> There are already signs that startups may not spread particularly well. The spread of startups seems to be proceeding slower than the spread of the Industrial Revolution
Government was much smaller at the time of the Industrial Revolution.
The theory of "regulatory capture" says that government is responsive to those firms that are affected by government regulation and have the money to buy favorable legislation.
Thus, it's much easier for big-dumb-and-slow firms to use government to put roadblocks in the way of startups than it was for inefficient firms to put roadblocks in the way of factory production 200 years ago.
Want to sell wine over the Internet? The established players will make sure that it's illegal for you to do so.
Want to cut a deal with your early stage employees where they get no pay for 6 months? It's illegal in most places to not pay someone a wage. In MA it's illegal to fire someone for blatant incompetence and they pay them their final paycheck on the regularly scheduled payroll 4 days later - you need to hand them a physical check then and there.
Want to buy a bunch of servers? In a lot of locales (MA, again), you have to pay taxes on the assets every year.
Want to sell real estate on the Internet? Hope you're a licensed broker!
Want to write an expert system that generates legal documents? You'd better have passed the bar exam. Actually, wait, even though that was good enough for Lincoln, these days you need to have actually ATTENDED lay school - a mere demonstration of proficiency is not enough.
Etc., etc., etc.
If this is how things are in the free-wheeling US, I imagine that it's far worse in Europe.
I think that people across the globe have what it takes to launch startups. I just think that the legal culture makes it very hard in most places.
Government was much smaller at the time of the Industrial Revolution.
Is this true? Is it really such an open-and-shut case?
In some senses it is obviously true: There were a lot fewer people, a lot fewer filing cabinets and a lot fewer regulations. A lot of the tools needed for our big modern system -- hygiene, communications, literacy -- just weren't there yet. There were essentially no worker protections in the early stages of the Industrial Revolution; there were few environmental regulations. And travel and communications were harder, which made it easier to evade certain regulations -- they say that, as late as the 20th century, movie companies relocated to Hollywood in large part because it was the US city farthest from Thomas Edison's patent lawyers.
But government-sanctioned monopolies are not a new invention. (That's what the British East India Company was all about.) Enormous tariffs levied on mass-produced goods are not a new invention. Taxes are not a new invention. Gangs of Luddites (not metaphorical Luddites -- the real thing) who break your equipment and sabotage your business model are not a new invention.
Wine monopolies? Here's a book that deals, in part, with how the English ban on French wine imports boosted the English brewing industry in the 17th century:
Mercantilism was all about big-government management of the economy through tariffs and subsidies. But, of course, that was arguably a big improvement over the medieval model: Small-government "management" of the "economy" by private armies who threatened to burn your crops and besiege your house unless you paid tribute.
You ask a very astute question that is too rarely asked. The way this plays out in the United States is that many entrepreneurs complain about the growing role of the federal government in regulation of business over the last two centuries, which plainly is a correct description of the historical trend. But only legal historians, with their specialized knowledge, notice that many state laws have become much less restrictive of business activity over time. It's actually rather difficult to figure out what the overall historical trend is at all levels of government in the United States. Some businesses may be easier than ever to form, while others are harder than ever to form.
Yes, government was much smaller at the time of the Industrial Revolution, at least in the US. The federal budget, as a share of the GDP, was 3% right before FDR. Now it's some giant, horrible amount I don't even want to say out loud.
Given this, the idea that there could be less (or about the same) regulatory interference of business today is, frankly, ludicrous. For one thing, simply reducing business budgets (taxes) at the rate the government does is an onerous burden all by itself. Arguments can certainly be made that it's a net gain for society, but please do not be disingenuous in stipulating the facts.
the idea that there could be less (or about the same) regulatory interference of business today is, frankly, ludicrous.
That depends crucially on which business you're talking about. Exceptions abound. Consider that example of the mail-order wine business. Do you know how hard it was to start such a business in the days before FDR? There was a lot of "regulatory interference", personified by guys like "Eliot Ness". You couldn't start a wine business in the USA, full stop -- unless you kept it underground. That era also had a devastating effect on our brewing industry that took a generation to recover from.
One could also ask some of the people who run porn sites about the complex history of "regulatory interference" in their business.
And, as we now know all too well, there have been times when regulation of the financial industry was much more strict than it is today. That's not a simple history, either: Corporations and stock exchanges have existed since the 17th century, but their degree of regulation has changed in complicated ways over time.
And a relentless focus on taxes obscures other forms of interference that were more important at certain times, for certain industries. However onerous it is to start a telecom-based company today, it was probably harder back in those days before FDR, when AT&T held a government-sanctioned monopoly over all of US telephony:
In 1934, the government set AT&T up as a regulated monopoly under the jurisdiction of the Federal Communications Commission, in the Communications Act of 1934... by 1940 the Bell System effectively owned most telephone service in the United States, from local and long-distance service to the telephones themselves. This allowed Bell to prohibit their customers from connecting phones not made or sold by Bell to the system without paying fees. For example, if a customer desired a type of phone not leased by the local Bell monopoly, he or she had to purchase the phone at cost, give it to the phone company, then pay a 're-wiring' charge and a monthly lease fee in order to use it.
That the Internet managed to come into existence at all is the result of a complex interplay of antitrust law, the rise of government-funded research under guys like Vannevar Bush, and the Cold War (DARPA, which funded the Internet, was born in the days after Sputnik).
Today, our cable and phone companies also wield virtual monopolies with the help of the FCC. The situation is pretty bad in the USA. But I'm not sure it's significantly worse than it was back in 1940.
> Consider that example of the mail-order wine business. Do you know how hard it was to start such a business in the days before FDR? There was a lot of "regulatory interference", personified by guys like "Eliot Ness".
Prohibition was passed in 1919 (under Wilson) and "repealed" in 1934. (The repeal amendment federalized some restrictions that hadn't existed before 1919.)
No, but a reasonable percentage of those convinced that "it's all the government's fault" are going to continue to see it that way no matter what (and get offended when that gets pointed out, apparently). Just like other people think that more government/laws/rules is always the answer. For that matter, I'm pretty convinced of my "radically moderate" views which are somewhere around The Economist's.
Which is why these sorts of discussions rarely go anywhere, even over a beer, to say nothing of short, sharp answers on internet forums.
I'm fairly willing to be convinced that, on some point or another, the best course of action lies close to what I held to be true, but differs by a few degrees. I also try and think about the holes in my argument and to consider the fact that I am not actually omniscient, and may very well be wrong about things (it is rare, but does happen:-).
I was thinking about this, and part of the problem is that these threads often feel like someone selling me something, rather than trying to figure out how things really work. They'd be much more interesting if people "let down their guard" a bit, upped the reasoning, and lowered the propagandizing. I guess it may not feel like propaganda to those involved, but I really like reading comments like mechanical_fish's, which cast a wide eye on things, considering all points of view, rather than people pulling out all the stops to fit the facts to their point of view.
"moderate" is a cop-out. If you're a moderate now you would presumably have been a moderate 50 years ago. And the moderate of 50 years ago would be fringe today. The moderate of 100 years ago would be a radical today.
Don't surrender your reason to social contructivism.
That's not a bad point, especially where it comes to how we treat other people. Indeed, I am not a moderate in everything, clinging to the center. However, in terms of government intervention in the economy, I am, and I also think it's far too large and interesting, and nuanced of a subject to adequately cover it in these sorts of forum. It's best left to be talked about over a bottle of good wine.
Why is it bad that in order to automatically generate binding legal documents, someone involved needs to be educated in the subtleties of legal language? A walking, talking lawyer should be personally responsible for the quality of the documents being generated. If there's no such person, you're putting the onus of validation on the client, which I think misses the entire point.
In most of these cases, it's not the government protecting existing companies, it's the government protecting the naive population. You won't fire someone after six months of unpaid labor, but someone else will. Your expert system might create air-tight contracts, but someone else will cut corners.
These laws do not exist solely to frustrate you. Part of being in a uniformly free society is ensuring that naive people are not at the mercy of the better informed or less scrupulous. Things can't work any other way.
> Why is it bad that in order to automatically generate binding legal documents, someone involved needs to be educated in the subtleties of legal language?
I wouldn't put up a huge stink if the law was "you must pass the bar exam".
However, the law is "you must attend 3 years of law school".
> These laws do not exist solely to frustrate you
Agreed. They exist to decrease the supply of lawyers, thus increasing the price that lawyers can demand.
I know a fair number of newly-minted laywers, and my understanding is that most of the concrete material tested in the bar exam is taught in the summer following graduation from law school. Law school is largely concerned with abstract concepts that are important, but more difficult to test.
I sympathize with your frustration, but the reason for the prerequisites could just as easily be that they don't believe they can create a test which properly encompasses everything a lawyer should know.
Also, at least in San Francisco, the job market seems to be saturated with lawyers, even before the economy went south.
>>In most of these cases, it's not the government protecting existing companies, it's the government protecting the naive population. You won't fire someone after six months of unpaid labor, but someone else will. Your expert system might create air-tight contracts, but someone else will cut corners.
This untrue, throughout history companies get laws to make upstarts more difficult. There is no reason you need to go to law school or even pass the bar, if the buyers don't think it is a good tool they won't use it.
In fact things can work the other way, you are protected the few who may be swindled (they're clearly visible), but ignore all those who are harmed by the lack of new products. It's very hard to look at the other side of the equation, but it is normally quite significant.
Do you want to back up that assertion? I'm not saying regulatory capture doesn't exist, I'm just saying it's not the primary driving force behind legislation. You don't get to just say I'm wrong and make vague references to historical trends.
Unless a company already has lawyers on staff, they have no way of determining whether a legal expert system is a good product or not. For that reason, it cannot be a substitute for a living, breathing lawyer. It could certainly be a tool for lawyers, and streamline the hell out of most legal processes, but to treat it as something more is to give up on any lowest common denominator for legal advice. And while this would certainly harm lawyers, it would also harm a great many other people.
Well there are many cases of regulator capture, the AMA, the railroad industry, the airplane industry (pre-deregulation), the telephone industry (pre-deregulation), and the cable industry.
The recent mandates on testing toys for lead, has pretty much outlawed small toy producers. You can't prepare food at your house and legally sell it (even if you say this was cooked at my house). The FDA debacle with juice pasteurization vs more modern UV processes.
> And while this would certainly harm lawyers, it would also harm a great many other people.
Once again you make a statement that claim to be true with no real backing. Say 1 person is harmed by this newer cheaper law advice, and 10,000 are saved due to the reallocation of resources to medicine -- this is of course a hypothetical example, but it has just as much grounding as yours.
The lesson to learn from your example is: don't work for someone without getting paid, unless you /really/ trust them.
Not: we need the government to be our nanny. That results in all kinds of really bad side effects, including those illustrated by the person who started this thread.
I agree. Instead of trying to protect the masses from their own stupidity, why not work harder to educate the masses? I realize that there must be a balance in all things, but I think the scale is tipped heavily in the direction of the government holding our hands and making up our minds for us at the same time.
We've funded lots of startups and we hardly ever see cases where they're crushed by big companies. The biggest exception is music. But though the record labels have certainly captured congress, their lawsuits tend to destroy startups before government regulation does.
>> their lawsuits tend to destroy startups before government regulation does.
In the broad sense, this is government regulation as well, but with the laws being written to transfer the regulatory function from the Executive branch to the Judicial branch.
To build on this comment, I disagree with pg's statement that startups have and will spread somewhat randomly. The biggest defining factor is having a (more) free market in a given place (inc. lower taxes and a /proper/ legal system).
In fact, that's largely why northern Europe had the Industrial Revolution and why most startups are in the US today. Because the US seems to be becoming more statist, that could definitely change.
From Europe, whilst many elements of our culture are less conducive to entrepreneurs than the US, these examples are not among them (except I don't know about the legal docs thing - I agree with earlier replies that critical information providers need to have rigorously maintained qualifications). My mate set up an internet wine company here without any problems, it's easy to pay directors of a company in equity [but non-directors get minimum wage protection], and the real estate industry is 'self-regulated' here in the UK. Oh, and we have free universal healthcare.
All that said, I still believe the next game-changing startups will come mostly from the US - which is still the destination of choice. And I agree with your central argument that 'big government' generally stifles innovation.
I felt the effect of being in Silicon Valley the moment I moved out here. The area is not only teeming with endless talent, energy, and enthusiasm, but (and this is important) we have an entire culture that makes this talent, energy, and enthusiasm possible in the first place.
When I moved to San Francisco from Chicago, I felt the change immediately. People simply get me out here. They understand why I'm doing what I'm doing. They ask intelligent, insightful questions. They give me invaluable advice. They share their knowledge freely. And they do all of this because they believe in -- and belong to -- the same amazing culture that I now do.
Funny story: I was sitting at a bar out here in Bushwick, Brooklyn, just finishing up my dinner and sipping on a nice stout. Woman next to me glances up from her book and says, "You aren't a programmer, are you?"
"Um.. yes, actually. I do user interface development. How'd you know"
"Oh, I just moved here from the Valley... I guess I just know the type when I see it."
Northeast Kingdom - Troutman and Wyckoff, right between the two entrances to the Jefferson stop on the L. We've got a Dunkin over by the Dekalb stop, but Wyckoff Starr is better (starr & wyckoff) ;)
The flip side of that is there's a complete lack of any criticism, and to me it feels like that will be its downfall (Detroit met its downfall, and so will Silicon Valley). A recent case in point is the whole DiggBar controversy. I'd suggest that the criticism has come entirely from outsiders or people "not of the valley," Meanwhile there's been a whole tide of "digg is taking over the world," "digg is growing up" from the usual denizens of the echo chamber. And then there's Robert Scoble - a one-man black hole of uncritical praise for all things SV. Add to that the complete disappearance of IPOs or anything resembling an exit (in fact the big trend now seems to be the opposite of exits -- Stumbleupon, Skype) and this piece feels a bit pollyanna-ish.
You must have seen a strange slice of the Valley. People here think much more critically about startups than in the rest of the US. People in Nebraska aren't worrying about Twitter's business model.
You must have spent some time 'out west' during the Viaweb acquisition. How much did you see of the valley in those years? I lived in San Francisco in 1999 and early 2000, and thought things were actually pretty crazy. I don't think I knew anyone who wasn't doing something related to computers, and there was a very unhealthy "free money! wheeee!" attitude amongst a lot of people, so much of it was being sloshed around in a way that obviously wasn't sustainable.
Also, there were virtually no children, no old people, not many "middle class" people... it all just struck me as very out of kilter.
That's not to say that there aren't lots of good things, or that the area is a bad place or somewhere people shouldn't go, just that it has its downsides too. I got the feeling that it has never really cast off the 'gold rush' mentality: get in, get rich, get out. Sure, some do stay and make it their home (you, for instance), but so many move on that I never felt much sense of community, something that I do enjoy over here in Italy, and have found more of in other towns in the US.
> People in Nebraska aren't worrying about Twitter's business model.
They're busy getting rich with more traditional stuff that they understand:
> People in Nebraska aren't worrying about Twitter's business model.
Very true, but we should not be so dismissive of non-startup small businesses. I'm new here, and have worked at tech startups since I got my CS degrees and started my career a few years ago.
Starting a new business on the basis of a great idea and hard work is nothing new. All around me, I see examples of small businesses being creative within their domain, and they are growing profitable businesses using the same tools we use (in theory: great graphic design, viral/word-of-mouth marketing, accessible customer service, and hard work) without any promise of making it big. These businesses see a niche and try to serve it uniquely.
I appreciate SV and the hungry tech culture as much as anyone, but we're not operating on fundamentally different principles.
By voting me up, you guys have proved me at least slightly wrong. The one thing that Northern California is demonstrably great at (over the long term) is providing room for the outsider. It seems like that's being eroded lately.
PG might be on to something here. I used to live in Austin back in the mid-90s. It has has great weather, a great university and several startups that went onto become big (Dell, Pervasive, etc). And yet Austin has failed to compete with Silicon Valley and in the last decade might have even become less relevant in the tech sector.
I've really been amazed at the caliber of people that I meet through the Hackers and Founders meetup that are from Austin. There's really a great little sub-culture here of ex-Austinites that are interested in founding startups.
At the beginning, there is always a 'center', but that does not negate the influence of the other districts.
I suspect that PG has gotten caught up in a bit of the valley hype.
There are many more counterexamples. Microsoft (does it not count because it was pre-internet?) was a start-up outside the valley at one point. MySpace didn't start out in the valley. So why skype? because a valley company overpaid significantly for them? (I love skype though, i just don't think they are the best counterexample).
Detroit thought it had all the 'talent' and 'knowledge' in building cars, and then got it handed to them on both sides, cheaper and better from Asia, more expensive luxurious and better from Europe.
No question the valley has investors, and very talented ones at that. As costs for internet start-ups decreases, amazing open-source technologies like hadoop continue to spread, and talent realizes they can excel anywhere, why would the valley continue to as the center?
What will be the draw to keep people in the valley? Angel financing can now accomplish what only VC could a few years ago, and that continues to change. Ideas are not limited to the valley, technology is not limited, for many start-ups their are no physical goods, and when there are, those are created overseas.
So, though Silicon Valley may be the center today, what will make it the center tomorrow?
This comment seems to disregard the point of PG's essay. He's describing a network effect. Just as it's hard to create a startup community (not a single startup, as in your examples) you need the types of people who build startups present or on their way. The valley is full of people who are interested in startups and work with them to some capacity. It would be difficult to attract someone interested in startups to move to New Haven when there is already so much attraction to the valley. It's true, you could start a startup anywhere. It's probably going to happen in the valley.
When you interpret Paul's "startup" as a "newborn Internet media company" the essay makes a lot of sense, especially the comparison to movie business.
But if you interpret "startup" as "newborn consumer electronics company" or a "newborn software company" or anything else more generic not squeezed into a web page it falls short: first, the percentage of GDP generated by internet media is miniscule, the whole thing isn't that big, it's a sizable niche in a software development world but not more. For instance, when measuring life in "robotic startups" one may conclude the SV is not relevant, but Boston very much is, and Japan is taking over the world, I'm sure Trevor would agree with some of this.
The kind of companies Paul is talking about appear to be software companies, in general. More new software companies originate in Silicon Valley than anywhere else in the world by a wide margin. Most new software companies (Google, for example) tend to be internet companies, but that's just because the internet is the most sensible way to distribute most software now.
Robotics other than military, industrial, and medical is really nonexistent at this point. Japan's robotics are cosmetic and hardly taking over anything. With something like robotics it is difficult to weigh who will end up funding you. The answer today is probably the government, not VCs, so location is irrelevant.
I think pg has misjudged what stage of the start-up revolution we're in.
He notes that the industrial revolution spread quickly because there were geographical areas with middle class populations and cultures that let people keep the large sums of money they made. He also notes that the culture necessary for large numbers of start-ups is not present in many areas.
IMO, we're no way near the railroad stage of the start-up revolution. I think we're in the early renaissance. The valley is like Venice in the 1400s. It's a very localised concentration of merchant traders and a culture of what we now call "modern values", meanwhile the rest Europe is living in the "dark ages". It wasn't until the plague spread across Europe and gave the surviving working classes more influence that the culture started to change. Even then, it's another 300 years before society really starts to change.
Looking at Europe in the 1400 hundreds you might well have thought that Venetian style modernism would be a very localised revolution and that these new fangled "banks" would only ever be established in small areas.
Start-ups are not the "railroads" of the industrial revolution. They are the "banks" of the industrial revolution.
The main problem with PG's article is that he is, as always, overly focused on startups. For something to be a revolution, it has to last for a while (otherwise it's just a rebellion, and a failed one on top of it), which is something directly opposite to the nature of startups.
By definition, startups don't last -- they either disappear (either by failing or M&A), or they stop being startups and turn into "regular" companies. But, that being said, I still think that Paul is on to something -- the future model of a succesful company is not a large multinational conglomerate like IBM, Microsoft or Shell, which is too large for its own good and where, like with dinosaurs, it takes ages for a signal to travel from limb to brain and back.
Instead, I believe that the future of companies lie in small, flexible and agile companies which have simple operations and global reach. This doesn't mean that a successful company should somehow prevent its own growth -- but it should grow in small units, sprouting new, smaller companies which operate on their own.
My favo(u)rite example, which proves that this model works, is the Virgin Group, which includes lots of small to medium sized companies which operate in a number of different industries. For a traditional, centralized company such diversity would be a suicide, but in this case the companies are connected by the common brand and its messages.
"The most likely scenario is (1) that no government will successfully establish a startup hub"
Take a look at HaiDian/ZhongGuanCun in Beijing, chock full of startups big and small. From my understanding, the reason entrepreneurs flocked there originally was because HaiDian offered huge tax breaks versus the rest of China (for the explicit purpose of developing a high-tech industry).
In my eyes, HaiDian is a huge counter-example for the first point--not forgetting of course that the economic climate in China is one of urbanization unseen since the American Industrial Revolution.
PG does appear to assume ease of movement by potential founders to current startup hubs. Large cultural and language barriers remove this ease of movement. Perhaps a more rigorous claim is that there will be very few startup hubs per largely-different (culture,language) region.
You're making a substantial assumption, which is that startups look for tax incentives when founding. Since nearly all startups are unprofitable for a while, I suspect that most don't really care what the tax codes in the area are, as long as they aren't ridiculous (insert your version of this accordingly).
I feel tweaking tax codes to encourage startups won't have its intended consequence, but rather encourage established companies to try to pretend that they're startups. They are much more likely to have the resources to chase tax incentives than a real bootstrapping startup.
1. Innovation over the past couple centuries has not spread evenly. Cities always got a disproportional deal. That is why even today small towns all over the developing AND developed world are still losing people to the big cities.
2. A big part of the high-tech startup concentration seems to be VC clustering. But if startups are to be a truly revolutionary change I hope we get away from the investor model and move closer to the paying customer model. And this would make the valley's regional advantage smaller.
Especially nowadays, with the direct costs of a web-based software startup approaching zero; there is no reason for fincanciers to be the bottleneck. When three motivated people working out of their homes can build a fully functional site; the direct costs to the founders aren't that different from their living expenses and there is no need for large amounts of capital until after product development is completed.
The main advantage of regional hubs is that of social density; easy access to expertise and enough demand for experts that they can survive and thrive.
In practice I doubt any government would have the balls to try this
I'm a bit surprised to see a metaphor about testicles used.
A few thoughts:
This isn't a revolution if it is localized.
Also, there might be a way to create a government that had the will needed.
It is going to get easier to start countries. Rich individuals like the leaders of Qatar are buying their way into a good university system, e.g. http://www.qatar.cmu.edu/
These two trends mean an empowered individual could start kickstart a startup hub with a clean slate of government. Immigration law, taxes, and other burdens could be removed, making it particularly appealing.
Schools for the kids of founders and university employees would be the hardest part, because we haven't seen an efficient pattern for schooling that is also effective.
I don't see irony, just wordplay between "local" and "remote". If you are given a tool that can remotely connect resources then it's not unexpected that you should only build the stuff in one place. You use the remote-connection tool to reach everyone, making anyone doing the same thing elsewhere mostly redundant.
I do wonder if future improvements in communications technology will render the need to go to Silicon Valley moot.
If FutureHN can provide both a valleyesque enthusiastic social network and a contact with investors, will the benefits of migrating to S.Valley become less important?
It'll certainly be interesting to watch. I'd like to move to SV, but it's not really practical for me for a variety of reasons. So I'll keep plugging away out here; I don't really need investors (yet), but the idea of having so much technical talent in one place is really what makes Silicon Valley so unique in the world.
I'd still like to see you talk some more about the economic factors that point to startups being very common (Coase et al). In other words, what fundamentals have changed so that starting your own company is the best option for a sizable number of people. I think to predict a true revolution, you'd have to demonstrate that these conditions don't just exist in a particular high tech sector with amazingly low capital requirements, but across a much broader swath of the business world. For one thing, will capital requirements drop in other sectors too? Or is the web just a blip, and eventually capital requirements will eventually rise again as it takes more 'stuff' just to get off the ground? What other things come into play that mean that more people will strike out on their own. Are there historical precedents?
finance -- New York, Boston
automobiles -- Detroit
entertainment -- LA
etc.
There are natural reasons for that: new companies are often founded by execs who think they can do better than the big companies they leave, infrastructure gets built up that benefits all in that given industry, talent begins to congregate in those areas, etc.
I think that different types of startups might be congregated around different geographic locales: biomedical startups would likely flourish near big research hospitals, etc. Service industry startups likely aren't clustered: restaurants, etc.
With technology startups, yes there is a natural clustering in Silicon Valley and Boston, but I think that's because tech startups are fairly new. I hear lots of good things coming out of Chicago, Austin, Columbus, Minneapolis, and elsewhere. It's definitely spreading.
This is on the line of why I think this article is actually pretty weak; it's acting like a bunch of known, understood factors are somehow new and revolutionary.
Yes, industrialisation "spread" in the sense that it was possible to use an engine anywhere, but in practice most countries developed tightly clustered industrial areas (as in heavy industry being the predominant economic engine), while others areas of the country remained either agricultural or mercantile or service economies.
The reality is that in any human endeavour (steel production, summer blockbusters, LCD screen manufacture, grunge, salmon farming, dadaism) there are clusters of activity - they do not preclude the activity being carried elsewhere, but the people who do carry out the activity at the cluster have both synergy gains, and act as a magnet so create a snowball effect.
Obviously the SV is a cluster of excellence for startups. Why is this expected to have a social impact, even if they do represent a new economic phase? (which personally I doubt - the numbers of people involved are just too small). I don't get it.
"Startups don't seem to spread so well, partly because they're more a social than a technical phenomenon"
It's this social aspect that makes me think startups will spread beyond some major centers. The Internet has proven to be great for "finding the others," be it for social/political/transactional kin.
Also, as pg suggested in an earlier essay, vc is only growing less necessary for many startups. The shrinking dependence on vc, and the social lubrication of the Internet suggests more startups will bloom in all sorts of places.
I appreciate the caution expressed in the essay; that it is unclear both that an event is on its way and that location is necessarily decided. No doubt SV is the capital of startups, but I still think in the world of startups it is slightly different than Hollywood or Wall Street because you can do it with little money or connections. The distribution of startups across the world remains to be seen, but I agree often it will depend on who is offering the money.
To address the randomness of the revolution, I sincerely hope someone out there is developing some kind of portal to find all of these neat tools being developed. There are several YC apps in faq.html that, when put together, give many people no reason to really have any desktop software at all.
Just as an example, let's say a school teacher uses her computer for email, web browsing, managing files, photo editing, and maintaining her second grade web site. Gmail, (some browser), dropbox, snipshot, and splashup/weebly would be all she needs.
If all these websites were somehow organized in a way that said teacher can easily find them, then I think a 'startup revolution' would be more likely: people could go to a central website to browse for startup that meets their needs.
"People are dramatically more productive as founders or early employees of startups—imagine how much less Larry and Sergey would have achieved if they'd gone to work for a big company—and that scale of improvement can change social customs."
I question this assertion. I'd bet a single Google engineer who makes a 1% improvement in adsense targeting can create more economic value than most startup founders.
Exactly. Productivity only matters when it's directed properly. A brilliant MIT student working 20 hour days on some silly mash-up is not nearly as productive as a slow and obtuse engineer in some large corporation working 6 hours days. It's the results that count.
I would argue the contrary case to Paul's; I think that Silicon Valley has slowed down growth. They make it a lottery that you can win with having a cool, quick hit. The money in the valley has little love for hard problems and a lot of love for fads, much like record producers.
This sounds nice in theory, but I don't think it is always true. There are techies who have a career and those who have a job. It is better to get as many who have a job doing what they are passionate about, because they are probably wasting their time at their job.
And I wish that at least one of the 100+ companies that have come out of YC were IPO bound. I tried to hint that we were of that mindset when we had applied, no cigar. I'm not really sure why YC continues to advance sites which seem to go against their own "Derivative Idea" rule for why startups fail (e.g. this round: a group bookmarking site, a themed video site which piggybacks on YouTube). It would be interesting if YC decided to have one round where they choose nothing but "hard problems" with huge potential.
I couldn't agree more about YC. The application questions would change to "we're only interested in single founders, but we will consider a technical co-founder." I think it also just makes sense for the "spray and pay" style to fund 1000 startups focusing on hard problems with the potential of 1 making up for it all.
When we hit it, dbul, you and I can put this together. Deal? :-)
(1) that no government will successfully establish a startup hub
What about the Pearl Delta and Shanghai's Pudong district? Both seem to be smashing successes.
There are two very different types of startup.
I don't buy that there is a hard distinction between different types of startups (or even between startups and small businesses).
Any small business that tries to grow will create a differentiated offering. An ice cream shop might promote a special flavor, a restaurant will develop a certain vibe, a furniture company will create signature ergonomic chairs, a web form company will create an awesome UI, etc. The amount of uniqueness of the offering can be plotted along a continuum. When it passes a certain point, we call it a "tech startup".
The special thing about Silicon Valley is that it attracts a ton of many smart, talented, ambitious people. Therefore the offerings of their businesses tend to be very novel and differentiated. The supply of brilliant innovators is limited so it is not necessarily possible to replicate this everywhere (but I am guessing the world is very significantly under performing its potential).
Edit:
Perhaps another way of saying this is:
1) A startup is a small business that attempts to scale by developing a new technique (that technique can be a product or a way of doing operations).
2) Different sectors of the economy have different opportunities for developing new techniques. It's harder to develop new techniques in agriculture than in software.
3) In the absence of high transportation costs, industries tend to cluster ( autos in Detroit, entertainment in Hollywood, finance in New York, energy in Houston, education in Boston, software in Silicon Valley).
4) Silicon Valley has a cluster of industries that natural have a high level of new technique development (software and electronics). Therefore Silicon Valley has many startups.
That said, now I'm not sure I buy my own argument. Silicon Valley seems to have the most innovative companies in every industry. In addition to software startups, there are energy startups, automobile startups, etc. So Silicon Valley really is a cluster of companies that specialize in innovation. Perhaps the thing that limits the spread of Silicon Valley magic is that there are only so many innovative people, and only Silicon Valley has been able to attract a critical mass.
I agree technology may represent a new economic phase -- but startups are just the way they're introduced. And it has nothing to do with productivity and everything to do with uniqueness of vision. If big companies did more things like Google's 20% project, you'd have the same effect.
Also, I think you ignored that VC is becoming less and less important. And as it gets less important, the big funding hubs will become less important. I see a future of micro-equity and the big hits coming out of everywhere.
Also, the hubs have a horrible self-delusional feedback loop. Once you get an entire environment that thinks the same way, these feedback loops are inevitable and you spiral into silliness, e.g. Twitter.
The other question to consider is if the revolution is local, does that really matter for the society at large? The startups are creating value for the rest of society, regardless of where the startups are located. Using the internet as the means of distribution, decouples the location of the business from the consumers of the product, thus allowing the startups to have a central location.
The business revolution will only matter for a subset of companies, mostly software companies.
The fact that the Industrial Revolution was localized mattered a lot to the regions that were left behind. It made them the victims in the ensuing era of colonialism.
That's true, but meaningless. History is only our friend when it applies. A lot of technology is now software based and thanks to the web, it is immediately distributed evenly to all those who have computers with internet access. While that still leaves out a large chunk of the world, it gives no substantial benefit to the place where the technology was born.
Open source and internet help to spread software source. But the mindset of how a piece of software solve users' problems are still limited the location that the software is created.
It will be more like a culture barrier. For a culture that has no sense of custom satisfaction, the user interface for e-commerce will never be as good as Apple or Amazon. We as technologists sometime forget a lot of our assumptions are based on our environment.
That's a perfect reply. You're absolutely right, but I know a lot of companies who've made a fortune doing "Ebay for Switzerland", etc. Businesses will step in to fill that gap to a certain extent but the applications themselves reflect the hub culture. Excellent point.
A lot of technology is now software based and thanks to the web, it is immediately distributed evenly to all those who have computers with internet access.
The first half is correct, but not the second half. A few examples: the software that runs on spy satellites, the software used to design stealth planes, the software used in sigint, "smart bomb" guidance systems, the software used to design nuclear weapons, etc.
Regarding spreading, it seems to me that location doesn't matter as much as it could. I've never lived in the Valley, or in any other stereotypical "tech hub" though, so I could be wrong there.
I'm currently working for what I consider to be a startup in a small town in the UK. We're making software that is applicable to a specific industry, and the industry is very hungry for it. I'm currently the only programmer.
What looks like the probable course for us (and probably for other people like us - people starting up software dev stuff in smaller areas), is that once we've sold what we're working on right now, we'll probably move on to doing more interesting things, and form a small software development house of sorts.
What we're doing right now would be utterly impossible without the internet, if only due to communication-over-distance issues.
I think that the stereotypical "startup" is just the start of what we'll continue to see as more and more people realize that they can make their livelihood online. What I'm doing doesn't fit the stereotypical startup mold, but is the type of thing that I think we'll see more and more of, assuming nothing majorly bad happens to the internet or society in general.
Of course, SV would dominate, but I wonder by how much. It'd be REALLY interesting to see it as a ratio to total startups started, but I think that'd be a harder # to come by.
I think startup success is fueled by how many people have the desire to start something up and access to early-stage funding. The Valley clearly dominates the latter, but it seems (anecdotally) that hackers who want to spin something up are becoming more commonplace in non-valley locations.
The essay's stated thesis is that the Valley is and will continue to be a hub for startups because of the abundance of Venture Capital and talent, both of which are produced by previous startups, maintaining a strong cycle of success. At the same time, the essay points out that Seattle is a center because Gates and Allen wanted to move back home, and it has been suggested in past essays that Venture Capital is increasingly less necessary for startups.
My impression is that PG has unintentionally suggested that the current state of high housing prices and less Venture Capital could easily reduce the Valley's role in the startup market. The way for other cities to capitalize on this is to convince smaller, upcoming companies to move "back home", or not relocate to the Valley in the first place. As some commenters have pointed out, other states may have legislation that is less inviting to startups, and these should be repealed.
Ultimately, startups will probably flourish wherever smart, ambitious people want to live, as long as the government treats them well.
Starting little businesses to write microblogging and photo sharing websites does not matter in the scope of human history. Comparing web startups to the industrial revolution?! You gotta be kidding me, I don't care how much ajax or standards compliance or user generated content your social grocery networking website has, you're deluding yourself.
On the other hand, it's easy to break into the startup industry from anywhere in the world. You just need to gain a certain number of customers and following, and you will be noticed. Contrast that with the film industry, where the need for expensive equipment means that financing is critical to every venture.
It makes sense for the film industry to be concentrated in hubs because of the huge expenses in each film. It does not make sense for startups, because such expenses don't exist, and the need for connections, while important, is not critical.
So, though the trend may look like it's going towards major hubs for now, over time this will normalise itself as those players who are not able to go to the startup hubs will fragment and create successful companies all over.
There will be few major hubs for technology startups - but they won't be ever as concentrated as hollywood - simply because there is no need for them to be so.
I wonder if a smaller local revolution has happened before: formation and spread of great research universities. Govts have tried to set up great research universities by will, China is trying hard, India is making small steps, so are places like Germany. But they have not succeeded yet. Even in a small country like UK, only 3 great universities have emerged.
Within US, Stanford is today a great research university but it was not so in 1945. Great luck like the combination of cold war spending and the idea of steeples of excellence by Terman made it happen. Uni. of Florida or ASU may be today aspiring to greatness, but they are fighting the 'locality' of research universities. Maybe some random events will help them out.
Nobody seemed to mention (in a quick read) that the local/legal environment influences whether startups fly or not. Boston was a startup hub well before Silicon Valley was flying, but it's my understanding that IP and labor law in MA is (far?) more favorable to the employer than the employee - as compared to California.
If you're going to look for an environment that has "movement" of ideas and people, the environment has to be conducive to this. California has done well, but might strangle itself on commuting costs, the lack of infrastructure and the costs of living there.
It'll be hard for states to favor new vs. old businesses, but there's 50 of them out there, and a number of them will probably give it a go at some point.
Information technology is still just a fraction of the whole world economy. I don't think it is big enough for anyone to have the slightest idea of what the world will be like when we actually become an information-based economy.
I think the point that people like the google founders were many many times more productive than their equally-smart employees counterpat is important, but I think the same contrast can be found on people working on their free software projects compared to people working on their day-job projects.
I personally think that things like free software will have much more impact in the future economic revolutions than the concept of "start-up".
It is pretty neat to wonder if, during the Industrial Revolution, people had nearly-free ways to ship their manufactured products and electricity, if there'd have been more than a few hubs, or if it would have worked out about the same.
I am keeping my eyes on Boulder, CO. From what I hear, good university, good business climate, willing to take initiatives (Smart Grid being rolled out as we type).
Boulder _is_ a good example of how to foster a tech hub. Wired had a print story about this a few months back, but I couldn't find it on their site. Though they did claim that Colorado's Front Range is an optimal site for a new SV. Given time, I agree.
I can say that there is a very open and supportive start-up culture here (Boulder & Denver metro). Tech Meetups and LUG groups are pretty active; the new-tech group based in Boulder routinely overfills 300-seat auditoriums and is expanding to handle the traffic. Most tech folks here have the mentality of wanting others to succeed, no matter if it is with our own companies or with a different one. We are happy to share knowledge.
Sorry no beachfront though, you'd have to settle for the best mountain sports anywhere.
If you want to see startup revolution as the norm, just go to Asia. I'm sure Paul Graham has seen the scale of small business across the Mideast, China, Southeast Asia, for that matter the whole world, I just haven't been to South America so I can't speak about that.
It's kinda helpful to look past your own borders sometimes. There's a lot more lint in the world than just in your own belly button.
as i told Paul when he sent me this essay, he's got it all wrong. silicon valley's rate of startup growth is slowing and the rate of startup growth in many of the leading cities of the world (london, bejing, mumbai, banglore, new york, tel aviv, berlin, etc) is increasing. i think Paul's view has been skewed by living in Boston and Silicon Valley. Boston's startup ecosystem has been in decline for almost ten years now and when compared to Silicon Valley, it looks like game over. But if you chart the number of new startups funded by angels and VCs in all these cities since 1995, you'll see the right picture. I did that for my web 2.0 keynote and if i can find the chart, i'll post a link here
Excellent essay, as usual. Question: does this imply that Silicon Valley will continue to grow as a result of being the dominant location for startups? And if so, is there a point at which its size will diminish its superiority?
What about the spread of existing hubs? How much bigger is Silicon Valley than it was a couple of decades ago? How many start-ups and how much land will it encompass in another 20 years?
Paul, this post gives me hope. I'm now in a university belt and we're about to turn up the research heat and our students are seriously looking into startups - perhaps your localization will materialize before our very eyes. And yes, still no Silicon Valley equivalent to contend with at the moment!
If so, this revolution is going to be particularly revolutionary lol every revolution is revolutionary that is why it is called a revolution and I doubt you may have a particular revolutionary revolution lol
Anyway, you may say that the intellectual revolution of the enlightenment was local, it started in Italy and remained there for a while, some century I think. But this is not a start up revolution, it is a technological revolution, the revolution is not the start up culture but the advances made in communication technology. As such it is not any different than the industrial revolution as far as the importing of technological aspects is concerned. Nor is it any different from a power plant, every nation must have internet and with Italian internet comes Italian Google, which as it happens I think it is libero.it although that might be more like yahoo than google. So perhaps now the start up culture is localised to silicon valley, but I doubt it will be long until a silicon valley appears in London or Rome or Berlin.
Government was much smaller at the time of the Industrial Revolution.
The theory of "regulatory capture" says that government is responsive to those firms that are affected by government regulation and have the money to buy favorable legislation.
Thus, it's much easier for big-dumb-and-slow firms to use government to put roadblocks in the way of startups than it was for inefficient firms to put roadblocks in the way of factory production 200 years ago.
Want to sell wine over the Internet? The established players will make sure that it's illegal for you to do so.
Want to cut a deal with your early stage employees where they get no pay for 6 months? It's illegal in most places to not pay someone a wage. In MA it's illegal to fire someone for blatant incompetence and they pay them their final paycheck on the regularly scheduled payroll 4 days later - you need to hand them a physical check then and there.
Want to buy a bunch of servers? In a lot of locales (MA, again), you have to pay taxes on the assets every year.
Want to sell real estate on the Internet? Hope you're a licensed broker!
Want to write an expert system that generates legal documents? You'd better have passed the bar exam. Actually, wait, even though that was good enough for Lincoln, these days you need to have actually ATTENDED lay school - a mere demonstration of proficiency is not enough.
Etc., etc., etc.
If this is how things are in the free-wheeling US, I imagine that it's far worse in Europe.
I think that people across the globe have what it takes to launch startups. I just think that the legal culture makes it very hard in most places.