1
00:00:01,556 --> 00:00:05,622
One of the problems with corruption is
that sometimes you get the wrong outputs,

2
00:00:05,622 --> 00:00:07,165
the wrong kinds of output.

3
00:00:08,638 --> 00:00:11,252
Let's compare two kinds 
of government projects.

4
00:00:11,338 --> 00:00:16,267
The first kind of project is local,
small-scale, labor-intensive.

5
00:00:16,330 --> 00:00:18,823
Imagine educating children in a village.

6
00:00:20,228 --> 00:00:22,612
The second kind of project
is very glamorous.

7
00:00:22,612 --> 00:00:24,204
It's capital-intensive.

8
00:00:24,309 --> 00:00:27,244
It has a lot of hard-to-value
inputs and outputs.

9
00:00:27,312 --> 00:00:30,660
Imagine, for instance,
government building a large dam.

10
00:00:31,783 --> 00:00:35,578
With which kind of project
is it easier to skim money off the top?

11
00:00:35,814 --> 00:00:38,352
Well, when the inputs
of the project are opaque,

12
00:00:38,352 --> 00:00:40,939
when the project lacks transparency,

13
00:00:40,939 --> 00:00:45,215
and when there are large expenditures
running both in and out of the project,

14
00:00:45,215 --> 00:00:48,622
in that case, it's probably easier
to have bribery and corruption.

15
00:00:48,622 --> 00:00:50,299
So what's the result of this?

16
00:00:51,105 --> 00:00:54,889
A lot of governments will build
too many large-scale capital projects

17
00:00:54,889 --> 00:00:57,251
with hard-to-value inputs and outputs

18
00:00:57,251 --> 00:01:01,732
and perhaps too few small-scale,
local labor-intensive outputs,

19
00:01:01,732 --> 00:01:04,002
such as educating children in a village.

20
00:01:04,002 --> 00:01:07,079
And again, this is a distortion
driven by corruption.

21
00:01:08,616 --> 00:01:11,361
This argument comes from a paper
called "Corruption"

22
00:01:11,361 --> 00:01:13,271
by Shleifer and Vishny.