One XBox per child

Alex Tabarrok reports what happens when you give computers to children in Romania. Computer homework does not change. Computer use goes up due mainly to… computer games.

with all that game playing going on, the authors find that the voucher program actually resulted in a decline in grades although there was also some evidence for an increase in computer proficiency and perhaps some improvement in a cognitive test.

In fairness to the one laptop per child people, who could have predicted such a result?

Oh wait.

6 thoughts on “One XBox per child

  1. The problem with the ideology (constructivism) behind OLPC is that kids not doing their homework is a positive result. Their homework is the wrong thing for children to be doing, they claim, and whatever kids are doing on computers that is self-directed is a step in the right direction. Thus, OLPCs claims about the goodness of OLPC can never be falsified. If the kids use the computers great! If they don’t use the computers, well then its someone else’s fault for not helping the kids use the computers.

    FYI, Ray Fisman reported on this case about 18 months ago in Slate. An article I wrote for Miller McCune ( has more details about that study and others all of which suggest that OLPC is a colossal waste of money.

  2. Tim:

    Is kids not doing their homework necessarily a negative result, though? When I was a kid I got familiar with a computer by playing Oregon Trail. Today I’m an editor who’s thankful he got computer literate on the (Oregon Trail-equipped) machines his school brought in in the 1980s.

    It’s clear I’m coming in the middle of a conversation, but this minor drop in homework in completion seems a pretty light price to pay.

  3. Hacking and playing games on english language computers arguably builds more human capital than a lot of primary school systems run by deficient states. Maybe not in post-communist Romania, but I think this result isn’t so damning of computers as your post implies.

  4. who could have predicted it? anyone who’s ever worked in a civil service where people are paid rotten salaries – when staff get allocated new computers, the sequence of events tends to be:

    1) more efficient, better presented work
    2) brief (or not so brief) down period with little work, boredom alleviated by playing solitaire
    3) discovery of spider solitaire. work rate plummets, even when work is there to be done.
    4) discovery of tumblebug. work ceases altogether.
    5) discovery of youtube, music downloads and other filesharing softwares – computer compromised by more viruses than a hospital
    6) computer breaks down. work resumes by hand. productivity reaches medium level again.

    and repeat.

  5. I can’t see that the OLPC XO machines are involved here. All I can find in Google is that in 2007, the Romanian government turned down XOs

    We probably need more research or facts. XO developers made a different product: they aimed to build what the trade calls a “rich experience” around classwork, classroom experience, class collaboration, and technical prowess. The Romanian project seems to have merely subsidized ordinary PCs and ordinary commercial PC software — of which the bulk is certainly games.