A few days ago a post entitled Autoworkers of Our Generation floated across my radar. In his post, Greg Baugues argues that as developers, we have a short term advantage, and would do well to view our lot as a temporary anomaly. In this article I’d like to engage with his post, and respond.
The key analogy of Baugues’ post is the parallel between the Detroit auto worker of the 1980’s and the Macbook hugging developer of the new millennium. Without disagreeing with the latter characterisation, I believe the analogy with the auto worker is flawed.
Detroit, or more generally, the American car industry of the ’80s was highly uncompetitive. The Go-Go decade of Wall Street had distorted the market for home made vehicles, With the oil crisis of the previous decade a distant memory, local car manufacturers pursued larger, more expensive vehicles, presenting them as aspirational symbols of bull market opulence. Luxury sticker prices supported the wage claims of Motor City’s highly unionised workforce and for a time there was a virtuous cycle; an affluent customer base drove demand for more expensive vehicles supporting an inefficient manufacturing sector.
The results spoke for themselves. By the 90’s, foreign manufacturers moved in to capture the under serviced budget car market and created a beachhead to attack the local luxury brands from a lower manufacturing cost base. Responses from the lobbing industry in the form of import tariffs, did nothing to solve the underlying structural problems and delayed the restructuring of the domestic producers. As a last ditch measure, the local manufacturers aggressively pursued off-shoring and automation, decimating the lives of many of middle class Rust Belt families.
The parable that Baugues’ tells is clear; if you or your parents were manufacturing cars in the 80’s, you could be forgiven for thinking that you had it made. However, to accept Baugeus’ analogy of the misfortune that vested on Michigan, is to accept that software development is a purely mechanical action, prescribed, just as the creation of a motor car on a production line. I cannot agree with this.
My career in the computer industry started in the mid 1990’s. I was there for the hysteria of the Y2k fiasco, and saw the effects of a generation of university graduates, late to the party, but none the less eager to share the spoils of unbounded consulting hours, meet the reality of the dot com collapse. In 2001 a degree in comp sci was worthless. Yet, within a few years, the industry had bounced back. With the exception of periods of industry wide delusion the demands for software developers, or more correctly the software itself, has always been strong.
In conclusion, I accept Baugues’ thesis that an economic winter is coming. It must, if for no other reason that every other boom in a western economy is followed by a bust. But I believe the further conclusion that software development will become automated to the point it can be produced en mass, is false.
I will leave you with the recommendations of the original author:
Don’t get too comfortable. Don’t get locked into a language. Don’t burn bridges for short term gain. Keep your tools sharp. Learn soft skills. Build an audience. Save some money. Network. Read.
It’s an obscenely good time to be a developer. Enjoy it while it lasts.
Great analysis Dave. Developers, and knowledge workers in general are not like manufacturing roles, for any number of reasons. In the same way Doctors and Lawyers are not like manufacturing jobs, neither will Developers ever be impacted in the same way by the same forces as auto workers. I do however disagree with some of your economic history.
Luxury sticker prices supported the wage claims of Motor City’s highly unionised workforce
I think this is a bit backward. The wage claims came due to luxury sticker prices and the corresponding profits.
By the 90′s, foreign manufacturers moved in to capture the under serviced budget car market and created a beachhead to attack the local luxury brands from a lower manufacturing cost base.
The failure here wasn’t *solely* due to inefficiency. It was also due to the innovator’s dilemma – where correct decisions to make change to assist with current markets destroy ability to compete in new markets. They were not inefficient so much as very efficient in creating the wrong car.
every other boom in a western economy is followed by a bust
Only because we allow it to happen. The bust of the Great Depression was followed by a number of regulations to prevent these epic swings. It worked admirably as the US economy was stable/growing for 40 years post WWII. It wasn’t until the de-regulation during the 80s and 90s that we got widely fluctuating boom/bust cycles.
Yet, within a few years, the industry had bounced back.
Very true. One of the things I remember most about the pre-dotcom bust was very marginally skilled developers getting insane salaries. So many business and developer were maximizing for short term gains. I think this is also different from the Auto industry, and there really wasn’t a sense among those companies that they were building something to last. They were building something hoping to cash in on the lottery. Once the lottery closed shop, business that were trying to build viable products emerged.
Don’t get too comfortable. Don’t get locked into a language. Don’t burn bridges for short term gain. Keep your tools sharp. Learn soft skills. Build an audience. Save some money. Network. Read.
Can’t argue with that.
I’m really interested by this topic because I live in southeast Michigan where the Detroit auto industry grew up and sits. GM proving grounds is minutes from my home.
This issue is complex when you go below the surface. For instance, look at how machines have replaced many of the line workers. There are still automotive line workers but there are fewer. Yet, someone has to design and manufacture the machines. Those machines are all over.
Now, compare that to the engineers who build the complex systems as Google, Github, or one of the other companies out there. Is there something automated on the horizon to replace them? I don’t see it. But, look at the people who write WordPress plugins or Drupal modules. Look at the programmers (as opposed to those who engineer solutions). There are people already working to remove the need for programmers here. And, WordPress, Drupal, and Joomla power almost 25% of all websites.
Maybe think of it like other engineering disciplines. There are some engineers and they are paid well. Consider picturing civil engineers who handle roads, buildings, and so forth. Yet, there are many lower (I didn’t say low… it’s all relative) paid workers who pave roads. In software right now I see many people who are getting paid well for their work who could be placed in the category of the lower placed worker. Going back to the CMS examples, their job could be replaced by automation or a tool that business people could use as easily as excel.
I wouldn’t be surprised to see a shift in programming that separates out the engineers from the programmers.
Software developers are like car *designers*, not the people actually putting the cars together. To produce more copies of cars, you have an auto plant. To produce more copies of software, you just copy the bytes. I’m guessing car designers are still doing really well.
As a former owner of a 1978 Chevy Malibu, I can confidently state that it was the wrong car for nearly every situation.