Mr Clarke has to attract support from ordinary members
For all the acreage of print that has appeared about how individual members intend to vote in the ballot for leader of the Conservative Party, the simple truth is that nobody knows.
We have had surveys telling us the candidate preferences of the chairs of Conservative constituency associations, as well as the reported views of various groups of party activists.
However, no comprehensive national survey has been conducted among the estimated 300,000 party members to whom ballot papers will be distributed.
One further reason for caution is that ballots of party members have only existed since 1997 and there have been just four prior to the current exercise.
Hague supported
In September 1997, around 40% of members voted in the ballot to endorse William Hague's prior election (by Conservative MPs) as leader of the party: 81% supported him.
In March 1998, about one-third of members participated in the ballot to endorse William Hague's proposals to change the party's constitution: 96% voted in favour.
In September 1998, there was a 58.9% turnout in the ballot concerning William Hague's policy on the euro: 84.4% endorsed his view.
In September 2000, 16.7% of party members participated in the ballot to endorse the policy document "Believing in Britain": 99% approved it.
The only time members have participated in a candidate election was when those living in London cast their votes for the Conservative party's mayoral candidate in 1999.
On that occasion, 58.6% of the London membership voted and 70.5% of them supported Jeffrey Archer. This suggests that turnout for the present contest will be closer to 60% than 17%.
Lacking any reliable data on the candidate preferences of party members, the national opinion polls have, inevitably, concentrated on the views of respondents who say they will vote Conservative as well as those of all electors.
The July ICM/ Guardian poll was conducted before Michael Portillo's defeat in the final ballot of MPs and it found that among all respondents 14% said that an Iain Duncan Smith victory would make them more likely to vote Conservative, compared with 22% who said it would make them less likely.
Crucial questions
The comparable figures for Kenneth Clarke were 25% and 21% respectively.
The July MORI/Times poll was sampled after the final ballot of MPs and it tested among Conservative supporters three crucial questions facing Conservative party members as they vote this summer.
Conservative supporters were asked which of the two candidates would do the best job "as leader", at "increasing Conservative support across the country" and at "uniting the party".
On leadership, Clarke took a 15% lead over Duncan Smith; and this lead increased to 26% when it came to increasing support for the party across the country.
However, on the issue of uniting the party, Clarke's lead over Duncan Smith was a mere 3%.
Challenge for members
So, it is left to party members to square the circle.
Their overwhelming response in the euro ballot of September 1998 shows them opposed to Kenneth Clarke's stance and yet their own supporters in the country, as well as others they need to attract, express a clear preference for him as party leader.
My belief is that party members are much closer to the views of the general public than party activists: that is certainly Labour's general experience in membership ballots.
If that assumption is correct then clearly Mr Clarke will have the edge in the current ballot.
But if he does win the leadership then many party members will have swallowed some strongly-held political views before they voted for him.