Making cross-national comparisons using macro data Unit 2 Dave Fysh University of Portsmouth
We will cover • Problems of measurement in making comparisons • Where does comparative data come from? • Getting metadata • What series are harmonised? • Exchange rates as comparative data • Purchasing Power Parity exchange rates • PPP data series
Common Measurement Standards Comparing fruit diets Measured Ratio Bill Tanya Concept 1 apple, 4 oranges 4 apple, 1 oranges fruit 5 fruit 5 fruit 1 (equal) vitamin C 1 X 5 4 X 50 205mg 4 X 5 1 X 50 70mg 0.341 Tanya gets 34% vitamin C of Bill (Apple 5mg per 100g Orange 50 mg per 100g) MeasurementProblems .857 Tanya gets 86% vit. C of Bill Tanya’s apples weighs 200g Bill’s oranges weight only 50g Same sized fruit? Same standard of Testing fruit for vitamin C? Same period of time? Bill gets105 mg vit. C Tanya gets 90mg vit.C Same quality of fruit? Metadata and methodology Quality of measurement system
Making cross-national comparisons • ESDS International macro database publishes data series from 7 major IGOs and UK National Statistics time series data • 100,000s of series from all countries across the World Where do they come from? State statistical agencies IGO statistical departments Other international agencies Academic projects Commercial organisations What is an IGO? International body founded by treaty or similar arrangement e.g. UN, IMF, WB, OECD, ILO,IEA, European Commission What do the IGO statistical depts. do? Set quality standards Collate and present data Transform data if necessary Manage programmes to harmonise data
Some measurement difficulties in ESDS International macrodatabank series Do series refer to comparable: • Locations • measurement dates • splicing methods • base years and locations • methodologies • presentational formats • treatment of missing or estimated data • other matters Not always!
Can we find information to detect incomparability? Yes – in ESDS International there are a number of methods Look at the header of a series in report or even the item list Sweep the cursor over ‘numbered’ data cells Click the information icon Go to the ESDS Information Users Guide Go to the source documentation for the databank Information obtained from these sources is called metadata.
Harmonised series • OECD • IEA • Eurostat • IMF • ILO publish many harmonised series. If possible, use a single such databank when making comparisons
Concepts First look at exchange rates These transform incomparable national currency data to a common currency base But exchange rates also carry comparative information themselves: they are comparisons Consider the US dollar/pound sterling exchange rate: $1.80 per £ to $2 per £ Goes up from Tells us that £ holders can buy 11.1% more from US then previously for the same value of sterling Brits are therefore wealthier compared to Americans than before Note this is a relative statement – there is no information as to which country is absolutely wealthier
Limitations of exchange rate data Market rates are volatile All countries, permanently or from time to time seek to control market rates So a lot of contextual information is required to interpret market exchange rates or official exchange rates as comparative information
Using exchange rates to produce a common currency base for cross national comparison If all series denominated in local currencies could be expressed in a common currency then many sorts of comparison could be made. But not only are market rates volatile, they don’t seem to be able to capture difference in price level between countries. But they should if goods were fully tradable and currency fully ‘convertible’ e.g. A watch in the UK costs £20, the same watch in China 30 yuan At an exchange rate of 15 yuan per £, the watch costs only £2 to buy in China It seems likely that the exchange rate would change until the watch cost the same in China or the UK, or price of watches as measured in the two currencies equalised
What's wrong with market exchange rates as a basis for comparison? Flexible exchange rates don’t in fact equalise prices between countries because many goods are not traded between countries Governments don’t allow them to be traded the goods are inherently untradable For example – a haircut! The price of untraded goods will not be reflected in exchange rates So the cost of living in one country will not be valued on a common basis with another if market exchange rates are used to convert one currency into the other
Purchasing Power Parity Statisticians have found a way to avoid the problem of exchange rate imperfections A ‘PPP’ rate can be calculated from direct observations of prices, in local currencies of a similar bundle of goods in each country under comparison. So if the bundle of goods was – a meal, a pair of jeans , a haircut That might be observed to cost £25 in the UK and 100 yuan in China The PPP exchange rate would be 4 yuan per £ Provided the bundle of goods was sufficiently typical of consumption both in China an UK The PPP exchange rate would provide a base for a genuinely common currency unit of account for China and the UK Series such as GDP ‘deflated’ by this exchange rate could show real difference and similarities between the two countries
Purchasing Power Parity PPP exchanges rates and many series transformed to a common ‘unit of account are to be found in ESDS International databanks. Using PPP exchange rate series creates additional series for apparently similar variables. Have a look at the spreadsheet showing GDP measures for most countries in the World (LIMMD Materials Unit 3.8.4, first activity) Which is the correct measure?
Activities: Cross-national comparisons Form groups and • Read LIMDD Materials Unit 3, section 3.7.3 and other parts of 3.7 and undertake the activities • Read 3.8.3 and undertake the activities • Read 3.8.4, look at the first activity and undertake the second activity • Read 3.5.5 and try out all the ways metadata can be obtained from ESDS International macro databanks Report back to the whole session on what you learned (or didn't!)