Yves here. We’ll take some of the fun out of using this article as a reader critical thinking exercise by pointing out the obvious up front. This piece is garbage in, garbage out. Or perhaps more accurately, (not surprising for a piece about capitalism) it falls into the “drunk looking under the streetlight” category by analyzing readily available data, as opposed to squishier metrics that would better represent changes in news production and consumption over time.
The underlying study mined only at articles in large, mainstream publications. This approach assumes continuity when that is not at all the case. It ignores the rise of social and independent media and the continuing plunge of trust in mainstream media outlets. For instance, even here in Southeast Asia, when I run into expats who ask about me and I describe the site, most snort that they don’t trust the news and consume it less than they once did.
In addition, it ignores the large shift in the class composition of newsrooms over time. As we have mentioned often, save perhaps at the New York Times, reporters generally came from working class backgrounds. That gave most a native skepticism of taking what business and political leaders said at face value. But now, the surviving institutions have a heavy representation of graduates of Ivies or equivalent schools who want to be welcome at the “right” parties. On top of that, the loss of classified ad revenues has hollowed out newsrooms, making the press more dependent on “access” stories. That gives the powerful yet another leverage point over the media, since they can freeze reporters who aren’t sufficiently compliant out of the information tidbits they spread around.
Specifically, it seems difficult to believe that the perception of socialism in the US remained flattish over the time of Sanders’ two campaigns. It was at least somewhat reported that being a socialist had become popular among teens and young adults. But then again, that cohort is too young to be a big target for ad dollars, so major publications were unlikely to shift their messaging to increase appeal to this cohort.
To put it another way: the father of modern public relations, Eddie Bernays, determined in 1926 that half the stories on the front page of the New York Times were propaganda. What would you guess the proportion is now? And what do non-US readers guesstimate the shifts in sentiment
By Jay L. Zagorsky, Associate Professor of Markets, Public Policy and Law, Boston University, and H. Sami Karaca, Professor of Business Analytics, Boston University. Originally published at The Conversation
Capitalism, communism and socialism are the world’s three major economic systems. While the phrase “economic system” may seem like a yawn, countless people have fought and died in major wars over which one should dominate.
Shifts from one system to another, like the 1989 fall of communism in much of Eastern Europe, changed the lives of millions. And while researchers know that a country’s economic system dramatically impacts people’s living standards, less is known about how attitudes toward these systems have changed over time.
We are professors working at Boston University’s new Ravi K. Mehrotra Institute, which is trying to understand how business, markets and society interact. Given many recent criticisms of capitalism, we were surprised to find positive sentiment toward capitalism is slowly rising over time.
The Main Economic Systems Explained
Capitalism, communism and socialism are economic and political systems that differ in their principles and organization. Capitalism emphasizes the private ownership of resources and the means of production, driven by profit and market competition, with minimal government intervention.
Communism, on the other hand, advocates for a classless society where all property is communally owned. In communism, wealth is distributed according to need and there is no private ownership, which aims to eliminate inequality and oppression.
Socialism falls between these extremes. It focuses on the collective or state ownership of key industries and resources. This allows for some private enterprise, with the aim of reducing inequality through social welfare programs and obtaining a more equitable distribution of wealth.
Modern economies blend capitalism with socialism to address challenges like inequality, market failures and negative externalities, like when a business harms the environment. Governments intervene through regulations, welfare programs and public services to tackle issues like pollution and income inequality. This creates what economists call a “mixed economy.”
The amount of state involvement varies from country to country. At one end is market capitalism, where markets dominate with a limited government role. The U.S. is one such example.
At the other end is state capitalism, like in China, where the government directs economic activity while incorporating market elements. The goal is to combine market efficiency and innovation with measures to contain capitalism’s social and economic costs.
How to Measure People’s Attitudes Toward Economic Systems
Some surveys have asked people directly how they feel about these systems.
For example, the Pew Research organization’s most recent survey on the issue found the proportion of Americans with positive views of either capitalism or socialism has declined slightly since 2019, with capitalism remaining more popular overall. Nevertheless, Americans are split sharply along partisan lines. About three-quarters of Republican voters have positive views of capitalism, compared with less than half of Democratic voters.
Unfortunately, there are no long-running surveys tracking people’s feelings toward the three systems. Because of this shortcoming, we used artificial intelligence to analyze references to the three systems in more than 400,000 newspaper articles published over a span of decades.
We identified every news story that discussed capitalism, communism or socialism using ProQuest’s TDM Studio. ProQuest has digitized almost all the articles in major English-language newspapers – including The Wall Street Journal and The New York Times – starting in the mid-1970s, with partial archives from earlier years.
The AI model was designed to assess the tone of each article across several dimensions, including anger, surprise and happiness. After the model scored each article on those qualities, we combined the emotions into three categories: positive, negative, and neutral or unknown. For example, an article discussing capitalism might be rated as 60% positive, 20% negative and 20% neutral.
Using an AI large language model allowed us to track shifts in press attitudes over time – which, to be fair, might not match popular opinion.
How Views Have Changed Since the 1940s
When we looked at newspaper articles from the end of World War II to the present, we found something unexpected. In the 1940s, capitalism was not well regarded. The average article containing “capitalism” or “capitalist” got a 43% negative and 25% positive sentiment score. This is surprising, since we looked at newspapers published primarily in countries with capitalist systems.
However, just because capitalism didn’t get a high positive score doesn’t mean that newspaper writers loved communism or socialism. In the 1940s, articles with those words also got relatively high negative scores: 47% on average for articles containing “communism” or “communist,” and a 46% negative rating for “socialism” and “socialist.”
Since that time, however, positive sentiment toward capitalism has improved. In the 2020s, the average article with capitalism got a more balanced 37% negative and 34% positive sentiment score. While capitalism clearly isn’t loved in the press, it’s also not disparaged as much as it was just after World War II.
The news media’s attitudes toward capitalism improved more than attitudes toward socialism or communism over time. In the 1960s, positive attitudes toward all three were roughly the same. Today, however, positive sentiment toward capitalism is 4 or 5 percentage points higher than the other two. The climb wasn’t steady, since the number of favorable articles about capitalism fell during recession years.
Still, some contemporary commentators fret that capitalism is in crisis.
Not long ago, The New York Times – a newspaper located in the world’s financial center – ran an op-ed headlined “How Capitalism Went Off the Rails.” A recent book review in The Wall Street Journal, a newspaper that is a bastion of capitalism, starts, “Our universities teach that we are living the End of Times of ‘late capitalism.’”
But while capitalism clearly isn’t beloved by all, we didn’t find evidence that it’s being overtaken by socialism or communism. Instead, using AI to process the attitudes reflected in thousands of newspaper articles, we found that people – or at least the press – are slowly warming to it.
I’d have to agree that much of the shifts in the sentiment measured in media reflects an increasing amount of propaganda whether it be by government influence or by its ever-increasing ownership concentration by extremely wealthy persons and corporations.
The other issue I’d have with it is that it does a poor job of defining exactly what they mean by communism, socialism, and capitalism and how that went into their measurements. I’d suspect if you asked the average American what they thought about capitalism they’d knee jerk provide a positive comment because of its basic framework of private ownership, however if you got them to talk about all of the massive corporations and their power and rent seeking behaviors they’d give you a much more negative take on capitalism from that perspective.
Memory is notoriously unreliable, but I recall a large advertisement, probably in “Life” magazine or “Ladies’ Home Journal,” which my grandmother subscribed to, and which I consumed voraciously whenever I visited her. It would have been in the 1950’s or late 1940’s, and probably around Thanksgiving. It pointed out that, while the Pilgrims’ original communities were Socialist, they never really took off until they embraced Capitalism. Enough of this sharing of the harvest with the local Indigenous families. Time to wipe ’em out and take over that land!
Love the intro by Mr. Smith…
put this into Columbias’s first seminar:
” the loss of classified ad revenues has hollowed out newsrooms, making the press more dependent on “access” stories. That gives the powerful yet another leverage point over the media, since they can freeze reporters who aren’t sufficiently compliant out of the information tidbits they spread around.”
As German legacy papers are concerned – it by now has become common knowledge that by various degress they all have tried to create two versions of their papers – one behind the paywall, one for free.
Depending on the paper, the free versions are the capitalist feel-good image of Germany/Europe/the progressive globe. All is good. There might be some fires – real ones and imagined – but things are potentially optimistic – even though the content of and cause for most articles is that things are looking bad.
So in order to satisfy the product-y glossy look, “feel” and “flavour” of the paper you purport one message while giving insight into the opposite – but you do so with an eventual positive spin in the end of most texts. Which is mainly due to one wisdom: capitalism WILL offer a solution.
The alternative is non-existent. And more than anything else, positive design is the embodiment of this fundamental insight. Of course mixing product placement in as WaPo and NYT had first introduced it with their cooking, travel and life-style obsessions.
p.s. As the first daily (and first ever left German paper) – TAZ – now by a high percentage a state PR paper (GREENS) – with horrible articles glossing over genocide – but still some decent reporting among the trash (ratio bad vs. good – 70/30 80/20?) – has announced to phase out print by Oct. 2025 I believe.
Giving up print eventually for the capitalist thrust, faith and look will only be beneficial.
Is this research part of the ad / marketing campaign for capitalism? I remember reading about some agencies getting involved with some kind of capitalism marketing campaign. This was after some polling on youth’s views on socialism.
These people are idiots. They pretend to measure the popularity of an idea by trying to measure what an unspecified group of journalists say about It. Not the same.
Not even what they say but merely the “tone” of their language. And this vacuousness removed yet an extra degree away from reality by the moderator being AI. Who are these guys kidding? Possibly only another AI Tone Survey. Cool. Then they can settle any disagreement with a big AI Tone Debate, having eliminated any actual meaning in the first “analysis.”
I noticed that the authors “are professors working at Boston University’s new Ravi K. Mehrotra Institute…” Here is a description:
“Boston University celebrated the launch of the Ravi K. Mehrotra Institute for Business, Markets & Society at Questrom School of Business on Thursday, September 19, 2024. Made possible by a landmark gift from Ravi K. Mehrotra, the mission of the Institute is to help students, industry leaders, regulators, and the broader public understand and appreciate the role business and markets do, can, and should play in creating lasting prosperity, advancing societal goals, and solving global challenges.”
https://www.businesswire.com/news/home/20240919399506/en/Boston-University-Commemorates-Launch-of-the-Ravi-K.-Mehrotra-Institute-for-Business-Markets-Society-at-Questrom-School-of-Business
Sounds a lot like the type of “research institute” the Koch brothers and other billionaires have been busy funding over the last many decades. Apparently one of the first tasks for its researchers was to show us that people really like capitalism after all. We can ignore internet comments about murdered insurance executives and such; this is AI-driven Science!
Mehrotra himself is apparently “self made,” working his way up in shipping in India, Iran, and globally. His company, Foresight Ltd., is based in London. I’m not sure what his connections are to Boston U.; maybe all the better business schools were taken. But I’m sure he really believes in “markets,” just like Peter Thiel, Elon Musk, et al. The billionaires will lead us!
“Capitalism, communism and socialism are economic and political systems that differ in their principles and organization”
Should more ‘ism(s) be introduced into the reductionist handling above..
Maybe… financialism, rentierism, neo-liberalism, loan-sharkism ??
I don’t know, I guess I should just zip-it, sit back and, let myself drift into a semi-unconscious and blissful ignorance while AI, the Alphas with their humility and generosity. take care of my comfort and needs as they deem fit.
Exactly. What about financial capitalism vis à vis industrial capitalism as Michael Hudson describes. I’m more interested in the comments about this ‘research paper’. I stopped reading this ‘paper’ after that first sentence. Complete rubbish.
Where’s the chart showing the uptick in positive sentiment towards capitalism closely tracking the uptick in media ownership by wealthy individuals and corporations? I’m willing to bet it exists somewhere. Journalists singing for their supoer by pandering to their capitalist owners’ sensitivities can hardly be considered instructive vis a vis a shift in sentiment towards capitalism.
Since, imo, Capitalism without conscience is killing our Climate and thus killing our species, this is a dialogue we all need to honestly foreground daily.
Despite the study’s shortcomings I’m glad they presented it. I am buoyed that “Today however, positive sentiment toward Capitalism is 4/5% higher than the other two.”
Could this also be viewed that of the three economic systems considered, 2 of the 3, in some form, would be preferred to Capitalism?
I do wonder if the AI could have a bias though, in its mining of articles.
In addition to the dataset issues and the center’s ideological leanings, the methodology seems fundamentally misguided because it doesn’t understand how written arguments work, or any writing for that matter. The AI model, according to the ProQuest TDM Studio link, scores individual sentences on an affective scale (“We then train a model using the sentence embeddings to predict the probability of each sentence being assigned to each affective state [i.e. ‘Anger’, ‘Disgust’, ‘Fear’, ‘Sadness’, ‘Happiness’, ‘Love’, ‘Surprise’, ‘Neutral’, ‘Other’])”. If the study authors then simply tally up the percentages, that doesn’t mean that an article depicts capitalism or anything else negatively or positively. One would assume that even an article advocating “kill all bosses” would start with some neutral-to-objective presentation in order to make its case. But then again, maybe the real thesis here is that there can be no arguments, that all writing is only vague immediate sentiments and impulses (or a series of like buttons). And that too might be its propagandistic import: that humans should read as if each sentence is a slogan, nothing more than an immediate emotional button-pusher.
this nonsense is Ivy Lee on steroids
Ah, definitions! Re “Capitalism emphasizes the private ownership of resources and the means of production, driven by profit and market competition, with minimal government intervention.” WTF? “Minimal government intervention?” Part of the definition of capitalism? Sez who?
Conservatives laud themselves by listing the functions of government that are appropriate: things like providing a stable currency, a legal system for enforcing contracts, etc. So, where is the line drawn, between “just enough government regulation” and “too much government regulation”? There has to be a line because capitalism doesn’t work without government regulation/intervention.
It’s not quite garbage. It’s probably a fairly accurate assessment of sentiment among journalists and editorial boards for English language mainstream media (although the AI ‘sentiment’ measure is very fluffy and typically boils down to saying nice things or speaking positively, and doesn’t include any kind of critical analysis). The problem is, as Yves points out, these are no longer a reliable proxy for sentiment among the population as a whole. The ‘English’ restriction will also skew things since there are relatively few English language publications in communist countries (maybe RT and SCMP, but we don’t know whether they met the ‘mainstream’ definition).
It’s at odds with some other recent findings, like Gen Z having a more favorable view of communism than capitalism. We could explain this by suggesting that mainstream media doesn’t accurately represent the views of Gen Z, and I think there’s plenty of other supporting evidence for that.
The author notes this briefly:
but does not make any attempt to measure this discrepancy or check whether it’s changed over time.
We might reasonably conclude that press coverage of capitalism, communism etc. is not meaningfully more or less favorable than it was (say) 30 years ago. 1940 is a more interesting data point to me as it suggests that all systems were at a low ebb in trust back then. However, this is based only on ‘partial archives’ and the typical writing style of newspaper articles has changed quite markedly since that time, so it could simply be that that LLM reads the older ones as less positive in general. That’s yet another potential confounding factor that would need controlling.