2006 Democratic gubernatorial primary

UPDATE: Scroll down for an analysis written just after the Democratic gubernatorial primary in 2006. You can get town-level election results as an Excel spreadsheet or a PDF. What you cannot do is get any of this data from the secretary of state's website.

The official election data (which you can buy for $20  at the State House Bookstore) show that 57.8 percent of the 4,535 voters who went to the polls in Wakefield turned in blank ballots in the governor's race. This is doubtlessly inaccurate (the runner-up in this category is Southbridge, where 9.4 percent of the voters turned in blanks), but it's doubtful that the secretary of state's office will ever make a correction, since the outcome of the primary statewide was not close. On the day after the election, the Boston Globe reported that Wakefield cast 1,798 votes for Patrick; 1,588 votes for Gabrieli; and 1,124 votes for Reilly, which seems like a plausible distribution of votes; it did not tabulate blanks.

Woburn may be seeing a lot of Kerry Healey this fall. The results of last week’s gubernatorial primary showed few weak spots for Democratic nominee Deval Patrick, and unless Healey can drastically change the dynamics of the race, her best chance at victory lies in the mix of suburbs and smaller cities between Route 128 and the New Hampshire border.

In the primary, Patrick won all 10 of the state’s political regions, as defined by CommonWealth, each of which casts about 10 percent of the vote in a general election (see “Shifting Ground,” Spring ’06), compared with seven regions won by Shannon O’Brien in the four-candidate Democratic primary in 2002, and fell below 40 percent in only one of them: Post-Industria. (See chart at the bottom of this page for more data.) This region includes the cities of Lowell, Lawrence, and Lynn, but most of its votes come from bedroom communities such as Chelmsford and Tyngsborough. Patrick won less than one-third of the vote in only 10 communities across the state, and five of them were in this region: Billerica, Revere, Saugus, Tewksbury, and Woburn. This once-Democratic region swung toward the Republicans in 1998 and 2002, and analysis of primary returns indicates that Healey, this year’s GOP nominee, will have to push Post-Industria even more to the right in order to keep her party in the corner office.

By some measures, Patrick is the strongest Democratic nominee since 1990, when the Republicans began their current winning streak. Healey may have wished him as an opponent (over the more moderate Chris Gabrieli and Tom Reilly), but she couldn’t have been happy about the size of his victory margin last week, or about the unexpectedly strong turnout in the primary. Patrick polled 452,000 votes, which represents almost half of the votes won by Shannon O’Brien in the close general election of 2002. It also equals 20.6 percent of all votes for governor that November. This gives him a bigger base of votes going into the general election than any of the four previous Democratic nominees.

If turnout equals that of 2002, and Patrick holds on to the people who voted for him in the primary, he already has up to 46 percent of the votes he needs to win in November. In order to top him, Healey will have to win about 62 percent of the voters who voted for one of Patrick’s opponents or sat out the Democratic primary. By comparison, Shannon O’Brien’s primary total in 2002 represented 11.1 percent of the total vote in the fall — meaning that Mitt Romney needed to win about 56 percent of the voters who were up for grabs after the primary. To win, Healey needs to do at least one of three things: win over most of the people who voted for Gabrieli or Reilly last week (and were most prevalent in Post-Industria); convince a significant number of people who voted for Patrick to change their minds; or dramatically increase turnout over 2002 levels — and then get most of the new voters to vote Republican.

DEVAL VS. THE DUKE

Patrick won last week’s primary by 23 points, the largest spread in a contested race for the Democratic gubernatorial nomination in at least three decades. His grassroots organization is frequently compared to that of the last Democratic governor, Michael Dukakis, who beat strong candidates — a sitting attorney general and a sitting governor — in the 1974 and 1982 primaries. (In between those victories, of course, he was the victim of an upset himself, by Ed King in 1978.) And the geographic contours of Patrick’s victory resemble that of the Duke in his rematch against King in 1982. Both ran strongest in the same three regions: Left Fields, which joins Cambridge, Somerville, the tip of Cape Cod, and most of the rural third of the state; Bigger Boston, which includes the capital city and a few outskirts; and Shopper’s World, the affluent and well-educated suburbs to the west of Boston. And they were the weakest in the same three regions: Post-Industria; the Brink Cities, which include Fall River, New Bedford, and Springfield; and Ponkapoag, which includes middle-class and largely Irish suburbs to the south of Boston.

Indeed, the biggest difference between the 1982 and 2006 primaries was greater geographic polarization, even though the ideological differences among the three candidates this year were not nearly as sharp as those between Dukakis and King. The city of Boston, for example, went from 53 percent for Dukakis to 59 percent for Patrick. And while Patrick ran about four points behind Dukakis’s 1982 showing statewide, his 67 percent in the Left Fields region was three points better. In Somerville, Patrick was up from 53 percent to 58 percent, but the differences were starkest in the western part of the region; in Northampton, he was up from 62 percent to 76 percent, and in Great Barrington, he was up from 65 percent to 83 percent.

Of course, with a quarter-century having passed, these changes probably have less to do with people changing their views than with who’s been moving in and out of specific communities, but the demographic trends seem to have benefited Patrick in September, even if November is still an open question. In many places where he fell short of a majority of the vote, he still did better than the liberal candidate of 1982. Taunton, in the Ponkapoag region, gave Dukakis only 35 percent (and gave liberal darling Robert Reich a miserable 12 percent in the 2002 primary), but Patrick polled 47 percent this time around.

On the other side, the Post-Industria region gave Dukakis 48 percent in 1982, Patrick fell to 37 percent last week. Stoneham went from 51 percent for Dukakis to a tad more than 33 percent for Patrick, and Billerica went from 48 percent to 32 percent. Part of the reason may have been that voters in these towns were especially familiar with Reilly from his days as Middlesex County district attorney, but that doesn’t explain why Patrick got only 31 percent in Essex County’s Saugus, a town that gave Dukakis 51 percent in 1982.

WINNING FROM TOP TO BOTTOM

At the beginning of this year, Patrick was more often compared to another Clinton appointee, Robert Reich, who tried to build a campaign around liberal grassroots activism but finished a rather distant second to Shannon O’Brien in the 2002 Democratic primary. But Patrick doubled Reich’s share of the vote, from 25 percent to 50 percent, and more than tripled his vote total, thanks to a higher turnout. The professorial Reich may have been doomed by his inability to connect with voters in urban areas outside of Route 128; in the Brink Cities, which has the lowest median income and the fewest number of college graduates of all the regions, he got only 16 percent of the vote. Last week, Patrick got 44 percent there. Compared with Reich, he improved from 15 percent to 53 percent in Springfield, from 14 percent to 43 percent in New Bedford, and from 12 percent to 39 percent in Fall River.

In essence, Patrick held onto Reich’s support at the top of the socioeconomic scale — Shopper’s World, the most affluent and highly educated of the 10 regions, went from 34 percent for Reich to 55 percent for Patrick — while building another base, almost from scratch, at the lower end of the scale.

It was the lower middle of the socio-economic spectrum that was less enthusiastic for Patrick in the primary, even if he still mostly prevailed in the communities that fall into that category. In Post-Industria, which falls a bit below the state average in terms of household income and the percentage of adults with college degrees, he posted the smallest gain over Reich’s showing, going from 17 percent to 37 percent.

Take two Post-Industria cities as examples. In Woburn (where 30 percent of adults over 25 had college degrees, according to the last national census, compared with a Bay State average of 33 percent), Patrick got 31 percent to Reich’s 19 percent. And in Methuen (with a 1999 median household income of $49,600, compared with a statewide average of $50,500), Patrick got 42 percent to Reich’s 28 percent. Improvements to be sure, but not on the scale found elsewhere in Massachusetts.

GOOD FOR THE GOP?

It’s not quite accurate to say that Kerry Healey has a chance of winning with the same voters that elected Republicans Bill Weld in 1990, Paul Cellucci in 1998, and Mitt Romney in 2002, since all three won with very different voting blocs. A replay of Bill Weld’s Bizarro World victory over John Silber — in which he won Brookline, Newton, and even the People’s Republic of Cambridge — is clearly out of the question. And even Mitt Romney’s success in the affluent suburbs a bit farther out from Boston may not be possible to repeat, given Patrick’s strength in those communities last week.

Romney ran up his biggest tally over Shannon O’Brien in the exurban region of Offramps, and in three other regions (Shopper’s World, Stables and Subdivisions, and Cranberry Country) he ran three to four points above Paul Cellucci’s 1998 showing — allowing him to compensate for O’Brien’s improvement over Scott Harshbarger’s 1998 vote in the state’s urban areas. But Patrick easily won all of these regions and ran especially strongly in the one that swung the most dramatically toward Romney in 2002. Shopper’s World, which went from a Democratic margin of 10,000 votes in 1998 to a Republican margin of 15,000 votes four years later, was Patrick’s third best region, and his total was the equivalent of 24 percent of all the votes cast in the region in the 2003 general election — which means that Healey might have get almost two-thirds of all voters in the region who didn’t vote in the primary in order to duplicate the GOP win of 2002. And her task may be even more daunting in the region’s biggest city, Newton, which gave went from a Democratic margin of 8,465 in the 1998 gubernatorial election to a Democratic margin of 5,926 four years later, giving Romney a net gain of 2,539 (his biggest in the state). Last week Patrick got 64 percent in Newton, and his total already equaled 32 percent of all votes cast in the city in the 2002 general election, leaving few opportunities for the GOP to pick up votes here in the fall.

If “Romney Democrats” seem to be a tough group for Healey to win over, “Cellucci Democrats” may be more receptive to the GOP. There were three regions where Cellucci did especially well in 1998, running more than five points ahead of Bill Weld’s winning percentage of 1990: Brink Cities, MidMass, and Post-Industria. Patrick ran behind his statewide showing in all three regions last week, and his total in Post-Industria equaled only 15 percent of all votes cast in the 2002 general election. With as much as 85 percent of all votes up for grabs in this region come November, this is where a Healey push could yield some dividends. Take Lowell, the region’s biggest community, for example. That city went for Democrat John Silber by a 3,000-vote margin in 1990 but Cellucci carried it by about 750 votes eight years later. Last week Reilly narrowly won there (it was one of only eight towns won by the attorney general in the entire state), and Patrick’s total equaled 16 percent of the 2002 general vote. Patrick fell below this 16 percent mark in several other towns that swung strongly toward Cellucci in 1998, including Post Industria’s Dracut, Methuen, and Revere; the Brink Cities’ Chicopee; and MidMass’s Blackstone, Douglas, and Millville. While there is much speculation about how much campaigning Romney will do for his lieutenant governor this fall, it may be his predecessor (not counting Acting Gov. Jane Swift) who could help the most.