A.L.P. Federal Election Results Since 1910

An updated version of the data on this page was published in 2016.

“The ALP is heading for its biggest defeat ever under Gillard,” I was told recently.

The confident assertion promptly fell to pieces when I asked for a definition of “biggest defeat ever”. A garbled account of seats, votes and swings followed. Such are casual political conversations. Few people know the figures.

But the question is a good one. How do you measure the extent of an election defeat? If the Gillard government is annihilated this year, what measures of comparison should we use?

Here’s a table showing ALP statistics for three different measures: the proportion of seats won in the House of Representatives, the two-party-preferred vote, and the primary vote. The ALP’s winning election years are shaded yellow.

The table includes every election since Federation, except for the first three: 1901, 1903 and 1906. These have been excluded since they took place before the formation of the two-party system as we know it. Since 1910, elections have been fought between the ALP and the non-Labor parties under a variety of names.

The ALP has won 14 of the 40 elections held since 1910. I have categorised the 26 elections it has lost into four groups:

1. Seven major defeats where the ALP won no more than a third of the seats in the House: 1917, 1925, 1931, 1966, 1975, 1977 and 1996.
2. Seven significant defeats where the ALP won between 33% and 40% of the seats: 1919, 1922, 1934, 1937, 1949, 1955 and 1958.
3. Nine moderate defeats where the ALP won between 40% and 50% of the seats: 1928, 1951, 1954, 1963, 1969, 1980, 1998, 2001 and 2004.
4. Three near misses where the ALP just fell short: 1913, 1940 and 1961.

By any measure, the ALP’s most successful election was John Curtin’s victory in 1943. Curtin won 66.21% of seats in the House. James Scullin won 61.33% in 1929 and Bob Hawke won 60% in 1983.

Curtin’s victory is also the only election in which the ALP polled in excess of 55% of the national two-party-preferred vote. [Note: Early figures for the two-party vote are not shown either because there are no precise figures available or because the election took place before preferential voting was introduced in 1918. Up until 1955, two-party figures contain a small element of estimation because some seats returned a member unopposed.]

Curtin’s primary vote of 49.94% in 1943 was also one of the best ever. However, Ben Chifley polled 49.71% in 1946, Gough Whitlam polled 49.59% in 1972, and Bob Hawke polled 49.48% in 1983.

The outstanding performer on primary votes is Andrew Fisher. At his two successful elections in 1910 and 1914, the ALP polled 49.97% and 50.89% of the primary vote.

The only other Labor leader to poll over 50% of the primary vote was Dr. Evatt in 1954. The ALP secured 50.03% but still lost the election. At the same time, Evatt managed to win an estimated 50.70% of the two-party vote.

What this points to is the reality of a party’s vote not being equally distributed amongst all seats. In the real world, electoral support can be concentrated in particular areas – just ask the Nationals.

However, a very real problem has emerged over the years. The ALP could be said to have been robbed of victory in the elections of 1940, 1954, 1961, 1969 and 1998 because in each case it won a majority of the preferred vote but lost the election. Similarly, the Coalition won the two-party vote in 1990 but lost the election. The simple principle of majority rule seems to be under threat. Afterall, that’s 6 elections out of 43 since 1901 where the party with the most votes after preferences has been defeated.

Hence, the two-party-preferred vote is an uncertain indicator of electoral success in a parliamentary system using single-member electorates. The ALP’s string of 9 consecutive defeats between 1949 and 1969, for example, shows some reasonably healthy two-party results, including the three elections where more than 50% of the electorate voted or preferenced Labor. But this was a dark time for the ALP which suffered a debilitating split and confronted Australia’s longest-serving prime minister, Robert Menzies. Statistics don’t always tell the real story.

For similiar reasons, the primary vote isn’t a perfect measure of electoral success. The existence of strong third parties since the 1950s (the DLP, Democrats and Greens) has adversely affected the primary vote of the major parties. Moreover, the use of compulsory preferential voting can mask a primary vote fall for a major party because the vote flows back in preferences from minor parties.

So whilst it’s really a statement of the bleeding obvious, it has to be said that statistically and practically the only definitive measure of electoral success is the number of seats won in the House. Political success is ultimately measured on the floor of the lower house. Yes, the two-party vote is usually a good guide to a party’s standing in the electorate. Yes, the primary vote is a good guide to the health of a party over time. But neither of these measures tells the full story. Another example: Matthew Charlton was wiped out in 1925 whilst still polling a primary vote of 45%.

If the ALP’s best election result in the House was 1943, its worst is 1931. Scullin’s divided government suffered a catastrophic defeat that year. Even including the NSW Lang Labor contingent with the ALP yields only 24% of the seats. Scullin also scored the second lowest primary vote on record with 37.67%, again including the Lang Labor vote with the ALP’s.

The lowest ALP primary vote ever recorded in a federal election since 1910 was Mark Latham’s 37.63% in 2004. Julia Gillard just nudged past him in 2010 with 37.99%.

The ALP’s second worst performance in the House was Whitlam’s in 1975. The ALP won just 36 seats out of 127, or 28.34%. At 30.64%, it wasn’t much better in 1977. These results were in the same ballpark with Scullin in 1931 and Frank Tudor’s 1917 defeat in the aftermath of the ALP split over conscription.

If a major electoral defeat is on the cards for the ALP in 2013, Scullin’s 1931 result is the ultimate benchmark. To come anywhere near those figures would be an electoral catastrophe of massive proportions.

But the 1975 and 1977 Whitlam defeats make for a slightly higher benchmark for a defeat which is still disastrous.

Paul Keating’s 1996 defeat is still close enough in living memory to be used as a benchmark. The ALP won just 33% of the seats and scored the then lowest primary vote since Scullin. In the number of seats won, it is on a par with Calwell’s defeat in the Vietnam election of 1966.

The Gillard government holds 72 seats, already a minority. If it loses 22 seats this year, it matches Keating in 1996 and Calwell in 1966. If it loses 30 seats, it matches Whitlam in 1975. If it loses 36 seats, it matches Scullin in 1931.

This is the above table reordered and sorted from best to worst results in the House of Representatives: