James Frayne is Director of Public First and author of Meet the People, a guide to moving public opinion.

Since the Government machine should be included in its overall spending, the Remain campaign significantly outspent the Leave campaign in the 2016 EU referendum. According to the Electoral Commission (although the final figures are a little unclear), both sides spent around £16 million in the referendum, but the Government’s pro-EU mailing to every household alone was said to have cost nearly £10 million. Priti Patel is alleging that some of Remain’s spending looks as though it should not have been spent at all, following other allegations against the Leave campaign. All that extra spending – and the Remain campaign still lost by a clear margin. This raises an important question: does spending really make a difference in the outcome of British elections?

The evidence suggests “not really”; or, perhaps more accurately, “not necessarily”. There are many examples where election campaigns have spent big but come away with little. The 2017 election is a good example. Here, the Conservatives began the campaign with a 20 point lead, massively outspent Labour by around £18 million to £11 million, and lost seats and their majority. In 2010, the Conservatives outspent Labour by around £16 million to £8 million, against a terrible candidate in Gordon Brown, and only inched into Government through a Coalition. While the victorious No campaign in the 2011 referendum on the Alternative Vote spent significantly more than the Yes campaign, in the North East referendum in 2004, the Yes campaign was devastated by a No campaign run on a shoestring budget.

In each of these examples, where high-spending campaigns under-performed, you could argue the higher spending side might have performed worse than they otherwise might have done. Had they not spent all those millions more, maybe the Leave campaign in the referendum would have won by a bigger margin. That’s unknowable. But while it’s a reasonable case to make, there are simply too many examples where money hasn’t worked for us to conclude other than it only works for campaigns with the fundamentals in place.

Above all, this means a clear and compelling message that resonates emotionally. Through all the noise, people have got to hear a simple reason to turn out for a particular side. It’s extremely rare – the unionist campaign in Scotland, maybe – for a campaign to win outright without a clear message. In those examples above – the Conservatives in 2010 and 2017, the Remain campaign, the Yes campaign in 2004 – these campaigns struggled because they didn’t give reasons for people to turn out and vote. You can touch voters online, in print or through the media but if the message is weak all that activity won’t make a difference.

Of course, that’s easier said than done: getting to that point requires an organisation that can create the right strategy and consistently execute the right decisions fast. And that in turn is easier said than done. Campaigns are very often divided by disagreements over strategy or over who’s in control. Vanity and jealously are common. In such a climate it’s hard to make any decisions, let alone the right ones. Analyse a campaign with a poor message and more often than not it’s down to a poor decision-making structure and a correspondingly weak strategy. (Donald Trump somehow managed to create a clear message out of what looks like complete dysfunction internally).

If money doesn’t make the difference that it’s commonly believed, what does this mean for the regulation of money in elections? Are we becoming too obsessed with money? Personally, I’d relax the limits significantly if not totally, but insist on near real-time transparency from campaigns over their permitted donors. That way, money could form a genuine part of ongoing political debate and voters (and commentators) could decide whether they were happy or not with the types and levels of donations campaigns were receiving. But ultimately those looking for reasons why their preferred campaigns didn’t work are always better off looking at the failure of those campaigns to create the right message.