Evolution of the System of National Accounts (SNA) and its Impact

W(h)ither the SNA? Anne Harrison

A depressing moment in 2007 When will the next revision take place? Tensions between those wanting change and those resisting it could militate against another update being successful How did this come about? 2

The early days of the SNA Rebuilding economies after WWII Marshall plan return to normal economic health, policy directed against hunger, poverty, desperation, chaos Beveridge plan to address five giant evils squalor, ignorance, want, idleness, and disease 3

The 1970s Opec and the oil price increase inflation Undermined IO models Breakdown of the Bretton Woods agreement on fixed exchange rates Rise of monetarism Computers become ubiquitous Microsimulation 4

Consequences End of golden age for SNA SNA seen as Keynesian economists lost interest Reduction of statistical resources Changing priorities in statistics management not technical skills 5

Why did SNA not die? Increasing use as an administrative tool UN committee of contributions World Bank IDA conditions Focus on GDP per capita Rise of ICP 6

1993 Revision Incorporated Prices and volumes Financial accounts Balance sheets Satellite accounts Downplayed but did not remove IO Increased emphasis on the system as the overall framework for economic statistics 7

Major concerns What about developing countries? Did not want SNA Lite Centrally planned economies? Even before collapse of iron curtain moves to reconcile MPS and SNA, then MPS overtaken by events 8

Post 1993 City groups emerged Household income Environment Not everyone as enthusiastic about the idea of the SNA as the overall framework as the national accountants ILO prices; informal sector; production boundary 9

ESA95 Becomes a directive Emphasis on cross-country comparability Focus on particular aggregates and ratios Concentration on most recent observations and timeliness Greatly reinforced the use of national accounts for administrative purposes 10

Tensions during the 2008 revision Europe vs. others Administrative uses vs. economic analysis Cross-country vs. time series Accuracy vs. looser, wider picture Observation vs. data from models 11

Two extremes GDP is discredited as a measure of wellbeing or even as a policy goal GDP serves a number of purposes very well I can tell you what you should add and why There are good reasons why it cannot or should not be done 12

The SNA withers? If nothing is done to answer demands to reach out to well-being and other concerns, the risk is the SNA becomes increasingly marginalised if not irrelevant At best becomes a heritage statistical system In the limit it withers and dies 13

Whither the SNA? The demands for change are numerous, not always completely thought out, sometimes may be in conflict with one another How to decide which to pursue, in what order and when? Where is the SNA headed whither? 14

A personal position Should the SNA survive and be revised YES Do I sympathise with the reasons why some wish to consolidate the present system rather than enlarge it Yes Do I think anything and everything should be incorporated into the system NO Can indicators be a useful supplement - YES Should some extensions be considered - Yes 15

First steps to increase the relevance of the SNA Make better use of what we have Use GNI as well as GDP and make the difference clear Use NDP and NNI when appropriate Disaggregate households Not just NPISHs but institutional households also Do more with corporate sub-sectors 16

Making changes Instead of nominating issues, consider how an update is organised Who is involved Why it takes so long What might help ease the process 17

How does the SNA change? ISWGNA sets an an agenda The expert group discusses Ultimately the UN Statistical Commission approves Note that only official statisticians involved Process is not very transparent 18

Why does it take so long? There is no such thing as a small change to the system In the past the topics for discussion have been agreed at the outset but the implications for the system as a whole were not worked through takes a lot of time Need for widespread consultation if the SNA is to remain suitable for all countries 19

Advancing the case for a change Need to consider the proposal in the light of the accounting structure of the SNA Not everyone making a proposal is very familiar with the system and the possible ramifications of a suggested change SNA Para 1.1 - the system is comprehensive, consistent, integrated Does the proposal fit within these constraints Does it follow the SNA accounting rules, classifications, valuation conventions 20

Example 1 Moving an item used in production from intermediate consumption to be a fixed asset Increases GDP; Increases NDP much less Changes the asset boundary Does not change the production boundary Data not observed, would need to be modelled Relevant more to industrialised than to developing countries 21

Example 2 Move an item from final consumption to be a fixed asset Does not change the asset boundary Does change the production boundary Data readily available May have much greater impact on developing countries than industrialised 22

Suggestion Based on the way in which the (commercial) International Accounting standards are revised Two part system Part 1 elaborate the proposal and consequences ask for comments Part 2 consider proposal in light of comments and decide whether to proceed 23

Part 1 A proposal for change (e.g. the papers before this meeting) should be accompanied by a document explaining in detail the implications for the system Would need constructive engagement between proponents of change and SNA experts These should be posted on the internet soliciting comment 24

What sort of implications? Does this proposal change the level of GDP or simply increase analytical depth? Does it change the growth rate of GDP? Is any change steadily increasing over time? Or does it smooth fluctuations over time? Does it change NDP as much as GDP? Does it change NNI? 25

Does this proposal.. Apply to industrialised and developing countries equally? Require changes to the production or asset boundary? Fit within the classifications of the system? Use a valuation method that can be used for all parties involved? 26

Feasibility Does the necessary data exist ? Could it be gathered by minor modifications to existing data collections ? Does it need a new data source? 27

Timing Would incorporation of the proposed change impact the resources needed to produce quarterly accounts on the current schedule? Do the results actually need to be presented quarterly or would annually or even less frequently be sufficient? 28

Integration Does your proposal really need incorporating formally into the SNA or could it simply be associated with it? Quicker to achieve Could still appear in the same publication 29

Part 2 A review body (similar to an editorial board for a journal) helps ensure all aspects are covered and and correct and in the light of comments and the merits of the case judges whether the proposal should go forward to the ISWGNA Should involve both economists and statisticians 30

Looking forward We must ensure the SNA does not wither and die We need to find a way to decide whither it is headed that as many people as possible can contribute, see how the decisions are made and understand why one path rather than another has been chosen 31