I'm 37. I'm willing to believe there is are two decent reasons for some apparent ageism. I think #2 here is probably in the not-safe-to-share-in-public category, but honest discussion requires embracing one of those every now and then, so here goes:
a) When supply of well trained programmers is increased, it happens with 22 year olds, who are graduating from college. There is a 30+ year lag on those folks becoming old programmers. To hire a 50 year old programmer, he would either have had to graduate back in 1985 when there were a lot fewer people studying programming, or else changed careers at some point. And a lot of people don't change careers. There simply aren't as many older programmers available because it's a new field as these things go.
b) I've wasted enough years on dead-end tech that I'm more cautious about what I spend my time on now. Things I have become an expert in that are now useless to me and probably not valued by an employer:
TCL, Oracle 8i and PL/SQL, Ada, Java 1.1/1.2, EJB, Java AWT & Swing, Struts, Delphi, Flex, Paradox, Access + VBScript, ASP.NET 1.0, ODBC/OLE/ADO.NET, Linq to SQL, Subversion/MKS Source Integrity/Rational Clearcase/Perforce, differences between bash/ksh/csh scripting, awk+sed, perl, prototype.
And those are just the dead-ends that I regret exploring, because I think I will never use them again and because I could have spent that time learning something more useful. I know the time wasn't completely wasted, but time spent on prototype would have been better spent on jquery, time spent on linq to sql would have been better spent on nhibernate, etc. Reflecting on that leads to a bit of analysis paralysis when looking at TodoMVC to pick which next framework to invest time and effort in - will this be another flash in the pan, is this nothing but Fire and Motion (1)? I expect many employers would probably benefit more from a team of people who lack that perspective, because they'll be giving up more of their evenings and weekends on things that might pan out. I still do that, but a bit less these days, and I wait a bit longer before I dive in. Some might value my caution, but they probably aren't starting a bootstrapped or VC backed change-the-world-through-web-programming moonshot business.
a) When supply of well trained programmers is increased, it happens with 22 year olds, who are graduating from college. There is a 30+ year lag on those folks becoming old programmers. To hire a 50 year old programmer, he would either have had to graduate back in 1985 when there were a lot fewer people studying programming, or else changed careers at some point. And a lot of people don't change careers. There simply aren't as many older programmers available because it's a new field as these things go.
b) I've wasted enough years on dead-end tech that I'm more cautious about what I spend my time on now. Things I have become an expert in that are now useless to me and probably not valued by an employer:
TCL, Oracle 8i and PL/SQL, Ada, Java 1.1/1.2, EJB, Java AWT & Swing, Struts, Delphi, Flex, Paradox, Access + VBScript, ASP.NET 1.0, ODBC/OLE/ADO.NET, Linq to SQL, Subversion/MKS Source Integrity/Rational Clearcase/Perforce, differences between bash/ksh/csh scripting, awk+sed, perl, prototype.
And those are just the dead-ends that I regret exploring, because I think I will never use them again and because I could have spent that time learning something more useful. I know the time wasn't completely wasted, but time spent on prototype would have been better spent on jquery, time spent on linq to sql would have been better spent on nhibernate, etc. Reflecting on that leads to a bit of analysis paralysis when looking at TodoMVC to pick which next framework to invest time and effort in - will this be another flash in the pan, is this nothing but Fire and Motion (1)? I expect many employers would probably benefit more from a team of people who lack that perspective, because they'll be giving up more of their evenings and weekends on things that might pan out. I still do that, but a bit less these days, and I wait a bit longer before I dive in. Some might value my caution, but they probably aren't starting a bootstrapped or VC backed change-the-world-through-web-programming moonshot business.
(1) http://www.joelonsoftware.com/articles/fog0000000339.html