average attendance was just 0.75% higher including their games.
total att. games mean
2007 all 3,270,210 195 16,770
TOR home 301,947 15 20,130
TOR away 221,650 15 14,777
totals w/o TOR 2,746,613 165 16,646
Chivas USA and Real Salt Lake joined in 2005, and average attendance was 3.82% higher factoring in their games. This was helped in part by the Superclassico which drove Chivas' "road" attendance significantly above the league average.
total att. games mean
2005 all 2,900,716 192 15,108
CHV home 273,284 16 17,080
CHV away 268,176 16 16,761
RSL home 288,586 16 18,037
RSL away 207,985 16 12,999
adj. totals 1,862,685 128 14,552
In 1998, the Fire and Miami Fusion joined the league. Average attendance was just 0.48% higher when their games are included. Oddly enough, the Fusion's road numbers were actually higher than the league average, but their home attendance more than offset this. If you factor out the Fusion, and just include the Fire's numbers, the attendance was 2.18% better off with them in the league.
total att. games mean
1998 all 2,747,897 192 14,312
CHI home 286,190 16 17,887
CHI away 220,734 16 13,796
MIA home 164,548 16 10,284
MIA away 253,269 16 15,829
adj. totals 1,823,156 128 14,243
This brings us to to this year. Seattle's phenomenal home support really has been a boon to the league average. Their 30,000+ average thus far has actually led to an 8.18% increase in average attendance even though their average road attendance is roughly 1,200 fans below the league average.
total att. games mean
2009 all 2,755,194 175 15,744
sea home 397,628 13 30,587
sea away 159,951 11 14,541
totals w/o SEA 2,197,615 151 14,554