Tuesday, February 22, 2011

Part 2 - Semiconductors

In this post I will look at the major area of technology which Professor Cowen misses.


The first and still most important product of the semiconductor industry is a device called a transistor. A transistor is a fast electronic switch which allows one electrical circuit to control another. Electronics, and even computers, did exist before transistors were introduced in 1954. They depended on another kind of electronic switch called a valve. Valves were light bulb like devices which were bulky, power hungry and produced substantial quantities of heat. Transistors were a huge improvement.

The real potential of transistors was harnessed when several were integrated together on a single piece of silicon. The number of transistors on a chip increased until by 1971 it was possible to put the entire brain of a computer on a single piece of silicon. The microprocessor was born.

In the chart below I look at the impact of semiconductors over time.

The microprocessor lead to pocket calculators, replacing slide rules and adding machines and making math calculations much easier.

The microprocessor also lead to affordable computer numerical control ( CNC) machine tools which automated parts of manufacturing in the 70s and 80s.

Most important of all, the microprocessor gave us the personal computer. Spreadsheet software was a very compelling application for the personal computer, because it could automate tedious financial calculations which once would have taken days. Spreadsheet software, faster mainframes and ATMs impacted the financial industry.

Database software helped businesses to keep records. When coupled to bar code scanners and cash registers, they automated the tedious business of keeping track of stock in stores and warehouses. Effective use of these technologies, together with satellite communications, helped Walmart to dominate retailing.

VCRs, cable TV, satellite TV and computer generated imagery (CGI) changed the information and media industries.

Finally the Internet has had many impacts, as even Professor Cowen admits. For example E-commerce has improved the market for airline tickets, and load factors on planes are about 10 percentage points higher than they used to be. Retailers like Borders and Blockbuster are being challenged by Amazon and Netflix.

Professor Cowen makes a fair point about how semiconductor technology has helped to make 'cheap fun' easier. Game consoles, DVDs and social networks have improved leisure time.


  1. I don't think Cowen would disagree that semiconductor technology - and computers in general- have wrought great change economically and societally. He would just argue that

    A)These impacts are not as great as earlier transitions, such as industrialization and the post-war period, and

    B)These developments don't generate anywhere near as many jobs.

    I actually think he's wrong on the second point, to some degree. The reason why iPods, for example, haven't created as many jobs as GM cars did is probably at least in part due to the fact that most of the supply chain is out of the country.

  2. It's not just IPods. Almost all of our electronic goods are imported. Consumer electronics has generated plenty of jobs, but they are on the other side of the Pacific. Of course, US production would make Ipods and computers a little more expensive.

    Imagine where America would be if Henry Ford had built the Model T assembly line in China and imported the cars into the US. That is what a modern globalized business would do.

    About your first point: One of the issues here is nobody seems to have any real quantitative measures of how much particular technologies contributed to growth.

    Cowen implies that there has been a sharp slowdown, a stagnation, in the development of technology. I don't think that is the case. He mentions the internet, but he doesn't talk about earlier things like ATM machines and spreadsheets which have some significant productivity implications.