Transcript Comments by Evan Kraft

Comments on “Socialist Growth
Revisted: Insights from
Yugoslavia”
Evan Kraft
American University
YES Conference, Dubrovnik June 2015
Contributions of the paper
• Model-based assessment of determinants of Yugoslavia’s economic
growth
• Brings up to date older growth-accounting work such as Sapir (1980)
• In line with recent literature reflecting on Communist economies’
ability to sustain fairly rapid growth for a certain period of time (eg.
Broadberry and Klein 2011 for Czechoslovakia)
• Potential contribution to understanding role of economic institutions
Findings of the paper
• TFP growth accounts for increasing share of growth over time
• Growth slowdown after 1979 mainly due to labor wedge
The model: Chari, Kehoe and McGrattan
(2007)
• Decompose deviations of output growth from trend into four frictions:
•
•
•
•
Efficiency wedge (time-varying TFP growth)
Labor wedge (labor income tax)
Investment wedge (investment tax)
Government consumption wedge (government consumption and net exports)—in
this paper called Income wedge
• Argument is that factors affecting growth can be bunched into these
categories for further analysis
• Acknowledge that these are reduced forms, so that the wedges are not
representations of the underlying model.
• Example: an economy with sticky wages and monetary shocks maps into a growth
model with labor wedges.
Christiano and Davis (2006): critique
• Show that wedges may not be independent
• Credit market imperfections that generate investment frictions may also increase labor
wedges, for example
• If so, identification of origins of frictions is unclear
• Also, the form of the friction may matter
• Get very different results for an “ROE tax” rather than a direct tax on investment. In the
case of the US Great Depression, credit market frictions of this form, a la Bernanke,
Gertler and Gilchrist, can explain a great deal of output variation, while the direct tax
formulation used by Chari et al does not explain much output variation
• (My own comment—growth really is not endogenous in this approach. The
trend rate of growth of labor augmented productivity growth is a fixed
parameter. Only wedges can change)
What happened to Yugoslav growth?
• Key turning point: 1979-81
• Economic event: sovereign debt crisis
• How should this show up in the model?
• Obvious effect should be difficulty accessing credit—investment wedge?
• Research by Vodopivec (1989) and Kraft and Vodopivec (1992) shows large
importance of “soft-budget” finance through
• Discretionary tax and subsidy policies
• Inflationary gains and losses, reaching great importance as inflation rises in the 80’s
• Regional redistribution—enterprises in Slovenia are net payers, all others net receivers,
with receipts correlated with income per capita
What reforms were undertaken after 1979?
• Few systemic reforms until 1988
• Initial austerity, export at all costs mentality (potentially decreasing
efficiency)
• Policy vacillation
• Sharp decrease in personal incomes, rise in unemployment, rise in
inflation
• Increase in labor unrest, especially in the second half of the 1980’s,
but not clear how much effect this had on output
• How should this show up in growth accounting?
Periods and breakpoints
• Several other years could perhaps be examined as breakpoints
• 1961: end of grant funding from West
• 1965: major economic reform, including major effects on labor supply (freedom to
travel leads to outmigration) and labor demand (greater decision-making power for
companies probably decreases labor demand).
• 1971: reaction against reform, beginning of new “self-management agreement”
period
• 1974: new Constitution
• 1976: Law on Associated Labor—major new labor code
• Could be worthwhile to examine these sub-periods
• Also, 1970’s are critical
• Failure to adjust to energy shocks
• Increasing foreign debt
• Similar to Poland more than Czechoslovakia
Labor market distortion: history
• Firms were labor managed for the whole period under study
• Open unemployment is seen from the mid-1960s at least
• Self-management agreements created detailed wage determination
structures in 1970s
• Set by bargaining process
• But strong Communist Party influence, pressure to limit wage disparities
• One would think that distortions would have increased substantially when this
system was introduced in the mid-1970s
• But in the 1980’s, firms were required to increase hiring based on numbers
given in Republic-level self-management agreements
• Would this account for labor wedge? Some deeper digging here could be
very helpful.
Regional issues
• Aggregate data may be misleading in a country where GDP/capita in
the most advanced republic was 7-8 times that in the least advanced
autonomous region
• Productivity differences associated with skill may vary: a PhD in the
more advanced regions may be substantially more productivityenhancing compared to secondary education than a PhD in the lessdeveloped regions
• In other words, returns to schooling may have varied considerably
across Yugoslavia
• This could pose issues for the conversion of educational attainment
into Mincerian human capital
To what extent was Yugoslavia a market
economy, and how does this affect modeling?
• Government involvement in investment decisions decreased after 1965—
central General Investment Fund closed
• But Republic level influence remained strong, perhaps underplayed in
paper
• Major projects clearly decided on non-economic basis by Government
• Belgrade-Bar railroad
• Aluminum, steel factories in each Republic held up as major new source of
inefficiency post 1971
• Can we distinguish between
• Inefficient investment decisions taken directly by Government--policy
• Tax/regulatory frictions--system
• Influence of worker-management--ownership?
Potential payoff: different story
• Broadberry and Kline, along with many others, point to inability of
socialist systems to make transition from “extensive” to “intensive”
growth
• But Yugoslav case shows inability to continue technological imitation
• Part of this is macro-level inability to manage foreign borrowing (a
common story in the 1970’s)
• Part of this is micro-level inefficiencies (labor and capital allocation
mechanisms)