The tale of the two Bobs
Using a spreadsheet, the Hamilton Spectator was able to explain a surprising election result

Sometimes you get lucky, and then sometimes you get really lucky.

That's what happened to us as the Hamilton Spectator on municipal election night Nov. 13.

It was the night a computer-assisted reporting project went from interesting to extraordinary because of an unexpected turn of events.

It had seemed to be a done deal. Our six-term mayor Bob Morrow was going to waltz back into office. All the polls said so. The pundits said so. Morrow thought so.

But then the voters spoke, and we were all wrong.

Instead of Morrow, the people's choice was Bob Wade, the quiet, polite mayor of Ancaster. The rich bedroom community is one of six towns and cities Ontario Premier Mike Harris has merged into the new city of Hamilton.

That night I had the job of collecting poll-by-poll election results from city hall, and typing them into a computer spreadsheet. It would provide us with an instant analysis of how Morrow won. It was to be a dissection of an anticlimax.

By 8:30 that night it was clear we'd be dissecting something much different.

Wade was ahead from the earliest results and never trailed. By the time the night was done, he had beaten Bob Morrow by more than 10,000 votes.

The spreadsheet helped explain why.

We had tried this kind of instant analysis in a provincial by-election a couple of months before the municipal vote. That analysis was the brainwave of reporter Dan Nolan, who developed the idea during one of the introductory CAR courses I taught at the paper several months ago.

We obtained the poll-by poll results of the June 1999 general election from the Elections Ontario web site and saved them in a Microsoft Excel worksheet.

We then added columns for the live results on by-election night and added formulas that would instantly calculate the raw change in the number of votes for each party, the percentage change and the voter turnout in each poll. We also set up the sheet to automatically calculate the overall results in each of the towns in the riding.

On the by-election night, the time between the first results and the copy deadline was barely more than an hour, so we pre-selected a range of polls where the ruling Conservatives had done particularly well over the years. We would write our early edition story based on that sample. The overriding issue in the by-election was the Harris government's decision to force the town of Flamborough to join with the new Hamilton. So if the Tories lost badly in polls they had dominated, it would help unlock the meaning of the overall results.

I sat in a small anteroom at the returning office that night, along with an intern, and madly entered more than 150 results. He read from paper slips. I typed.

The spreadsheet output spoke for itself. The Tory vote was plummeting, but the increase in votes for the Liberal, who won, was not nearly so large. The votes weren't going to the NDP either. The inevitable conclusion: Tory voters stayed home. It became a key piece of our coverage for the next day's paper.

Little did we know it, but we had essentially invented a new kind of election reporting, the instant spreadsheet analysis. Until now, this kind of poll-by-poll dissection has been done laboriously by hand, if at all. Few media ever get much beyond the aggregate results. We told a much more complete story.

The stage was now set for municipal election night, and the stunning Bob Wade victory.

When the results started pouring in, our carefully-constructed spreadsheet again told the story. As with the by-election, I wrote the first-brush analysis for the next day's paper, and followed up with a more detailed look at the voting patterns the day after.

As it turned out, Bob Morrow won the old city of Hamilton all right, but not by enough. Nearly 20,000 of his city supporters deserted him. To add to the misery, turnouts were lowest in polls where Morrow was strongest, so he got little bang for his electoral buck.

Wade, on the other hand, was strongest where turnout was most enthusiastic, in the suburbs and he won enough votes in the city to take him over the top.

In his hometown of Ancaster, more than half the people voted. That's considered impressive in a municipal election. Eighty per cent of those voters chose Wade.

He also had the support of the more affluent voters, who tend to come out in the greater numbers. By contrast, Morrow's strength was concentrated in poorer, blue-collar districts of industrial Hamilton.

The simple tool of a computer spreadsheet helped us unravel the complex puzzle of how an apparently-popular 18-year mayor finally went down to defeat. Again, the competition couldn't touch us. It's an easy story to do, one that is entirely portable. With the predictability of elections, it's also easy to plan this one ahead.

Sometimes the simplest ideas are the most inspired.

Fred Vallance-Jones is municipal affairs reporter with The Hamilton Spectator.