WEBVTT 00:00:00.000 --> 00:00:02.779 Welcome back to the deep dive. You know, there's 00:00:02.779 --> 00:00:04.900 this specific kind of anxiety that kicks in every 00:00:04.900 --> 00:00:07.200 four years in the United States. Oh, yeah. It's 00:00:07.200 --> 00:00:11.259 usually what, 7 .0 p .m. on a Tuesday in November. 00:00:11.779 --> 00:00:14.240 And you're just refreshing a Web page, staring 00:00:14.240 --> 00:00:16.679 at a screen, and you're just desperate for a 00:00:16.679 --> 00:00:18.660 sign. You're looking for that crystal ball. It's 00:00:18.660 --> 00:00:21.000 the holy grail of political science. Exactly. 00:00:21.260 --> 00:00:25.789 We want certainty in a process that is. I mean, 00:00:25.809 --> 00:00:28.550 it's inherently chaotic. And because of that, 00:00:28.629 --> 00:00:31.489 we tend to obsess over patterns. We look at bellwether 00:00:31.489 --> 00:00:34.250 counties in Ohio or we stare at the blue wall 00:00:34.250 --> 00:00:36.969 or we, you know, we analyze diner interviews 00:00:36.969 --> 00:00:39.409 and swing states. We convince ourselves that 00:00:39.409 --> 00:00:41.750 if we just look at this one spot, we can unlock 00:00:41.750 --> 00:00:44.210 the code for the whole country. It's just human 00:00:44.210 --> 00:00:46.469 nature. Yeah. We hate randomness. We want to 00:00:46.469 --> 00:00:48.310 believe there's a signal hidden in all the noise. 00:00:48.549 --> 00:00:50.829 Well, today we are going to talk about the granddaddy 00:00:50.829 --> 00:00:53.409 of all political signals. It's a phrase you have 00:00:53.409 --> 00:00:55.750 almost certainly heard. even if you're not a 00:00:55.750 --> 00:00:58.109 political janky. It sounds like this piece of 00:00:58.109 --> 00:01:01.450 ancient folklore. As Maine goes, so goes the 00:01:01.450 --> 00:01:03.630 nation. It really does have a nice ring to it. 00:01:03.710 --> 00:01:06.409 Sounds so authoritative, like Maine is some kind 00:01:06.409 --> 00:01:08.750 of mystical oracle tucked away up there that 00:01:08.750 --> 00:01:10.890 decides the fate of the presidency. But today 00:01:10.890 --> 00:01:13.849 we're going to destroy that myth, or at least 00:01:13.849 --> 00:01:15.329 we're going to dissect it. We're digging into 00:01:15.329 --> 00:01:18.670 the history, the data, and the Wikipedia entry 00:01:18.670 --> 00:01:23.170 for this specific maxim to figure out why it 00:01:23.170 --> 00:01:26.379 existed. why it worked for a solid century, and 00:01:26.379 --> 00:01:30.239 then, and this is the fun part, how it just spectacularly 00:01:30.239 --> 00:01:33.200 crashed and burned. It's a classic story of data 00:01:33.200 --> 00:01:35.299 failure, really. It's what happens when you rely 00:01:35.299 --> 00:01:37.700 on a metric that works perfectly right up until 00:01:37.700 --> 00:01:40.280 the moment it stops working entirely. So let's 00:01:40.280 --> 00:01:42.260 start with the why. I mean, if you look at a 00:01:42.260 --> 00:01:45.099 map, Maine's an outlier. It's geographically 00:01:45.099 --> 00:01:47.500 isolated, small population, and the winters are 00:01:47.500 --> 00:01:51.010 not for the faint of heart. So why on earth did 00:01:51.010 --> 00:01:52.969 anyone think Maine was the predictor? Well, the 00:01:52.969 --> 00:01:55.489 magic, if you want to call it that, wasn't political 00:01:55.489 --> 00:01:57.530 at all. It was purely logistical. It wasn't about 00:01:57.530 --> 00:01:59.370 who was voting. It was all about when they were 00:01:59.370 --> 00:02:02.069 voting. The calendar. The calendar. Yeah. Exactly. 00:02:02.129 --> 00:02:04.629 So you have to go back to 1820. That's when Maine 00:02:04.629 --> 00:02:06.590 splits off from Massachusetts, becomes its own 00:02:06.590 --> 00:02:09.650 state. And right from the jump, they set their 00:02:09.650 --> 00:02:11.650 election schedule differently. They held their 00:02:11.650 --> 00:02:15.150 big elections governor, Congress, in September. 00:02:15.150 --> 00:02:17.289 September. So two months before the rest of the 00:02:17.289 --> 00:02:20.699 country. Why the gap? It really comes down to 00:02:20.699 --> 00:02:23.240 what the source material calls the frigid November 00:02:23.240 --> 00:02:26.340 weather. I mean, you have to picture Maine in 00:02:26.340 --> 00:02:29.520 1820. No paved roads, no snow plows. No cars. 00:02:29.719 --> 00:02:32.439 No cars, right. Getting to the town square to 00:02:32.439 --> 00:02:35.039 vote in late November wasn't just a civic duty. 00:02:35.219 --> 00:02:37.419 It was a survival situation. Okay, that makes 00:02:37.419 --> 00:02:39.400 sense. A big snowstorm hits and your turnout 00:02:39.400 --> 00:02:42.259 just plummets. It drops to zero. And on top of 00:02:42.259 --> 00:02:45.340 that, Maine was a big agricultural state and 00:02:45.340 --> 00:02:47.560 the harvest cycle ended earlier up there. So 00:02:47.560 --> 00:02:49.719 it just made practical sense. Get it all done 00:02:49.719 --> 00:02:51.659 in September while the roads were still passable. 00:02:51.780 --> 00:02:53.979 So they voted for governor in Congress in September, 00:02:54.080 --> 00:02:56.099 but they still voted for president in November 00:02:56.099 --> 00:02:59.240 like everyone else. Yes. The presidential election 00:02:59.240 --> 00:03:01.199 date was set federally. They couldn't change 00:03:01.199 --> 00:03:03.759 that. But by the time November rolled around, 00:03:04.080 --> 00:03:06.659 Maine had already elected its entire state government. 00:03:06.840 --> 00:03:08.840 Which means the rest of the country got a massive 00:03:08.840 --> 00:03:11.729 spoiler alert. A sneak preview. Think about it. 00:03:11.750 --> 00:03:14.729 In the mid -1800s, you didn't have Gallup polling 00:03:14.729 --> 00:03:17.009 or, you know, Nate Silver running simulations. 00:03:17.449 --> 00:03:21.189 You had no hard data except for Maine. It was 00:03:21.189 --> 00:03:24.449 the only tangible data point available. Precisely. 00:03:24.669 --> 00:03:27.610 If a party did well in Maine in September, it 00:03:27.610 --> 00:03:30.090 was the only real -world evidence of how voters 00:03:30.090 --> 00:03:32.430 were feeling about the national platform. So 00:03:32.430 --> 00:03:34.729 this whole maxim wasn't just some superstition. 00:03:34.750 --> 00:03:37.090 It was based on a pretty solid track record, 00:03:37.169 --> 00:03:38.949 wasn't it? Oh, it was incredibly accurate for 00:03:38.949 --> 00:03:41.199 a very long time. The source material points 00:03:41.199 --> 00:03:44.020 out that between 1820 and 1932, that's a 112 00:03:44.020 --> 00:03:47.479 -year span, Maine's September results correctly 00:03:47.479 --> 00:03:49.840 predicted the party that would win the presidency 00:03:49.840 --> 00:03:53.520 in 21 out of 28 elections. 21 out of 28. That's 00:03:53.520 --> 00:03:56.300 a 75 % success rate. If a stock picker had that 00:03:56.300 --> 00:03:58.300 record today, they'd be a billionaire. It's impressive. 00:03:58.699 --> 00:04:01.180 And that's what built the legend. It really started 00:04:01.180 --> 00:04:04.419 to clutch on around 1840. The Whig Party candidate, 00:04:04.599 --> 00:04:07.379 a guy named Edward Kent, he wins the governorship 00:04:07.379 --> 00:04:10.419 in September. Two months later, The Whig presidential 00:04:10.419 --> 00:04:13.199 candidate, William Henry Harrison, wins the White 00:04:13.199 --> 00:04:15.139 House. Tippecanoe and Tyler, too. That's the 00:04:15.139 --> 00:04:17.220 one. Yeah. So people started connecting the dots. 00:04:17.279 --> 00:04:21.540 Then it happens again in 1888. Maine votes solidly 00:04:21.540 --> 00:04:24.519 Republican in September. Later that year, Benjamin 00:04:24.519 --> 00:04:27.199 Harrison, a Republican, wins the presidency. 00:04:27.560 --> 00:04:29.899 Now, wait a second. My history is a little rusty, 00:04:30.000 --> 00:04:32.500 but Benjamin Harrison lost the popular vote nationwide 00:04:32.500 --> 00:04:34.480 that year, didn't he? He did. Yeah, it was a 00:04:34.480 --> 00:04:36.839 messy election against Grover Cleveland. But 00:04:36.839 --> 00:04:39.379 Harrison won the Electoral College. So even in 00:04:39.379 --> 00:04:43.240 a weird split verdict election, Maine still technically 00:04:43.240 --> 00:04:45.699 called the winner. That's when the phrase really 00:04:45.699 --> 00:04:48.000 got locked into the American lexicon. OK, but 00:04:48.000 --> 00:04:49.860 let's dig into the data here because this is 00:04:49.860 --> 00:04:52.079 where it gets tricky. You said Maine was pretty 00:04:52.079 --> 00:04:54.339 heavily Republican for most of that era, right? 00:04:54.379 --> 00:04:56.459 It wasn't really a swing state. That is the crucial 00:04:56.459 --> 00:04:59.939 context. After the Civil War, it becomes a stronghold 00:04:59.939 --> 00:05:02.319 for the new Republican Party. It was not, you 00:05:02.319 --> 00:05:04.860 know, Pennsylvania or Florida today. It was deep 00:05:04.860 --> 00:05:07.959 red, to use our terms. So here's my question. 00:05:08.540 --> 00:05:11.220 If Maine is always voting Republican, how does 00:05:11.220 --> 00:05:13.519 that predict anything? I mean, if a Republican 00:05:13.519 --> 00:05:15.500 wins Maine and a Democrat wins the presidency, 00:05:15.800 --> 00:05:18.279 the rule is broken. But if a Republican wins 00:05:18.279 --> 00:05:22.180 Maine and the nation, isn't that just Maine doing 00:05:22.180 --> 00:05:24.540 what Maine always does? And you've just hit on 00:05:24.540 --> 00:05:26.399 the most sophisticated part of this whole thing. 00:05:26.779 --> 00:05:29.199 It's the part that pundits today often forget. 00:05:29.600 --> 00:05:31.819 It wasn't about the identity of the winner in 00:05:31.819 --> 00:05:34.500 Maine. It was about the margin of victory. OK, 00:05:34.579 --> 00:05:36.980 unpack that for me. How did the margin tell the 00:05:36.980 --> 00:05:39.339 story? Think of Maine. Not as a crystal ball, 00:05:39.459 --> 00:05:42.660 but more like a barometer for enthusiasm. Everyone 00:05:42.660 --> 00:05:44.680 knew the Republicans would probably win the governor's 00:05:44.680 --> 00:05:47.920 race. The real question was, by how much? So 00:05:47.920 --> 00:05:49.980 they were looking at the spread, like in sports 00:05:49.980 --> 00:05:53.279 betting. Exactly like sports betting. The baseline 00:05:53.279 --> 00:05:57.860 was Republican win. So if the GOP won Maine by 00:05:57.860 --> 00:06:00.939 a massive landslide in September, say 20 points 00:06:00.939 --> 00:06:03.160 or more, that was a signal that the party base 00:06:03.160 --> 00:06:05.519 was energized and that momentum would probably 00:06:05.519 --> 00:06:08.120 carry over to the real swing states in November. 00:06:08.420 --> 00:06:11.379 But if they barely scraped by. That was the danger 00:06:11.379 --> 00:06:14.000 zone. If the Republicans won but the margin was 00:06:14.000 --> 00:06:16.920 narrow, say less than 5%, that was seen as a 00:06:16.920 --> 00:06:19.439 disaster. It meant the Democrats were overperforming. 00:06:19.639 --> 00:06:22.160 And if Democrats were overperforming in a stronghold 00:06:22.160 --> 00:06:23.540 like Maine, they were probably going to crush 00:06:23.540 --> 00:06:25.180 it in the states that actually mattered. Wow. 00:06:25.300 --> 00:06:28.899 Yes. So a close run Republican victory in September 00:06:28.899 --> 00:06:31.870 was basically a warning siren for November. It 00:06:31.870 --> 00:06:34.310 predicted a Democratic national victory. That 00:06:34.310 --> 00:06:37.050 is fascinating. So a Republican win in Maine 00:06:37.050 --> 00:06:38.970 could actually predict a Democrat in the White 00:06:38.970 --> 00:06:41.569 House. It did. And the parties took it so seriously. 00:06:41.730 --> 00:06:44.129 The source notes they went to considerable lengths 00:06:44.129 --> 00:06:46.430 to win these elections. They would pour money 00:06:46.430 --> 00:06:49.170 and resources into Maine in August. For a state 00:06:49.170 --> 00:06:51.509 with just a handful of electoral votes, it was 00:06:51.509 --> 00:06:54.449 all psychological warfare. 100 percent. It was 00:06:54.449 --> 00:06:57.110 about shaping the narrative. If you could dominate 00:06:57.110 --> 00:07:00.550 the headlines in September with a GOP smashes 00:07:00.550 --> 00:07:03.829 main story. You could demoralize the opposition 00:07:03.829 --> 00:07:06.670 for two months. It works until it doesn't. And 00:07:06.670 --> 00:07:09.029 this brings us to the villain of our story. The 00:07:09.029 --> 00:07:12.589 year 1936. The year the bellwether broke. I love 00:07:12.589 --> 00:07:14.569 this part. It feels like a movie script. You've 00:07:14.569 --> 00:07:16.949 got the setup, the hubris, and then the fall. 00:07:17.089 --> 00:07:20.189 So set the scene. Okay. It's 1936. The Great 00:07:20.189 --> 00:07:22.790 Depression is in full swing. Yeah. FDR is in 00:07:22.790 --> 00:07:24.790 the White House with the New Deal. And the Republicans 00:07:24.790 --> 00:07:27.730 are desperate to get back in power. They nominate. 00:07:28.000 --> 00:07:30.319 Alf Landon, the governor of Kansas. And the Republicans 00:07:30.319 --> 00:07:33.339 are feeling surprisingly good, right? They are. 00:07:33.620 --> 00:07:36.139 And a huge reason for that confidence comes from 00:07:36.139 --> 00:07:39.160 Maine. In September 1936, the Maine Republicans 00:07:39.160 --> 00:07:41.800 have a field day. They elect a Republican governor, 00:07:41.980 --> 00:07:44.540 Louis Barrows. They elect an all -Republican 00:07:44.540 --> 00:07:47.079 congressional delegation. A clean sweep. A total 00:07:47.079 --> 00:07:49.399 clean sweep. And the party leadership went wild. 00:07:49.540 --> 00:07:51.259 They thought the tide had turned. They looked 00:07:51.259 --> 00:07:53.259 at their trusty barometer and saw clear skies 00:07:53.259 --> 00:07:56.160 ahead. They were convinced this predicted a national 00:07:56.160 --> 00:07:59.500 rejection of FDR. Oh, absolutely. The vice presidential 00:07:59.500 --> 00:08:02.180 candidate, Frank Knox, actually said, as Maine 00:08:02.180 --> 00:08:04.740 goes, so goes the nation, is still the gospel 00:08:04.740 --> 00:08:06.879 truth. They were trumpeting this everywhere. 00:08:07.160 --> 00:08:09.120 So September looks great for them, then November 00:08:09.120 --> 00:08:11.480 arrives. And the floor just drops out. How bad 00:08:11.480 --> 00:08:14.300 was it? It wasn't just a loss, it was an annihilation. 00:08:15.459 --> 00:08:18.779 FDR didn't just win. He won one of the greatest 00:08:18.779 --> 00:08:21.300 electoral landslides in American history up to 00:08:21.300 --> 00:08:23.019 that point. And what happened to Alf Landon, 00:08:23.139 --> 00:08:25.740 the guy whose party won the bellwether? Poor 00:08:25.740 --> 00:08:28.850 Alf Landon. He carried exactly two states. Let 00:08:28.850 --> 00:08:31.370 me guess. Maine was one of them. Maine and Vermont. 00:08:32.149 --> 00:08:34.450 That's it. Just those two states huddled together 00:08:34.450 --> 00:08:37.269 in the corner. Landon got a total of eight electoral 00:08:37.269 --> 00:08:43.029 votes. FDR got 523. 523 to eight. That's astounding. 00:08:43.169 --> 00:08:45.029 It's still the smallest electoral vote total 00:08:45.029 --> 00:08:47.590 for a major party nominee since the two -party 00:08:47.590 --> 00:08:50.649 system really began in 1856. So the bellwether 00:08:50.649 --> 00:08:52.870 didn't just fail. It was basically the only place 00:08:52.870 --> 00:08:55.179 in the country that was wrong. It was perfectly 00:08:55.179 --> 00:08:57.799 out of sync with America. It was an island. And 00:08:57.799 --> 00:09:00.320 the maxim, as Maine goes, so goes the nation, 00:09:00.559 --> 00:09:03.259 became an instant joke. And this gives us one 00:09:03.259 --> 00:09:06.080 of the absolute best clapbacks in political history 00:09:06.080 --> 00:09:09.440 from James Farley, FDR's campaign manager. He 00:09:09.440 --> 00:09:12.700 didn't miss a beat. When he saw that map, just 00:09:12.700 --> 00:09:15.799 a sea of blue with two little red dots, Farley 00:09:15.799 --> 00:09:18.679 announced to the press, as Maine goes. So goes 00:09:18.679 --> 00:09:21.700 Vermont. As Maine goes, so goes Vermont. That 00:09:21.700 --> 00:09:24.580 is just savage. It was the perfect dismissal. 00:09:24.600 --> 00:09:26.559 He's saying, sure, you two are statistically 00:09:26.559 --> 00:09:28.799 correlated. You're just totally irrelevant to 00:09:28.799 --> 00:09:31.080 the rest of us. And that pretty much killed the 00:09:31.080 --> 00:09:33.039 superstition on the spot, didn't it? It shattered 00:09:33.039 --> 00:09:35.460 the credibility. If you look at the data after 00:09:35.460 --> 00:09:39.220 1936, the correlation just evaporates. Between 00:09:39.220 --> 00:09:43.320 1940 and 1956, Maine's September winner only 00:09:43.320 --> 00:09:45.480 matched the November presidential winner in one 00:09:45.480 --> 00:09:47.679 of those five elections. One out of five. You'd 00:09:47.679 --> 00:09:50.139 be better off flipping a coin. Much better. The 00:09:50.139 --> 00:09:53.659 magic was gone. And finally, in 1959, Maine changed 00:09:53.659 --> 00:09:56.799 its laws. Since 1960, they voted in November 00:09:56.799 --> 00:09:59.580 with everyone else. The preview era was officially 00:09:59.580 --> 00:10:01.620 over. I want to circle back to that Farley quote 00:10:01.620 --> 00:10:03.899 for a second. As Maine goes, so goes Vermont. 00:10:04.600 --> 00:10:06.500 The source material calls them the Twin States. 00:10:07.210 --> 00:10:09.210 Was Farley just making a joke or is there real 00:10:09.210 --> 00:10:12.929 data behind that? No, he was spot on. Mathematically, 00:10:13.190 --> 00:10:16.009 the political correlation between Maine and Vermont 00:10:16.009 --> 00:10:19.970 is incredibly strong. Almost unnerving. How strong 00:10:19.970 --> 00:10:22.029 are we talking? Since the Republican Party was 00:10:22.029 --> 00:10:25.450 born in 1854, they've voted for different presidential 00:10:25.450 --> 00:10:28.190 candidates in the same election only twice. Wait, 00:10:28.250 --> 00:10:32.669 only twice in over 150 years? Only twice. Through 00:10:32.669 --> 00:10:35.289 the Civil War, the Gilded Age, the Depression, 00:10:35.590 --> 00:10:39.350 the Cold War. They have walked in lockstep. OK, 00:10:39.429 --> 00:10:41.210 I have to know, what were the two times they 00:10:41.210 --> 00:10:43.990 broke up? The first was 1912. That was the Bull 00:10:43.990 --> 00:10:46.350 Moose election. You had Teddy Roosevelt running 00:10:46.350 --> 00:10:48.590 as a third -party candidate against the incumbent 00:10:48.590 --> 00:10:51.090 Republican, Taft. Right, the great Republican 00:10:51.090 --> 00:10:53.809 split. Exactly. Teddy was very popular in Maine, 00:10:53.929 --> 00:10:56.509 and he split the vote so much that the Democrat, 00:10:56.710 --> 00:10:59.009 Woodrow Wilson, snuck through and won the state. 00:10:59.210 --> 00:11:01.549 But Vermont stayed loyal to the official Republican 00:11:01.549 --> 00:11:04.450 candidate, Taft. So it took a former president 00:11:04.450 --> 00:11:06.690 blowing up his own party to separate them. What 00:11:06.690 --> 00:11:09.789 was the second time? 1968. Richard Nixon for 00:11:09.789 --> 00:11:12.169 the Republicans. Vermont voted for Nixon, but 00:11:12.169 --> 00:11:13.970 Maine voted for the Democrat Hubert Humphrey. 00:11:14.110 --> 00:11:16.870 Why the split that time? Humphrey had an ace 00:11:16.870 --> 00:11:19.230 up his sleeve. His running mate for vice president 00:11:19.230 --> 00:11:22.669 was Senator Edmund Muskie. Ah, the favorite son. 00:11:23.009 --> 00:11:26.330 Exactly. Muskie was a beloved figure in Maine 00:11:26.330 --> 00:11:29.350 politics. His presence on the ticket was just 00:11:29.350 --> 00:11:31.700 enough to flip Maine to the Democrats. while 00:11:31.700 --> 00:11:34.460 Vermont stayed Republican. It's pretty telling 00:11:34.460 --> 00:11:36.440 that it took these extraordinary circumstances 00:11:36.440 --> 00:11:39.519 to drive a wedge between them. Otherwise, they're 00:11:39.519 --> 00:11:41.940 politically identical. It shows deep cultural 00:11:41.940 --> 00:11:44.879 similarities, although it is worth noting they 00:11:44.879 --> 00:11:47.259 aren't Republican strongholds anymore. Right. 00:11:47.419 --> 00:11:50.259 Today, the Northeast is pretty reliably blue. 00:11:50.360 --> 00:11:53.379 And they shifted together. The source notes that 00:11:53.379 --> 00:11:56.480 starting around 1992, both Maine and Vermont 00:11:56.480 --> 00:11:59.059 became reliably Democratic in presidential years. 00:11:59.179 --> 00:12:00.940 So they're still twins. They just switch sides 00:12:00.940 --> 00:12:02.919 together. I think there's a bigger lesson here. 00:12:03.059 --> 00:12:04.899 We started this talking about how much we crave 00:12:04.899 --> 00:12:07.899 patterns. The Republicans in 1936 weren't being 00:12:07.899 --> 00:12:10.120 stupid. They had 100 years of data on their side. 00:12:10.220 --> 00:12:12.340 They were following the trend line. They were 00:12:12.340 --> 00:12:15.299 following a trend line that had expired. And 00:12:15.299 --> 00:12:17.159 that's the danger of political data. It's not 00:12:17.159 --> 00:12:20.360 physics. Gravity works the same way in 1820 as 00:12:20.360 --> 00:12:24.259 it does in 1936. But in politics, the underlying 00:12:24.259 --> 00:12:27.740 variables, demographics, party platforms are 00:12:27.740 --> 00:12:31.039 constantly shifting. The as Maine goes maxim 00:12:31.039 --> 00:12:33.940 worked when the political divide in Maine sort 00:12:33.940 --> 00:12:36.639 of mirrored the national divide. Correct. But 00:12:36.639 --> 00:12:39.620 the New Deal realigned American politics. It 00:12:39.620 --> 00:12:42.559 brought in coalitions of urban workers and minorities 00:12:42.559 --> 00:12:45.379 that just didn't really exist in the same way 00:12:45.379 --> 00:12:47.980 in Maine. So Maine stood still and the country 00:12:47.980 --> 00:12:50.500 moved on. The bellwether didn't break. The flock 00:12:50.500 --> 00:12:52.919 just flew to a different pasture. And the Republicans 00:12:52.919 --> 00:12:54.860 missed it because they were looking at the one 00:12:54.860 --> 00:12:56.720 data point that confirmed what they wanted to 00:12:56.720 --> 00:12:59.159 believe. They mistook a local weather report 00:12:59.159 --> 00:13:01.899 for a national climate forecast. That's a perfect 00:13:01.899 --> 00:13:03.620 way to put it. It really makes you wonder what 00:13:03.620 --> 00:13:06.000 our main is today. We don't use that phrase, 00:13:06.080 --> 00:13:07.840 but we have others. We talk about the Blue Wall 00:13:07.840 --> 00:13:10.039 or bellwether counties. We're always looking 00:13:10.039 --> 00:13:13.379 for that shortcut. We absolutely are. The source 00:13:13.379 --> 00:13:15.820 material mentions things like the Missouri bellwether. 00:13:16.570 --> 00:13:18.929 We're always desperate for a simple rule that 00:13:18.929 --> 00:13:21.110 tells us the whole story so we don't have to 00:13:21.110 --> 00:13:23.190 deal with the messy complexity of 50 different 00:13:23.190 --> 00:13:26.389 states. But the lesson of 1936 is that those 00:13:26.389 --> 00:13:28.809 shortcuts work right up until the moment they 00:13:28.809 --> 00:13:32.210 spectacularly fail. And usually that failure 00:13:32.210 --> 00:13:34.909 happens during a realigning election when the 00:13:34.909 --> 00:13:36.750 rules of the game are changing under our feet. 00:13:37.200 --> 00:13:39.379 That's when the old metrics are most dangerous. 00:13:39.559 --> 00:13:41.620 If you're looking at a map from 20 years ago 00:13:41.620 --> 00:13:43.720 to predict tomorrow, you're going to end up like 00:13:43.720 --> 00:13:46.080 Al Flandin. Winning two states and becoming a 00:13:46.080 --> 00:13:49.059 trivia question. Precisely. So the next time 00:13:49.059 --> 00:13:52.059 you hear a pundit say, this county has predicted 00:13:52.059 --> 00:13:55.500 the winner for the last 50 years, maybe, take 00:13:55.500 --> 00:13:57.519 it with a grain of salt. Or a grain of granite, 00:13:57.519 --> 00:13:59.879 in Maine's case. Fair enough. Well, that brings 00:13:59.879 --> 00:14:02.500 us to the end of our deep dive. A story of frozen 00:14:02.500 --> 00:14:05.580 roads, margin analysis, and the danger of assuming 00:14:05.580 --> 00:14:08.159 the future will always look like the past. A 00:14:08.159 --> 00:14:10.179 good reminder to always question your assumptions. 00:14:10.500 --> 00:14:12.559 Thanks for listening. We'll catch you on the 00:14:12.559 --> 00:14:14.080 next deep dive. Goodbye, everyone.