[cross-posted at the TechLearning blog]

A few back-of-the-envelope calculations here (estimating conservatively when in doubt)...

A. Number of students and teachers

50 million public school students
+
3.3 million public school teachers (full-time)
=
53.3 million teachers and students

B. Cost per laptop (a regular laptop, not the OLPC laptop)

$1,993 average district cost per client computer per year [from the three One-to-One CoSN Total Cost Per Ownership (TCO) Case Studies]
x
1.5 (I'm adding 50% just to err on the safe side)
=
$2,990 average district cost per client computer per year (let's call it $3,000)

C. Total cost to give every student and teacher a laptop

53.3 million teachers and students (see A above)
x
$3,000 average district cost per client computer per year (see B above)
=
$159.9 billion (let's call it $160 billion)

D. Gross domestic product (GDP)

$13 trillion (United States GDP, overall)
x
3.4% (percentage of United States GDP spent on K-12 education)
=
$442 billion (amount in United States spent on K-12 education)

E. Percentage of GDP

$160 billion (see C above)
/
$442 billion (see D above)
=
36% of the overall United States K-12 education expense to give every teacher and student a regular laptop

$160 billion (see C above)
/
$13 trillion (see D above)
=
1.2% of the overall United States GDP to give every teacher and student a regular laptop

Obviously this is very rough, but hopefully it's also thought-provoking. It is highly possible that my numbers are incorrect somewhere. If you think I left something out or miscalculated, let me know. Also, of course, opportunities for savings abound (e.g., open source software, bulk discounts, buying OLPC laptops instead of regular ones) and those would have to be factored in as well.

So can we afford to give every child (and teacher) in America a laptop? You tell me...