In my suburban area - I could get a mortgage for a $100k house with at least 2br for about $500/month.
Of course people at Google probably don't get paid exactly $100k but keeping in mind a good sized chunk of that is just going toward living there.
[1] - http://www.calculator.net/take-home-pay-calculator.html?cann...
[2] - http://www.zillow.com/homes/for_rent/Redwood-City-CA/house,c...
To re-emphasize, 1k/month in lower rent is worth 15-18k of salary.
I have friends paying more than $2k/month more than me, which is around a negative 40k raise (I live in San Francisco, but share).
It's very easy for humans to misinterpret recurring costs.
Assuming $2k/month is $40k - that is almost half of $100k. So almost half your salary is going towards just rent.
Let's say you made $50k in a suburb area - using your numbers you would only pay $10k/year for a mortgage ($500/month) - with only $20k difference in salary.
$100k in CA ~ $60k $50k in suburbs ~ $40k
Realistically you would probably make $60-$70k in suburb areas which, according to your numbers, would make the same if not more AND own a house.
Of course this is ignoring other factors like food, utilities, bills and taxes.
That all depends. Many comp sci students are brainwashed to think that if a salary isn't at least 100k then it's not worth their time.
The fact is that it depends on location and cost of living.
For example - if you get hired to work for Google in their HQ in CA then you are guaranteed to get at least 100k salary. However, 90% of that will go towards food/rent (you will most likely rent as houses cost up in the millions...) - which making $60k in a smaller suburb area (where housing is in the affordable $50-$200k range) would arguably be the same net income as if you were working at Google.
I will say that you will earn more in the private sector - but I wouldn't use miserable to describe government salaries (more like underpaid). When I worked for the man it was defiantly an interesting experience but it was a very relaxed environment/startupy feeling.