The Economics of Software

The Economics of Software
An Introduction

Prof. David Bernstein
James Madison University

Computer Science Department

bernstdh@jmu.edu

The "Go or No-Go " Decision

The Situation:
- One sometimes has to decide whether to accept a contract to develop a piece of bespoke software
The Issue:
- Do the revenues exceed the costs?

The "Go or No-Go " Decision (cont.)

The "Buy or Lease" Decision

The Situation:
- One sometimes has the option of buying software or leasing/renting software (in both cases, subject to terms of a license) sometimes even from the same provider
The Issue:
- The alternatives have different cash flows hence one has to consider the time value of money

The "Buy or Lease" Decision (cont.)

The Cash Flow for a Purchase:
- "Large" payment up front for the purchase of the original product
The Cash Flow for a Lease:
- "Smaller" periodic payments

The "Buy or Lease" Decision (cont.)

Comparing using the Present Value:
1. Choose an interest rate (a.k.a., discount rate)
2. Calculate the present value of the lease
3. Compare the present value of the lease with the cost of the purchase
Comparing using the Internal Rate of Return:
- Find the interest rate that makes the present value of the lease the cost of the purchase
- Compare that rate with the actual/anticipated market rate

The "Buy or Lease" Decision (cont.)

An Example:
- I can buy the software for $4000 (in year 0)
- I can lease the software annually for $1000 in years 0 and 1, $1100 in year 2, and $1150 in year 3
Using the Present Values:
- Using a discount rate of 5%
- So, if you think that's the correct discount rate, the lease is the better deal
Using the Internal Rate of Return:
- The discount rate that equates the two is 4.02%
- So, if you think the appropriate discount rate is that value or higher, the lease is the better deal

The "Make or Buy" Decision

The Situation:
- One sometimes has the option of either "making" (i.e., developing either from scratch or by modifying existing software) or "buying" (i.e., "acquiring" which may be buying, leasing or renting) software (whether it be a component or a system)
The Issues:
- The alternatives have different cash flows
- The alternatives have different probabilities of success

The "Make or Buy" Decision (cont.)

How to Proceed:
1. List all of the alternatives
2. Choose a discount rate
3. Calculate the present value of the cash flows for each of the alternatives
4. Estimate the probabilities of success of each alternative
5. Calculate the expected present value of each of the alternatives
Organizing the Probabilities of Success:
- A decision tree that contains the alternatives (and sub-alternatives) and their probabilities (and conditional probabilities) can be helpful

The "Make or Buy" Decision (cont.)

An Example with Probabilities/Conditional Probabilities

The "Make or Buy" Decision (cont.)

Including the Estimated Costs (in Present Dollars)

The "Make or Buy" Decision (cont.)

With the Calculated Expected Values (in Present Dollars)

Outsourcing and Contracting

The Concept:
- Hire an outside party to perform some or all of the software development process
The Rationale (2014 SourcingLine Survey Results):
- Reduce or control costs - 44%
- Gain access to resources not available internally - 34%
- Free-up internal resources - 31%
- Improve business or customer focus - 28%
- Accelerate reorganization - 22%
- Accelerate project - 15%
- Reduce time to market - 9%

Outsourcing and Contracting (cont.)

Harte Hanks Market Intelligence CI Technology Database:
- 1998: 18.6% of companies outsourced programming and design
- 2000: 30.1% of companies outsourced programming and design
Congressional Research Service:
- 2001-2002: 7.1% permanent job loss due to offshoring in IT
Some Recent Data:
- The cost in India used to be 20% of the cost in the US but is now about 70%
- Salaries in China have been rising 30% per year

The Economics of Open Source Software

Definitions:
- Freeware/Shareware - The "executable" software product has a price of 0
- Open Source - The source code for the software product is available (but its use may be restricted)
- Public Domain - The source code for the software product can be used without restriction
History:
- 1960s-1970s: Software was often developed in academic settings (and similar corporate research labs), source code was made available, and property rights were not formalized
- 1980-1982: AT&T began claiming intellectual property rights to UNIX
- 1983: Stallman started the Free Software Foundation which introduced a General Public License (GPL) for the GNU OS; Berkeley introduced the Berkeley Software Distribution (BSD) license
- 1997: The Open Source Definition was introduced

The Economics of Open Source Software (cont.)

Who Works on Open Source Software:
- Paid employees of "open source" companies
- Paid employees of traditional companies
- Volunteers
Relative Quality:
- Raymond (1999) argues that open source software should be of superior quality ("to many eyes, all bugs are shallow")
- Kuan (2001) shows that in 2 out of 3 product pairs she investigated bugs were fixed faster in the open source project

The Economics of Open Source Software (cont.)

Theories of Volunteer Motivation:
- Raymond (1999) - It is a form of gift culture based on altruistic motives (arising from abundance)
- Shapiro and Varian (1999) - Software is an information good exhibiting positive network/scale economies
- Bezroukov (1999) - It is a social phenomenon in which status depends on both contributions and social activities
- Lerner and Tirole (2002) - Skills may improve; Excitement of working on interesting projects; Contacts
Empirical Results:
- Hann et al. (2004) - People with high rank in the Apache project earn 14%-29% higher salaries whether or not their work is related
- Haruvy et al. (2003) - The promise of higher future earnings is an important driver
- Lakhani and von Hippel (2003) - The overwhelming driver is the need to solve their own programming needs
- Boston Consulting Group (2003) - Intellectual curiosity is the most important factor

The Economics of Open Source Software (cont.)

Motivations of Traditional Companies:
- Acquire expertise in a segment of the market that is complementary (e.g., IBM)
- Learn about the strengths and weaknesses of the process
- Learn about the strengths and weaknesses of the (presumably competing) product
- Generate good public relations
Motivations of Free Open Source Companies:
- Sell advertising
- Sell support
- Give away the razor but sell the razor blade (e.g., software is free but charge for hosting to generate recurring revenues)
- "Freemium" (i.e., some features are free and charge for others)
- Donations
- Improve lagging market share (e.g., Netscape's Mozilla)
- Reduce fears of small partners