It’s his party: American primaries

Hello and welcome to The Intelligence from The Economist. In New York, I'm Jon Fassman.

And in London, I'm Jason Palmer. Every weekday, we provide a fresh perspective on the events shaping your world.

Eyes in the skies increasingly have a new tool at their disposal. It's called wide-area motion imaging, and it's a better way to look for changes on the ground. We ask how to sift through the slew of data it produces and what threats to privacy it presents.

And food prices are going up all over, but there's a spike coming for a particular snack in Britain: berries. We ask why the war in Ukraine will have such a pointed effect on the strawberries and cream industrial complex.

First up, though. Yesterday, primary voters in 5 states went to the polls. In Idaho and western North Carolina, the establishment had a good night. Idaho's incumbent governor, Doug Little, fended off a challenge from a far-right insurgent backed by Donald Trump. And in North Carolina, Madison Cawthorn, A young congressman who called Ukrainian President Volodymyr Zelensky a thug and who claimed without evidence to have been invited to orgies by his colleagues lost to an establishment-backed state senator despite having Donald Trump's support. But elsewhere, Mr. Trump's endorsement proved effective. Ted Budd, who he endorsed in North Carolina's Senate primary, will face Cheri Beasley, a former state Supreme Court Chief Justice, in a contest to replace Tom Tillis, a retiring Republican senator. But the most closely watched races were those for governor and senator in Pennsylvania. The results aren't fully in yet for the Republican Senate race, which is neck and neck. At his election night party, one contender, Mehmet Oz, thanked his most prominent supporter.

Let's start with 45, President Trump. God bless you, sir, for putting so much effort into this race. I will make you proud.

His opponent, David McCormick, sounded equally confident.

Unfortunately, we're not going to have resolution tonight, but we can see the path ahead.

We can see victory ahead, and it's all because of you.

Pennsylvania is a perennial swing state. Yesterday's elections there have national implications, both in which candidates emerge on top and in what the races say about broader dynamics in American politics.

I spent yesterday going across the state of Pennsylvania trying to take the temperature of the voters ahead of the big races on Tuesday. So I started yesterday morning in Scranton in the Northeast, this small city, traditionally Democratic and working class, then spent much of the day in rural areas around the Appalachian Mountains and ended the day in the city of Harrisburg, which is the state capital in the South.

Daniel Remler reports on the United States for The Economist.

These races really matter. So for the governor's race, of course, whoever wins this gets to appoint the secretary of state who administers the elections. That'll be critical as allegations of voter fraud probably crop up again in 2024, and then as potentially laws around abortion will need to be debated and possibly passed. The Senate race is incredibly important because Democrats currently have a razor-thin majority in the upper chamber of the U.S. Congress. And this is one of the very few seats currently held by a Republican who is retiring that Democrats could flip. And so they will need to do that if they have any hopes of retaining their majority in the chamber.

And so, Daniel, with so much on the line, how did things turn out?

Well, it's really not clear. So on the Republican side for governor, Doug Mastriano, state senator, won that race pretty handily. He's about as closely aligned with Mr. Trump as you can get. And he's very extreme in his policy views as well. So he supports a total ban on abortion with no exceptions for rape or incest. And he reiterated his support for that in his victory speech on Tuesday.

Democrat governors around the nation here want to kill babies even up to birth, and some are talking about after birth. That's extreme. That's denying the science. That's immoral. Every baby deserves a right to life. What do you think?

He supported abolishing property taxes that fund local schools, as well as making gun laws completely unenforceable in the state of Pennsylvania. And he did get Mr. Trump's endorsement, but that came just a few days before the primary, when it already seemed like he was very well ahead. And on the Senate side, it seemed like Mr. Trump's endorsement didn't prove to be the winning ticket. So there he backed Mehmet Oz, who's a celebrity TV doctor, but he ended the race in a dead heat with Dave McCormick, a former hedge fund executive. And that race now looks like it's headed to a recount.

Tell us a little more about these two. Who are they? What do they stand for?

Dr. Mehmet Oz is a celebrity TV doctor who had a long-running show on network television, and he has no political experience before. And many voters in Pennsylvania told me that they sort of saw him as not really a Pennsylvanian. He famously or infamously lived in New Jersey for much of his life, and even though he did spend much of his earlier life in Pennsylvania, seems to be sort of a newcomer to the state. But from what I've seen of him on the campaign trail, he's proved to be sort of a political natural. He has a very easy way of speaking to voters and clearly connects with enough of them to have built a significant base of support in the state. Dave McCormick is the former CEO of Bridgewater Associates, a huge hedge fund. Prior to that, he worked in the Bush administration as an undersecretary of the Treasury. He'd been in the military. He likes to say that he grew up on a Christmas farm in Western Pennsylvania. So he has a sort of mix of the worldly background, but he's been trying to ground it in a sort of more personal story. He, like Dr. Oz, sort of suffers from the slight impression that he's this very rich guy who hasn't spent that much time in the state recently. And maybe isn't the most conservative candidate in the race. But conversely, that also makes him more appealing as a possibility to take on Democrats in November.

And so Oz, despite having Trump's endorsement, couldn't win outright. Does that tell us anything about, about, about President Trump's influence over the Republican Party, about his ability to get candidates elected?

Mr. Trump's ideas and energy seem to really be holding the Republican Party together, But there seem to be pretty strict limits on his personal influence. So all of the candidates in both the governor's race and the Senate race really cast themselves as sort of Trump-like figures on policy terms. And you've seen this in other races too. So a few weeks ago, J.D. Vance won the Ohio primary after getting Mr. Trump's endorsement and sort of mimicking a lot of his policy views. But then yesterday, a few of his picks went down to defeat. So in Idaho, the lieutenant governor of the state, Janice McGeachin, she opposed the incumbent governor, also a Republican, and she was backed by Mr. Trump, but she lost to Brad Little, the incumbent governor. Nevertheless, it's clear that the party has basically recast itself in Mr. Trump's image. So Dr. Oz and Dave McCormick, neither of them had ever run for office before. They didn't have much of a public profile when it came to politics. But they both adopted pretty Trump-like policies when it came to immigration, crime, opposing China, all the things that you associate with Mr. Trump's policies, they kind of took on board.

So that's how things look for Republicans in Pennsylvania. What about on the Democratic side?

The night for Democrats looked a lot simpler. So the current lieutenant governor, John Fetterman, won the Senate nomination pretty handily. He's this huge 6-foot-9 man with a goatee and tattoos and often is seen wearing a hoodie and shorts walking around in the snow. And so he's sort of the picture of strength. That's the word that I kept hearing from Democratic voters across the state, actually, to refer to him. But he did suffer a stroke over the weekend, uh, and he's been laid up in the hospital. He had to have minor surgery to put in a pacemaker. And he was previously sort of perceived as very progressive. He endorsed Bernie Sanders in his presidential run in 2016. But during the campaign, he's really pivoted to the center and taken on a lot more moderate positions, which puts him very well positioned against the Republican come November. And on the governor's side, Josh Shapiro ran completely unopposed. He's the current attorney general for the state, and he looks very strong.

Okay. And as we said at the top, Pennsylvania was not the only state holding elections. What type of candidates did we see from Democrats in other races?

Even beyond Pennsylvania, you saw Democrats choose more moderate types over more progressive types, especially in kind of these closely contested races that they'll need to win in November if they want to keep control of Congress. So in North Carolina, Democrats selected Cheri Beasley, who's a former state Supreme Court justice, as their nominee for Senate. She's an African American woman with very moderate policy positions, and that's a very good profile for the state.

And Daniel, what does last night portend for November, for the midterms themselves?

Well, Democrats were already facing pretty fierce headwinds. President Joe Biden's approval ratings are deep underwater. The economy isn't doing so well, there's a war in Ukraine, you can expect that the incumbent president's party will suffer in midterm elections anyways, and things aren't looking good. That being said, Democrats are doing themselves some favors by picking moderate candidates with profiles that fit their states, both demographically and on policy terms, while Republicans are a little bit all over the place. They've picked a few good candidates in, in some circumstances and have not blown chances, but in other cases appear to be in danger of picking very extreme candidates like Doug Mastriano in Pennsylvania for the governor's race that could blow eminently winnable races.

All right, Dan, thanks so much for joining us.

Thank you, John.

If this has whetted your appetite, our US politics podcast, Checks and Balances, is going to be keeping an eye on the Pennsylvania Senate race until the midterms in November. The first episode was a few weeks ago, and it's a great primer on why this race is so fascinating and so crucial for the balance of power in Washington. That's in the podcast feed and called "Welcome to Pennsylvania." And look out for more special episodes between now and November.

A desert convoy on one of the most dangerous missions in Afghanistan.

It's trying to find the biggest killer of coalition troops in this war.

The bombs that are taking 3 out of 4 British and American lives.

This patrol— Few weapons have shaped the past 2 decades of conflict like the improvised explosive device, or IED. During wars in Iraq and Afghanistan, thousands of soldiers and civilians lost their lives to these indiscriminate weapons. But in 2006, a new surveillance system called Constant Hawk began to scan the actions of militants from the skies. It was the first time that a technique called wide-area motion imagery had been deployed. It recorded an aerial view of an army base in an area of about 25 square kilometers 24 hours a day and let analysts rewind through footage if anything seemed awry. 15 years later, this technology is now creeping into the civilian world and imperiling privacy.

WAMI, or Wide Area Motion Imagery, is a surveillance technique which was first developed by the military for use in Iraq.

David Hambling writes about technology for The Economist.

Now, advancements in AI automation, machine learning, and new types of sensor mean that WAMI is becoming an increasingly powerful surveillance tool and also an increasingly troubling one.

Okay, so how does it work?

It is basically a load of digital cameras stuck together. The original one developed in 2006 was literally just that, commercial cameras connected via software. So it gives you this reasonably good resolution imagery of a very large area. Basically, it's been compared to a Google Earth view. So you've got the area of a whole city and you can then zoom in and zoom in and look at one particular area of interest. But what's significant about it is that while you may only be looking at one small area at a time, it's actually recording everything that's happening. So afterwards, you can rewind the video and look back and see what was happening through the whole period of observation.

And let's go back through the history here. How, how was this first deployed in Iraq?

They had a huge problem with IEDs, and they simply were not managing to catch people who were planting bombs. The aim was you could record what was happening in a large area. If there was a bomb explosion, you could then rewind the video, find out where the bombers came from, and trace them back to their point of origin. And they were credited with taking out more than 1,000 insurgents. Using this technology and the related systems.

But that was a lot of footage to get through.

There is an insane amount of footage. I mean, you're recording far more video than anyone can go through by hand. And this is the big problem. While you can go through small areas of it, for example, you can use the video to follow a vehicle back and forth through time, there's simply too much there to check through everything to see every possible event of interest. At the moment, they're collecting vast amounts of data and they're simply archiving it without looking at it because there just aren't enough people to do that.

And so that is inevitably where the AI comes in.

Yes, exactly. Because there's far too much data for humans to cope with, this is where we need machines to do the sifting for us. The machine learning system can scan a very large area and then just home in on the few interesting things that might be happening down below. So, for example, when they were looking at vehicle movements in Iraq, they found that the bad guys tended to have some particular patterns of behaviour. So, for example, if you saw a vehicle which kept coming up to a checkpoint and every time it approached reached a checkpoint, it would do a U-turn or veer away. That would be something that would be worth following. And again, if you saw several vehicles driving in convoy, that would be something that was worth following. So rather than having human analysts doing this, this is something that the Whammy system itself has built in, and it will highlight to the operator to say, you may want to have a look at this.

So you mentioned that these Whammy systems are getting more sophisticated, the sensors are getting better. What can the new systems achieve?

The idea here is that you're actually mating the Wham-E with a number of other sensors so they can actually help each other and work cooperatively together. There's one version of the multi-sensor pod which is being used by the US Navy. They're testing at the moment, and that has 3 different cameras in it. It's got a Wham-E sensor, so that's looking over a wide area, and it's got a hyperspectral sensor which can see things that the human eye can't, and then it's got a high-resolution inspection camera. So, for example, if the WAMI sensor spots something that looks to be of interest below, it's only got a certain amount of resolution on that. It's a bit blurry. So, it then cues the high-resolution inspection camera, which can then zoom in. And that's like got a very long zoom lens that can get a very detailed image of what's happening below. And another way in the technology is advancing, as well as getting smarter, WAMI sensors are also getting quite a lot smaller. Originally, it was a fairly substantial piece of kit. It took up all the space on an aircraft or drone. So the originals were like over 600 kilos. The new versions is down to about less than 20 kilos. And there's even a micro version called Micro Kestrel, which weighs just 2.5 kilos. So that only scans an area about 3 square kilometres. But that's something you can put on a small drone, which can keep an eye out over a military base or similar type of area.

Okay, so this was born as an advanced military technology, but what about other applications for it?

There's also versions of this that are being used for things like search and rescue. There's a thing called Vaidar, which is like visual radar. So that uses a wide area video camera with a machine learning system. And the idea is that can scan the surface of the sea and look for life jackets because these orange life jackets stand out against the sea. And the makers, which is a company called Sentient Vision, they reckon it's about 300 times faster at searching than just a human being by eye alone. And they can spot someone with a life jacket even in really rough seas with 6-meter waves.

So in essence, still saving lives. But what about when it is inevitably more widely deployed?

Certainly the people who are developing it would like it to be more widely deployed. And certainly I think we've seen a trend with a lot of these technologies as they become lower cost, more affordable, smaller, they do filter down from the military to security and police services. And there's already been one use where the Baltimore Police Department actually had a Whammy system which was funded by a private donor, which they used for a period to observe things in the city. And that created a certain amount of furore when it was used for monitoring Black Lives Matter protests. In 2020, that became quite contentious.

And is there any way to protect privacy in a world where we have whammies all over the sky?

I mean, in a sense, it's just a creeping ratchet forward of what we've already got because there are satellites which are watching your every move from up above anyway. If there's one thing that Google Earth tells you, it's that there are eyes looking down on you. The concern though here is just in how much detail it's doing it and how much it can then link up. To other sources of information. Like, for example, if you start linking the camera data to mobile phone movements, you can then identify which car a particular mobile phone is in, and then you can track someone on foot going from that car. As with other forms of surveillance, it's the fact that it's building up far, far more detailed view of what's going on on the ground below, and that is potentially very alarming. The only solutions we have to that are legal ones. And ensuring that there are protections in law to prevent this data from being misused. I mean, there's this old saw that the innocent have nothing to fear, but I think certainly the, the way data gathering stories have gone in recent years, that's no longer quite seen as such as a watertight case.

Thanks very much for joining us, David.

Thank you.

One of the classic tastes of the lamentably short British summer, especially among the tennis crowd, is strawberries and cream. This year, that treat is probably going to come at a dearer price. There's a surprisingly strong link between getting those berries to market and the war in Ukraine.

Just over 60% of all the food that's eaten in Britain is produced on British farms. And between May and October, all of the berries in British supermarkets are homegrown.

Rebecca Jackson writes for The Economist.

In recent years, many of the migrants who have come to pick the fruit on British farms have been from Ukraine and Russia. And this year, most of them won't be coming back.

So how does that foreign labor figure into what's going on in Ukraine?

So 4 firms are licensed to recruit foreign workers. Into Britain's farms. And this year, one of those recruitment firms called AG Recruitment, they plan to recruit 50% of their workers from Ukraine and 30% from Russia. Because there is martial law currently in Ukraine, Ukrainian men cannot leave. They have to stay at home and fight. Also because of the war, the British visa application centers in Ukraine, which process the biometric checks that people need in order to apply for visas, have been shut. And so the women who could theoretically come to the UK to work are also not able to apply for visas. So what this means is that AG Recruitment, one of the four firms, isn't recruiting at all in Ukraine. And they also said that they're not recruiting any Russian nationals either.

So how big a deal then is it to the fruit and vegetable industry in a general sense that that supply of Ukrainian and Russian pickers has dried up?

So this is a pretty big deal. The UK kicked off this seasonal worker program after the Second World War. And for a long time, they were recruiting lots of European workers to come pick fruit. And then since Brexit, those European workers haven't been coming back. And so the UK government developed a new scheme to get temporary work visas for other people outside of the EU. So last year, 67% of the seasonal workers that came to Britain were from Ukraine and 8% were from Russia. The fact that these workers aren't coming back means that there's a gaping hole for farmers and they desperately need new workers. And so AG Recruitment told me that they've been going back and trying to recruit from Bulgarians, from Romanians, the sort of European cohort that used to make up a large number of pickers. And they're also looking more farther afield into Central Asia to find new workers. So this is pretty consequential for Britain's farms because the summer fruit industry really does rely on this foreign labor.

So do you think they'll be able to make up the shortfall by looking to those previous sources or indeed in Britain itself?

So Brits really don't want to do this kind of work. It's backbreaking work and even more so it's seasonal work. And Brits tend to want to do work that is full-time and year-round. So far, the recruitment agencies have scraped by, managing to get just enough workers each month. But for example, AG Recruitment still needs 1,000 workers for June. Also, many of this season's pickers are first-timers, and they'll need some training. And while this is an unskilled job technically, it does require some work to get workers up to speed. British Summer Fruits, a berry industry lobbyist, they estimate that newcomers are 30% less productive than returnees.

And so what if this problem essentially doesn't get solved? Doesn't that leave a lot of crops just sitting there to rot?

So it certainly seems like it will be a problem this summer and in summers ahead. A shortage of pickers will probably mean that fruit will be left to rot on trees and bushes. Last year, nearly 8,000 tons of berries, which was worth roughly £36 million, were wasted, and we can expect that number to grow. Nick Marston, who is the chair of British Summer Fruits, has said that he expects that there will be fewer berries on shelves, and that farmers will be producing less in the coming years.

All of this is increasing growers' production costs massively. You know, we have a less productive workforce, and in addition to that, the Home Office this year has attached a minimum wage of £10.10 an hour to the visa scheme compared to the living wage of £9.50. So we're having to pay substantially more to these folks than we would have to pay to a UK resident or an EU resident.

So all of this means that basically the farmers are unlikely to turn a profit this year, and that's pretty consequential for consumers as well. British Summer Fruits estimated that a punnet of raspberries or strawberries will be about 20p more expensive this summer. And if farmers can't turn profit this year, then they also won't be able to invest more in future summers.

Rebecca, thanks very much for joining us.

Thank you.

That's all for this episode of The Intelligence. Let us know what you think of the show. You can get in touch at podcast@ghion.ai. @economist.com.

And you can subscribe to The Economist at economist.com/intelligenceoffer. The link is in the show notes. We'll see you back here tomorrow.