Historical National Accounts The Case of Argentina

1
Historical National AccountsThe Case of Argentina

• Luis Bértola

2
This presentation

• Is about pre-1935 GDP estimates. The presentation
  will be divided in three sections
• 1900-1935
• up to 1900
• benchmarks for international comparisons.
• Post 1935 national accounts are concisely
  reviewed in Ricardo Gabriel Martínez (1998),
  Recopilación de Series Históricas del Producto y
  del Ingreso.

3
1900-1935

• All estimates of Argentine GDP 1900-1935 are
  related to ECLAs estimate El Desarrollo
  Económico de la Argentina (Anexo).
• The previous assessment includes IEERAL
  (Domenech), della Paolera Taylor, Hoffman,
  Maddison.
• An exception is Cortés Conde, who is responsible
  for the only significant revision of this series.

4
Cortés Condes (RCC) revision(following Eduardo
Salazar no title, no date)

• Cortés Conde produced two estimates of which only
  one is available for me that of 1994. In 1997 he
  published La Economía Argentina en el Largo
  Plazo, which reproduced a new series (see graph)

5
(No Transcript)
6

• RCC. ECLAs shortcoming uses the economic
  structure and relative prices of 1950 to estimate
  growth in the primary-product-based, export-led
  period. Thus, export-led growth was
  underestimated. ECLA was aware that the impact of
  manufacturing on total output could be
  overestimated in earlier periods.
• Instead, RCC uses the economic structure arising
  from the 1914-year Census for 1900-1935.
• According to Salazar, if the weights used by RCC
  and ECLA are inter-changed, the results only
  affect growth volatility, but not the trend.
  RCCs weights with higher agricultural shares,
  produce more volatile results.
• Thus, differences have to do with the sectoral
  series.

7
(No Transcript)
8

• The main difference between ECLA and RCC is the
  series for manufacturing output.
• Growth is lower according to RCC (4,2 vs 5,4).
• Growth is distributed differently faster than
  ECLA prior to 1913 (7,8 vs 7) more slowly
  thereafter (4,1 vs 8,4).

9
Why?

• Because of the different structure of the
  industrial sector used by Cortés Conde. The share
  of the dynamic sectors is much lower in his
  manufacturing structure. See next.

10
(No Transcript)
11
1900-1935 in short

• It seems that a better estimate should arise by
  using RCC sectoral weights and ECLAs sectoral
  series. Probably RCCs series for the public
  sector may be used instead of ECLAs, but its
  impact on average growth is low. The expected
  result is a similar trend than ECLAs, but with
  higher volatility.
• In any case, a better alternative is to use a
  Divisia Index.

12
1875-1900

• RCC uses the 1895-year weights (according to the
  Census) in order to estimate 1875-1900. The
  growth rates achieved are very high, psoibbly
  reflecting the higher share of the dynamic
  sectors in the end years 0f 1895-1900.
• The estimate by Della Paolera Ortiz may be a
  better alternative so long. However, as it is
  based on aggregates, this estimate does not
  constitute a path to be dependent on if GDP
  estimates are to be improved.
• Can Angus explain his 1870 data point? His
  1900-1935 series is that of CEPAL his 1890 is
  that of RCC his 1870 is compatible with RCC, due
  to the crisis of the early 1870s.

13
Argentine GDP, 1870-1900
14
Benchmarking

• We are all dependent on different benchmarks for
  very late periods.
• Earlier benchmarking faces the problem of
  information coverage we know there are large
  sectors of economic activity around the 1900s
  which are not accessible. Projecting later
  benchmarks by growth rates is what we have done,
  but the risks are high. Is it less riskier to
  attempt to produce contemporary benchmarks?
  Probably, some experiments may be done with
  smaller economies, as that of Uruguay.
• Possible benchmarks for Argentina are 1895, 1914
  and 1935.
• What do our benchmarks say today about the 19th
  Century? See next.

15
Relative per capita GDP
Argentine per capita GDP 1830-1870 projected
with help of Salvatore Newland 2003.
16
What I see

• 1830 was Argentine per capita GDP as low as half
  that of the USA?
• 1830-1870 no change in relation to the UK
  lagging behind AU moderate catch up with USA
• 1870-1913 Did Argentina really catch up with the
  US? From 50 to 72? Or was the starting point
  higher?
• In any case, she never came closer than 83
  (Australia), 79 (UK), 72 (US).
• My guess higher original levels, not a so
  significant catch-up 1870-1913, lower growth rate