I happen to like longer-form analysis. I subscribe to journals and a few magazines for that reason (and substacks). I also do it to be entertained. Thanks, Noah, for providing analysis and entertainment!
But let’s not try to compare value between the NYT at $25 and a substack at $10 or a bunch of substacks at $65. They are different animals/products. Remember that many people spend $150 a month on Frappuccinos. Consuming is about pleasure (often). The NYT and substackers try to keep their niche audiences pleased.
I am semi-retired and now do geopolitical and economic forecasting for fun and a little cash.. Long-form stuff lets me read other people’s takes and occasionally opens my eyes to data or approaches I hadn’t weighted highly enough. Or sometimes it is just useful to know what certain types of partisans are making excuses for on any given day. There is no such thing as “analysis” without bias and opinion. It is like the old Irish joke about a lost traveler asking for directions and being told “I wouldn’t start from here”. Everyone is starting from the “wrong” place, whether due to institutional history/constraints, partisanship, habit, audience expectations, ignorance etc. The key to being right is to change your assessment/forecast in response to new info. Everyone builds a model of how they think the world or a very micro part of the world behaves or will behave in response to “X” happening. If a model can’t make predictions it is useless. Toss it out. Sometimes people do change their models (or toss them out) in response to new outcomes or data, but most people make excuses and won’t change their models, or would lose their paying audience if they did.
On the subject of editors- they are responsible for maintaining the consistency of a branded product. Substackers are kidding themselves if they don’t think they are creating branded products. Most substackers could also benefit from editing for brevity (me too). It is a reason why I never waste time on podcasts (occasionally skim transcripts). Editors will be hired by the more successful individual substackers and platforms like the free press and pirate wires already use them.
One major problem with the Substack model of journalism is that while it seems excellent for analysis (here I completely agree with you), it doesn’t fund (or intrinsically care about) reporting and data gathering.
More and more of our media is just people expressing opinions or writing analyses from their own home, never having to actually experience the thing they’re talking about. Now, that’s probably fine for a macro econ substack like yours which is focused mainly on the US by a person living in the US (but even in your case, your writing about Japan or Poland massively benefits from having actually been there, and is much better than from people who only read macro stats about these countries!). However, it’s fundamentally problematic for analysis about things like a war in foreign countries written by people who never visited that foreign country and often don’t even speak the language. A person who actually goes to the frontline and speaks to the people directly affected by the war and contributing to it will come up with a very different analysis, which just can’t be replaced by reading OSINT reports or theorising. For example, as a person from central/eastern Europe, reading American analysis of the Russian war with Ukraine, you can immediately tell the difference between people who have actually been there and spoken to some Ukrainians, and the people who didn’t, and the analysis of the latter group is often just frustratingly stupid (pardon me the expression).
One benefit of the “legacy” media was that their business model allowed funding of both reporting and analysis/opinion and by often integrating both under one house guaranteed some basic integrity and accessibility of the reporting and data. I worry that while we’re having better options for reading analysis, we’re drastically losing people who do the actual reporting and investigative journalism, and a big part of it is that it’s just not being funded enough. Because while people (including me) are quite happy to pay for a substack from their favourite analyst (like you), we are much less likely to subscribe to an investigative journalist, war reporter or even a data collector.
As you well know, any analysis is only as good as the data it’s working with. But who will fund it in the new model? The government directly? The government through taxing social media? Analysts themselves through subscriptions to reporters/data collectors? Or?
Newsrooms have already dramatically shrunk, and will continue to do so. Opinion masquerading as News, is much, much cheaper to produce. Or, in the case of much of RW media, to creatively fabricate out of whole cloth. "Infotainment" may further degrade from lack of actual information to just entertainment.
As newsrooms further wither, Substack writers like yourself will increasingly have to rely on increasingly authoritarian national Ministries of Truth for data.
I think it speaks to one weakness many journalists have - self-mythologising. Both the practice of separating news from opinion and the shortness of columns have to do with the physical constraints of printing a newspaper. As does the editorial layers, which make more sense when you have one daily deadline and print schedule. But despite most of their business being online now, the basic principles have become sacrosanct.
My pet theory is that the mainstream press doesn't champion analysis because institutionally, they don't have the skills needed to support and understand analysis. Sure, they hire some very capable people that can do analysis, but institutionally, they just don't have the DNA or infrastructure needed to support analysis.
Maybe I've got this totally wrong, but I wonder what percent of NYT writers and editors know how to use Excel or have even a basic understanding of statistics. By use Excel, I mean could create a relatively complicated spreadsheet and use it to analyze something, not just scroll through a spreadsheet someone else created.
Most people don’t want analysis- it is not what newspapers are selling. There is a certain type of person who think better about themselves as “intellectuals” for reading what they think is analysis - as long as the outcome conforms to their views. Those types subscribed to the New Yorker, the Atlantic, Foreign Affairs. Now there is substack- more niches. More topics. A good thing.
Foreign Affairs' articles are often pretty wonky, with impressive footnote counts. But yes, even those often just pander to the siren song of confirmation bias.
But does the rise of Substacks like Noah's and Krugman's presage the cannibalization and eventual demise of legacy media? Will the only media choices in a decade or two be Substack...and FOX News?
Murdoch said decades ago that he believes that all US media will eventually consolidate into two outlets. And that he intends FOX News to be one of them.
100%! I've often wondered if journalists have enough experience and perspective to know who is telling the truth and who is bullshitting, which is a key part of why I would pay for their work.
The reality is that the best and the brightest are not heading into journalism... And their target audience is not the best and brightest either. Ask any specialist about literally anything in their realm of expertise and they'll talk about how the media has it all wrong. Finance, tech, science -- our journalists have no ability to cover these things seriously. We are only getting the pop culture veneer of what's actually going on. Or we are getting 800 words of infotainment that's only somewhat attached reality.
You're definitely right. When an article has been written in which I've been consulted as an expert, it's been clear almost every time that the journalist only had a very casual understanding of the subject, not deep enough to do serious analysis.
Just to clarify, I believe journalists are plenty smart enough to use Excel. As I understand it, journalists at mainstream outlets come from top schools at a much higher rate than other professions and I'd guess they're likely have very high IQs on average.
My contention is more that journalism seems to select for wordcels versus people more interested in and experienced with analysis type work (and Excel).
Agreed. I know a fair few journalists and although the plural of anecdote is not data (ha ha) they seem more than capable of learning Excel to a proficient level. At least no less capable than me, and I did it despite not having a STEM background. I think it’s instead because it doesn’t fit in with the product they’ve traditionally been selling and, like many big institutions, they’re slow to adapt.
I have a very different experience. I was a Substack author first and subsequently started writing for Bloomberg Opinion as well (where my pieces are principally analysis not opinion!) Substack allows more flexibility, but having never had an editor before, I value the input they provide at Bloomberg and the word limit fosters discipline. With Bloomberg specifically there is also the benefit of a professional audience. One thing I have noticed though is that companies I write about are much more sensitive to Bloomberg comment than Substack comment, even though Substack readership is higher. The PR industry has yet to catch up.
In my experience there is a huge age divide between perception of the media. People over 50ish have a massive bias towards legacy media institutions which completely disappears among younger people. And guess which age group the higher ups tend to belong to.
It's also a (not unrelated) respect for institutions. If I write it on Bloomberg, then as far as the company I'm writing about is concerned, 'Bloomberg says...' If I write on my Substack, then 'some guy on Substack says...' PR people probably don't have the resources to manage the proliferation of media outlets yet either.
I qualify as over 50 years old. In turn, my contempt for the legacy media continues to deepen despite my purchasing some of their products. Gradually, my respect for Substack products has increased over time.
This piece is itself analysis, of course, and it illustrates a technique that I think should be a part of every argument for change. You present the history of how the old way came to be and, crucially, why it had advantages at the time. Then you explain what specifically has changed since then that reduces or eliminates those erstwhile advantages, and how a new idea fits the new circumstances better.
I wish every argument for social change could include this element.
Yes to more analysis. I would add that, while editors at legacy media companies often add value to contributors, the gate-keeping behavior is grating, adds friction, and slows things down. It’s much easier to go direct and build your own audience, using outside freelance editors as needed.
Legacy media outlets using outside freelance editors would save money. This change would also complement the layoffs of journalists, dumping of office space, and the shrinking of printed product.
> Now here’s the thing — news readers want analysis. They don’t want to just read the reporting from the Associated Press or some investigative journalists and then have to put it all together and decide what it all means.
Thank you. That’s exactly why I read this blog. And accidental rabbits or travel tips, obviously 😁
You have provided superb analysis that is data driven.
I find the biggest faults with predominantly Conservative pundits who cherry pick data, use a convenient swiss cheese of data and have biased boundary conditions.
Generally, non technical "national authorities", again predominantly Conservatives, have 2 other ultra biased flaws;
A. Very poor assessment. The foundation of analysis.
B. Poor root cause analysis as they choose to opine rather than derive the deeper causes
Thank you for a useful post. I do like Substack but realize that I have to do my own curation and be careful about it. There are lots of writers who I would like to follow but there are two major constraints: time & money (are not these the constraints for a lot of other things as well?). Some writers offer free Substacks such as Krugman (the exception being his weekend writings) while the remainder have yearly subscription fees. Those fees can add up pretty quickly. I'm not picking on Noah, who I have followed for years since his blogging days at Michigan, but reading posts such as today's takes time and thought. When you multiply this by say 10-15 writers, that's a lot of time out of the day. It's the same with podcasting!! How many podcasts can one listen to each day. I guess if one were a hermit this would all be OK but then we would sacrifice community at the Substack and Podcast altar.
One further problem with Substack is the lack of an introductory trial period (or if there is one I have not found it). One needs to subscribe in order to get all the posts and that payment is locked in. I would rather see a 5-10 "issue" intro sub before your card is charged. If the content is no good, you just opt out. There are three writers who I've cancelled over the past three years (each was a $100/year subscription) because I grew tired of the same old stuff they were writing. Right now I maintain about a dozen subscriptions, a few of which are free or low cost ($50/year). That's about all that I can deal with given everything else going on. Content providers such as Noah and others need to be cognizant of their readers in this respect.
Interesting. Now that I think about it, one of the reasons I mistrust op-ed a bit more is that many times they seem to start with the opinion they’re trying to support and then provide cherry picked evidence and motivated reasoning to support it. Not all but it feels that way sometimes. And now learn this is actually forced.
When you start with the “analysis “ and the process for how to think about X, it allows for more nuance and allows me to add it to my epistemic model. An analysis can still support different conclusions or “opinions.” I believe this is missing from our public discourse.
With respect to the “data journalism” point, you can’t not mention John Burn-Murdoch at the FT. He creates conversations with his fresh data analysis and somehow still manages to be succinct. I think he is the best in the biz (and no, I am not John).
"Analysis" to me used to be the domain of the weekly/monthly news magazine; your Times, Newsweeks, Economists etc. Now they're a dying breed, but it sounds like they had a different newsroom culture than a daily paper.
re: length of articles, most of the substacks I subscribe to have a sweet spot of 2000 - 2500 words. This one is right in there at 2600 and flowed pretty well. Longer than that and honestly my eyes start to glaze over.
I wonder what you think of Ezra Klein's work at the NYTimes. He seems to get it right, in my view, and has his own voice and audience. Is that because he came to it from independent blogging and podcasting, as opposed to, say, Krugman, who came to op-ed from a professional economics background?
Everyone thinks their opinions are rooted in solid analysis. (Yours usually are, Noah.) But that makes the line between the two inherently blurry and often invisible.
If analysis involves making recommendations (which everyone wants), it's going to have to have some opinion mixed in. No getting around that. But the opinion is the cheap throwaway part...Anyone can say "Trump sux!!", and lots of people can find more eloquent ways to say that, but most can't do the analysis to back it up... ;-)
I happen to like longer-form analysis. I subscribe to journals and a few magazines for that reason (and substacks). I also do it to be entertained. Thanks, Noah, for providing analysis and entertainment!
But let’s not try to compare value between the NYT at $25 and a substack at $10 or a bunch of substacks at $65. They are different animals/products. Remember that many people spend $150 a month on Frappuccinos. Consuming is about pleasure (often). The NYT and substackers try to keep their niche audiences pleased.
I am semi-retired and now do geopolitical and economic forecasting for fun and a little cash.. Long-form stuff lets me read other people’s takes and occasionally opens my eyes to data or approaches I hadn’t weighted highly enough. Or sometimes it is just useful to know what certain types of partisans are making excuses for on any given day. There is no such thing as “analysis” without bias and opinion. It is like the old Irish joke about a lost traveler asking for directions and being told “I wouldn’t start from here”. Everyone is starting from the “wrong” place, whether due to institutional history/constraints, partisanship, habit, audience expectations, ignorance etc. The key to being right is to change your assessment/forecast in response to new info. Everyone builds a model of how they think the world or a very micro part of the world behaves or will behave in response to “X” happening. If a model can’t make predictions it is useless. Toss it out. Sometimes people do change their models (or toss them out) in response to new outcomes or data, but most people make excuses and won’t change their models, or would lose their paying audience if they did.
On the subject of editors- they are responsible for maintaining the consistency of a branded product. Substackers are kidding themselves if they don’t think they are creating branded products. Most substackers could also benefit from editing for brevity (me too). It is a reason why I never waste time on podcasts (occasionally skim transcripts). Editors will be hired by the more successful individual substackers and platforms like the free press and pirate wires already use them.
One major problem with the Substack model of journalism is that while it seems excellent for analysis (here I completely agree with you), it doesn’t fund (or intrinsically care about) reporting and data gathering.
More and more of our media is just people expressing opinions or writing analyses from their own home, never having to actually experience the thing they’re talking about. Now, that’s probably fine for a macro econ substack like yours which is focused mainly on the US by a person living in the US (but even in your case, your writing about Japan or Poland massively benefits from having actually been there, and is much better than from people who only read macro stats about these countries!). However, it’s fundamentally problematic for analysis about things like a war in foreign countries written by people who never visited that foreign country and often don’t even speak the language. A person who actually goes to the frontline and speaks to the people directly affected by the war and contributing to it will come up with a very different analysis, which just can’t be replaced by reading OSINT reports or theorising. For example, as a person from central/eastern Europe, reading American analysis of the Russian war with Ukraine, you can immediately tell the difference between people who have actually been there and spoken to some Ukrainians, and the people who didn’t, and the analysis of the latter group is often just frustratingly stupid (pardon me the expression).
One benefit of the “legacy” media was that their business model allowed funding of both reporting and analysis/opinion and by often integrating both under one house guaranteed some basic integrity and accessibility of the reporting and data. I worry that while we’re having better options for reading analysis, we’re drastically losing people who do the actual reporting and investigative journalism, and a big part of it is that it’s just not being funded enough. Because while people (including me) are quite happy to pay for a substack from their favourite analyst (like you), we are much less likely to subscribe to an investigative journalist, war reporter or even a data collector.
As you well know, any analysis is only as good as the data it’s working with. But who will fund it in the new model? The government directly? The government through taxing social media? Analysts themselves through subscriptions to reporters/data collectors? Or?
Yep, this is an important question, and I don't have a good answer!
You need a United Artists of Like The Economist to be big enough to generate the information for the Analysts to work with? Is that AI’s place.
Newsrooms have already dramatically shrunk, and will continue to do so. Opinion masquerading as News, is much, much cheaper to produce. Or, in the case of much of RW media, to creatively fabricate out of whole cloth. "Infotainment" may further degrade from lack of actual information to just entertainment.
As newsrooms further wither, Substack writers like yourself will increasingly have to rely on increasingly authoritarian national Ministries of Truth for data.
I think it speaks to one weakness many journalists have - self-mythologising. Both the practice of separating news from opinion and the shortness of columns have to do with the physical constraints of printing a newspaper. As does the editorial layers, which make more sense when you have one daily deadline and print schedule. But despite most of their business being online now, the basic principles have become sacrosanct.
My pet theory is that the mainstream press doesn't champion analysis because institutionally, they don't have the skills needed to support and understand analysis. Sure, they hire some very capable people that can do analysis, but institutionally, they just don't have the DNA or infrastructure needed to support analysis.
Maybe I've got this totally wrong, but I wonder what percent of NYT writers and editors know how to use Excel or have even a basic understanding of statistics. By use Excel, I mean could create a relatively complicated spreadsheet and use it to analyze something, not just scroll through a spreadsheet someone else created.
Most people don’t want analysis- it is not what newspapers are selling. There is a certain type of person who think better about themselves as “intellectuals” for reading what they think is analysis - as long as the outcome conforms to their views. Those types subscribed to the New Yorker, the Atlantic, Foreign Affairs. Now there is substack- more niches. More topics. A good thing.
Foreign Affairs' articles are often pretty wonky, with impressive footnote counts. But yes, even those often just pander to the siren song of confirmation bias.
But does the rise of Substacks like Noah's and Krugman's presage the cannibalization and eventual demise of legacy media? Will the only media choices in a decade or two be Substack...and FOX News?
Murdoch said decades ago that he believes that all US media will eventually consolidate into two outlets. And that he intends FOX News to be one of them.
Thanks for your salient thoughts.
100%! I've often wondered if journalists have enough experience and perspective to know who is telling the truth and who is bullshitting, which is a key part of why I would pay for their work.
The reality is that the best and the brightest are not heading into journalism... And their target audience is not the best and brightest either. Ask any specialist about literally anything in their realm of expertise and they'll talk about how the media has it all wrong. Finance, tech, science -- our journalists have no ability to cover these things seriously. We are only getting the pop culture veneer of what's actually going on. Or we are getting 800 words of infotainment that's only somewhat attached reality.
You're definitely right. When an article has been written in which I've been consulted as an expert, it's been clear almost every time that the journalist only had a very casual understanding of the subject, not deep enough to do serious analysis.
Just to clarify, I believe journalists are plenty smart enough to use Excel. As I understand it, journalists at mainstream outlets come from top schools at a much higher rate than other professions and I'd guess they're likely have very high IQs on average.
My contention is more that journalism seems to select for wordcels versus people more interested in and experienced with analysis type work (and Excel).
I don't think it's a question of intellect, but of sufficient familiarity and understanding of the subject matter.
Agreed. I know a fair few journalists and although the plural of anecdote is not data (ha ha) they seem more than capable of learning Excel to a proficient level. At least no less capable than me, and I did it despite not having a STEM background. I think it’s instead because it doesn’t fit in with the product they’ve traditionally been selling and, like many big institutions, they’re slow to adapt.
I have a very different experience. I was a Substack author first and subsequently started writing for Bloomberg Opinion as well (where my pieces are principally analysis not opinion!) Substack allows more flexibility, but having never had an editor before, I value the input they provide at Bloomberg and the word limit fosters discipline. With Bloomberg specifically there is also the benefit of a professional audience. One thing I have noticed though is that companies I write about are much more sensitive to Bloomberg comment than Substack comment, even though Substack readership is higher. The PR industry has yet to catch up.
In my experience there is a huge age divide between perception of the media. People over 50ish have a massive bias towards legacy media institutions which completely disappears among younger people. And guess which age group the higher ups tend to belong to.
It's also a (not unrelated) respect for institutions. If I write it on Bloomberg, then as far as the company I'm writing about is concerned, 'Bloomberg says...' If I write on my Substack, then 'some guy on Substack says...' PR people probably don't have the resources to manage the proliferation of media outlets yet either.
I qualify as over 50 years old. In turn, my contempt for the legacy media continues to deepen despite my purchasing some of their products. Gradually, my respect for Substack products has increased over time.
Amen, brother!
This piece is itself analysis, of course, and it illustrates a technique that I think should be a part of every argument for change. You present the history of how the old way came to be and, crucially, why it had advantages at the time. Then you explain what specifically has changed since then that reduces or eliminates those erstwhile advantages, and how a new idea fits the new circumstances better.
I wish every argument for social change could include this element.
Yes to more analysis. I would add that, while editors at legacy media companies often add value to contributors, the gate-keeping behavior is grating, adds friction, and slows things down. It’s much easier to go direct and build your own audience, using outside freelance editors as needed.
Legacy media outlets using outside freelance editors would save money. This change would also complement the layoffs of journalists, dumping of office space, and the shrinking of printed product.
> Now here’s the thing — news readers want analysis. They don’t want to just read the reporting from the Associated Press or some investigative journalists and then have to put it all together and decide what it all means.
Thank you. That’s exactly why I read this blog. And accidental rabbits or travel tips, obviously 😁
I certainly wouldn't mind an occasional cute bunny picture. 😁
You have provided superb analysis that is data driven.
I find the biggest faults with predominantly Conservative pundits who cherry pick data, use a convenient swiss cheese of data and have biased boundary conditions.
Generally, non technical "national authorities", again predominantly Conservatives, have 2 other ultra biased flaws;
A. Very poor assessment. The foundation of analysis.
B. Poor root cause analysis as they choose to opine rather than derive the deeper causes
Thank you for a useful post. I do like Substack but realize that I have to do my own curation and be careful about it. There are lots of writers who I would like to follow but there are two major constraints: time & money (are not these the constraints for a lot of other things as well?). Some writers offer free Substacks such as Krugman (the exception being his weekend writings) while the remainder have yearly subscription fees. Those fees can add up pretty quickly. I'm not picking on Noah, who I have followed for years since his blogging days at Michigan, but reading posts such as today's takes time and thought. When you multiply this by say 10-15 writers, that's a lot of time out of the day. It's the same with podcasting!! How many podcasts can one listen to each day. I guess if one were a hermit this would all be OK but then we would sacrifice community at the Substack and Podcast altar.
One further problem with Substack is the lack of an introductory trial period (or if there is one I have not found it). One needs to subscribe in order to get all the posts and that payment is locked in. I would rather see a 5-10 "issue" intro sub before your card is charged. If the content is no good, you just opt out. There are three writers who I've cancelled over the past three years (each was a $100/year subscription) because I grew tired of the same old stuff they were writing. Right now I maintain about a dozen subscriptions, a few of which are free or low cost ($50/year). That's about all that I can deal with given everything else going on. Content providers such as Noah and others need to be cognizant of their readers in this respect.
Interesting. Now that I think about it, one of the reasons I mistrust op-ed a bit more is that many times they seem to start with the opinion they’re trying to support and then provide cherry picked evidence and motivated reasoning to support it. Not all but it feels that way sometimes. And now learn this is actually forced.
When you start with the “analysis “ and the process for how to think about X, it allows for more nuance and allows me to add it to my epistemic model. An analysis can still support different conclusions or “opinions.” I believe this is missing from our public discourse.
This is exactly why I read you.
With respect to the “data journalism” point, you can’t not mention John Burn-Murdoch at the FT. He creates conversations with his fresh data analysis and somehow still manages to be succinct. I think he is the best in the biz (and no, I am not John).
I agree he's the best, and if you were him, it would be OK, since he is my friend.
"Analysis" to me used to be the domain of the weekly/monthly news magazine; your Times, Newsweeks, Economists etc. Now they're a dying breed, but it sounds like they had a different newsroom culture than a daily paper.
re: length of articles, most of the substacks I subscribe to have a sweet spot of 2000 - 2500 words. This one is right in there at 2600 and flowed pretty well. Longer than that and honestly my eyes start to glaze over.
I wonder what you think of Ezra Klein's work at the NYTimes. He seems to get it right, in my view, and has his own voice and audience. Is that because he came to it from independent blogging and podcasting, as opposed to, say, Krugman, who came to op-ed from a professional economics background?
Ezra is great, but I do think his format constrains him.
Over a decade ago I LOVED the geeky articles by Wall Street Journal's "The Numbers Guy" Carl Bialik.
Apparently he's working at Yelp now.
Everyone thinks their opinions are rooted in solid analysis. (Yours usually are, Noah.) But that makes the line between the two inherently blurry and often invisible.
If analysis involves making recommendations (which everyone wants), it's going to have to have some opinion mixed in. No getting around that. But the opinion is the cheap throwaway part...Anyone can say "Trump sux!!", and lots of people can find more eloquent ways to say that, but most can't do the analysis to back it up... ;-)
Just throw in a few solid stock picks now and then, and you can devour a huge chunk of Bloomberg's market share!
Correct.