YouGov versus Lib Dem Voice surveys: how do they compare?

As Stephen Tall mentioned earlier, we’ve done a little experiment with the latest Lib Dem Voice survey of party members by asking exactly the same questions as were very recently asked of party members in a YouGov survey. It turns out that the answers from our two different approaches are very similar:

As you may know, there is currently debate about whether or not the UK should replace its Trident nuclear weapons system. Current policy is to replace the Trident submarines with a new fleet of boats, and to replace the ballistic nuclear missiles they carry at a later date. Which of the following options would you favour most?
Replace Trident with a broadly comparable system: 7% YouGov, 3% LDV
Replace Trident with a cheaper system: 32% YouGov, 38% LDV
Not renew Trident and give up nuclear weapons altogether: 57% YouGov, 55% LDV
Don’t know: 4% YouGov, 4% LDV

As you may know, the UK government is planning to build a replacement for the Trident nuclear weapons system. Do you support or oppose plans to replace Trident with a ‘like-for-like’ system?

Strongly support: 4% YouGov, 1% LDV
Tend to support: 8% YouGov, 5% LDV
Neither support nor oppose: 8% YouGov, 4% LDV
Tend to oppose: 28% YouGov, 32% LDV
Strongly oppose: 49% YouGov, 56% LDV
Don’t know: 2% YouGov, 1% LDV

The questions and answers were worded the same, though the rest of both surveys was different and the fieldwork dates were slightly different.

Broadly speaking the figures are reassuring for both LDV and YouGov as we are both finding similar results. The differences on the second question suggest LDV may be polling people who are more opposed to nuclear weapons (potentially leading to speculation about them being more likely to be activists) but the variations on the first question do not support that.

All in all, this suggests that our survey results are generally pretty representative of the membership as a whole, as has also been suggested by the results in internal party elections which we’ve also surveyed.

The important caveat is that our survey results are skewed towards men. This does not appear to affect the results (e.g. surveyed party members of one gender do not give party figures of their own gender better or worse ratings than members of the opposite gender), but it is possible to imagine questions where it might.