Support for agricultural producers

The OECD has recently published Agricultural Policies in OECD Countries: At a Glance 2008 . One of the highlight of the report is that the main indicator the OECD studies to follow the level of support to farmers, the %PSE, is at its lowest level since 1986.

Support to farmers as a percentage of their income

1986: 38.8%
2007: 22.5%

Interestingly, the main explanation of this evolution is the rising food prices. To understand that, let’s see how agricultural support work.

A large part of the PSE indicator comes from Market Price Support (MPS). Market Price Support cover the money that the countries spend based on the difference between domestic prices and market prices. When the world prices raise, the countries spend less money on this item. In 1986, the OECD countries spent 168 USD billion on MPS, a number which is now down to 96.

For more information, see the Agricultural Policies of OECD Countries: At a Glance 2008 and check these data files:

• Evolution of the %PSE from 1986 to 2007
• Composition of the support from 1986 to 2007
• OECD PSE/CSE database

Road fatalities

Road fatalities per million vehicles

This graph shows the evolution of road fatalities. In OECD countries, these numbers tend to converge. There were only a couple of countries with less than 200 fatalities / million vehicle in 1990, now 24 out of 30 are under that treshold.

For some countries the drop can be drastic. See the Russian Federation, where the figures have dropped from 2556 in 1993 to 1177 in 2006. Within the OECD, see Hungary or Poland depicted in the graph.

Click on the graph to view how other countries have evolved. The complete dataset can be found here.

For more information about road fatalities check the Trends in the transport sector publication - the latest issue was released a couple of weeks ago.

How to access OECD data?

Among other things, the OECD is a first-class statistical agency. It produces data on a very wide variety of subjects - to name a few, economic indicators, gender issues, labour, migration or health - there are over 100 statistical publications per year. And its data are comparable across member countries.

So where to find it?

Everything published by the OECD can be found at sourceOECD, the Organization’s online library. Most things on SourceOECD require to have an account, but chances are you have access already: thousands of organizations, companies and universities have signed in and people who visit the site from there don’t need any form of identification. If you don’t have access, you can always request a trial account.

From SourceOECD, you can access OECD.Stat, the OECD statistical browser. This allows you to query all datasets and even to create tables across datasets. All the information about the data is also available directly from the dataset. By the way, until November 1st, 2008, OECD.Stat is not behind access control so anyone can access it with no restriction.

All OECD datasets are not yet available through OECD.stat, but all the databases are at sourceOECD.

Now all of OECD data are not in regular datasets. A lot of analytic books contain original tables or graphs, collected and compiled just for the occasion. This is where the “statlink” service comes in handy. Virtually every book produced by the OECD with tables or graphs is now powered by statlinks. At the bottom of every table or graph, you will find a link to the data, usually in excel format. It is that simple. Those links are DOIs so they will never rot and can always be used or cited with confidence.

By the way, all books can be browsed at the OECD online bookshop. So if you’re looking for a table or graph that you know is in a specific book, you can retrieve their statlink from there. If you have access to sourceOECD, you can also get the books in pdf format and merely click on the link to get the data you want.

In the next few months, when we will launch the new version of the sourceOECD site, we are going to roll out new (and free) data services : key tables, a list of the most useful OECD indicators, prepared by our statisticians and ready to use, and facts&figures, a collection of short stories around data.

CO2 emissions

CO₂ emissions
CO₂ emissions from energy use, in million tonnes

The International Energy Agency, linked with the OECD, publishes comprehensive energy indicators, which cover more than OECD member countries. Among them is the CO₂ emissions from energy use (that is, excluding CO₂ emissions from sources like wood or waste burning). The IEA prepares the yearly World Energy Outlook, a reference publication for long-term energy analysis.

This graph is based on the CO₂ emissions table presented in the OECD Factbook. The redish stack represents the OECD countries, the other countries are in the greenish stack. Click on the image to break down the chart and look at each country evolution.

Quick facts from the table:

Share of OECD countries in world CO₂ emissions:

1971: 66%
2005: 48%
2030: 36%*

CO₂ emissions of OECD countries in absolute values:

1971: 09.4 billion tonnes
2005: 12.9 billion tonnes
2030: 15.1 billion tonnes*

* according to World Energy Outlook projections

1971-2005 evolution from selected countries

Korea:                        +780%
China:                        +533%
India:                         +476%
Turkey:                      +421%

Sweden:                       -38%
Czech Republic:           -22%
Germany:                     -17%
UK:                               -15%
France:                         -11%

Buy the World Energy Outlook 2007

Patents

In order to compare the output of various countries’ research and development efforts - the inventions -, the OECD has come up with the concept of “triadic patent families”. They are a set of patents taken in all three major patent offices: the European Patent Office (EPO), the Japan Patent Office (JPO) and the United States Patent and Trademark Office (USPTO).

Patents
Number of triadic patent families per million inhabitants, 2005

This indicator varies greatly among OECD countries. The USA have the most triadic patent families in absolute values (16368 in 2005, over 30% of the world’s total) but, per inhabitant, have less than Japan or Switzerland. Other countries consistently above the OECD value are Germany, Netherlands, Sweden, Korea or Finland. On the other hand, 7 OECD countries have less than 10 triadic patent families per million inhabitants.

China, with 433 triadic patent families in 2005 (somewhere between Australia and Sweden in absolute values) has about 0.3 triadic patent families per million inhabitants.

The complete dataset can be found here. See also the work of OECD on patents.

Employment rates

Employment rates, 2006
People employed as a percentage of the 15-64 years old

The most-widely used labour indicator is unemployment rates. But employment is an even more interesting way to compare government policies. The proportion of people employed relative to the working age population varies significantly across the OECD countries, from below 50% in Turkey (mostly due to very low female labour participation) to over 85% in Iceland. Among G8 countries, these rates are low in Italy or France (around 60) and higher in anglo-saxon countries (70 or higher). Also note that China has stronger employment rates than any G8 country.

Factbook and trendalyzer

Trendalyzer is a software put together by the gapminder foundation, and which was later bought by google. It turns datasets into animated graphs which display as many as 5 variables at a time: dimension x, dimension y, colour, size of each point, and time.

In the words of Hans Rosling, who came up with the concept of trendalyzer, the goal of this software is to get rid of pre-formatted ideas about countries and to guide people on the way to evidence-backed thinking. Trendalyzer can be used to show convergence between countries that we thought different.

Take for instance this chart. It compares the health status of Denmark and Portugal. In the x axis, you’ve got infant mortality, and for the y axis, life expectancy. True enough, in the 1960s these two countries had very different health indicators. Portugal sat at the lower-right (”bad”) end of the chart. But over time, Portugal did catch up and in 2005, its figures were very close to the Danish ones. Click on the image to see the animation.

Trendalyzer can be used in a number of ways. Here is another. By setting time to the x axis, and by keeping the size of the bubbles minimal, it can also plot time series.

Here’s a comparison of the population growth rates of China, India and Japan. We see that China’s population grew very fast in the 1960s, but that this growth has now been put to a halt, while that of India is sustained and is only slightly decreasing now. Meanwhile, Japan’s population is growing slower and slower, and is now in fact decreasing.

Click on the image to play the animation.

OECD broadband statistics

Internet loves to measure itself, and there are quite a few companies that do a pretty good job at that: hitwise or internetworldstats, among others. The OECD angle is different. It collects data from official sources and presents them in a comparable way. This means that most indicators are available by country. Here, I feature two of them for a subset of the OECD countries - there were 42 in the last publication of OECD broadband statistics covering all 30 OECD member countries.

In these graphs, we learn that:

• on average in the OECD, there are 20 broadband subscribers per 100 inhabitants. But the contrast is stark between #1 Denmark with 35.1 and #30 Mexico, with just 4.3.
• The cheapest broadband plan in the OECD costs about $5 per month in Switzerland. However, in the Czech Republic, you can’t get broadband for less than $33. The most expansive broadband subscription is also in Czech Republic, over $300 per month!

Broadband penetration, december 2007
Broadband subscribers per 100 inhabitants

Range of broadband subscription rates, Oct 2007
Minimum and maximum values, US dollars per month, converted with PPPs

OECD population

OECD population vs. population of other countries and regions, 2006

This graph shows the population of the OECD countries compared to that of other large countries and regions. If you click on the graph, you can see this proportion evolve through time: from 26% in 1950, it is now down to 17%, and will be only 14% in 1950.

Click here for the whole data set in Excel. The data comes from the OECD Factbook 2008.

Update: the chart has been updated - the previous one was misleading about “Rest of the world”.

GDP per capita of OECD countries

GDP of OECD countries, 2006
Thousands of US dollars, converted using PPPs

This graph ranks the different OECD countries according to their GDP per capita. Due to the complex nature of this aggregate, slight differences between countries should not be taken into account.
The numbers are converted using purchasing power parities, in other words they take into account the price differences in the various countries.