Been looking over Antony Green’s recent report (pdf) on the 2011 NSW election and inspired by comments from Penny Sharpe thought it was interesting to consider what it reveals about the disappearance of Labor’s heartlands and how 2011 compared to Labor’s most severe previous defeat in 1932. For 2011 Green provides lists of the best and worse polling booths for registered parties. There were six (!) booths where Labor polled over 60% of the first preference vote. Green doesn’t say how many polling booths there were across the state but a rough estimate suggests well over 2000. Three of the 60%+ Labor booths are rural indigenous communities (two in the ultra-safe National Party electorate of Barwon) the others are areas of extreme poverty and social exclusion. There are almost no Labor heartlands left where the party can rely on huge local majorities and were Labor is the hegemonic local force, even when the party does very poorly elsewhere. Some research I have done on the 1890s illustrates this: Labor had a low statewide vote during its first decade, there was no inevitable march to power, but the party did develop core regional strongholds from it advanced outward after 1901.It is true that overall Labor in 2011 polled very poorly with a statewide vote of only 25.6%. However at occasions in the past Labor polled poorly but retained its heartlands. NSW politics in 1931-32 provides an example. At the 1931 Senate election there were two competing Labor parties in NSW, the largest was the Lang-controlled state branch which received 30.8%. At the 1932 state election the state ALP polled 40.2%. Both of these elections were severe Labor defeats, vast numbers of working-class voters did not vote for Lang’s party. In 1932 Lang Labor lost electorates such as Granville which after it was regained in 1938 remained a Labor seat until 2011. In 1932 Labor even lost Canterbury. There were very large swings against Labor. However despite this Labor retained significant regional heartlands where it was the hegemonic force. Polling booth returns are not extant for the 1931 Senate election and I do not have a database for the 1932 election by booth, but what I do are votes for the 418 subdivisions that made up state and federal electorates. In 1931 Lang Labor polled over 60% in 16 subdivisions all in the inner city, inner west and the Hunter coalfields. In 1932 it polled over 60% in 43 subdivisions. The number of polling booths in which Labor would have secured these levels of support would have been much higher. If in 2011 votes had been counted by subdivisions similar in size to those of 1931-32 there would have been none in which Labor polled over 60%. In the 1930s Labor had significant heartlands in 2011 it has none (even social disadvantage today is much more dependent on geographical location than it was previously as Bob Gregory from ANU has shown). The Hunter coalfields and inner city provided a Labor community in the 1930s now we have only a few pockets of extreme social exclusion where I would guess political activism is non-existent: how many ALP members in Wallaga Lake Koori Village (82% Labor in 2011) or Warilla North (64.5% Labor in 2011, some local background here)? In 1932 there would have been many ALP members in Waterloo (77.1% Labor in 1932).