Barter final



Yüklə 152,43 Kb.
Pdf görüntüsü
tarix27.03.2018
ölçüsü152,43 Kb.
#34928


1

Barter and Non-Monetary Transactions in Transition Economies:

evidence from a cross-country survey

1

Wendy Carlin, University College London and WDI

Steven Fries, European Bank for Reconstruction and Development

Mark Schaffer, Heriot-Watt University and WDI

Paul Seabright, University of Cambridge and CEPR.

December 1999



Abstract

This paper reports the findings of a survey of more than 3,000 firms in 20 transition

countries. It shows that barter and other non-monetary transactions (including the use of

bills of exchange, debt swaps, barter chains, and the redemption of debt in goods) are an

important phenomenon in Russia and Ukraine. Contrary to what is commonly believed

they are not negligible in Central and Eastern Europe. The causes and consequences vary

significantly between countries, but several conclusions emerge strongly. First, barter and

other non-monetary transactions are associated in all countries with financing difficulties

for firms. They appear to be helping to assure liquidity in an environment in which

contract enforcement (including tax enforcement) is uncertain. Secondly, the use of these

mechanisms is not significantly related to the restructuring and performance of firms that

use them in most countries except Russia. Thirdly, in Russia the nature of non-monetary

transacting is importantly different from elsewhere. It is much more associated than

elsewhere with market power and limited trading networks. It is also more costly in terms

of restructuring and performance. Firms that barter are less likely to improve their

existing products, probably because barter enables them to dispose of otherwise

unsaleable goods. They are also more likely to engage in internal reorganisation of a kind

designed purely to service existing barter chains. Internal reorganisation is strongly

associated with improved performance for firms that do not barter, but is unrelated to

performance for firms that do. Overall, in Russia but not elsewhere the findings are

consistent with the hypothesis that economic disorganisation, in the sense of Blanchard &

Kremer (1997), means that barter and other non-monetary transactions are both more

likely to occur and more damaging when they do occur.

                                                          

1

 Forthcoming in Paul Seabright (ed): The Vanishing Rouble: Barter networks and non-monetary



transactions in former Soviet societies, Cambridge University Press.  The views in this paper are solely

those of the authors and not necessarily those of the EBRD or other institutions with which they are

affiliated.



1

1.

 

Introduction

The persistence of barter transactions over a number of years in complex

industrialised economies has been one of the most puzzling paradoxes of the transition

from central planning to market organisation. Historically barter has characterised

relatively simple societies with a comparatively undifferentiated division of labour. It has

also been observed in more complex societies in the aftermath of serious crises such as

wars. For example, complex chains of bilateral exchanges of goods between firms and

payment in kind to workers were prevalent in the western zones of post-war Germany

between 1945 and mid-1948. In the context of a high level of uncertainty about the future

of the economy, with the collapse of the Nazi command economy, the freezing of prices

and wages at their 1936 levels and extensive controls over interregional trade, there was

an extreme shortage of goods. An assessment at the time captured the essential role of

barter in this episode: “Where neither trading for money nor redistribution of goods by

political authority, alone or in combination, can ensure a reliable division of labour,

bilateral exchange seems to be the safest line of economic retreat” (Mendershausen 1949,

pp. 657-8). The improved functioning of the costly and cumbersome barter mechanism

enabled production to recover from less than 20% of the 1936 level in mid 1945 to 50%

by the end of 1947.

But whenever barter has been observed in such situations of crisis it has been short-

lived. In mid 1948, there was a currency reform combined with the lifting of price and

wage controls. Barter and side-payments in kind vanished. There was also a clear shift in

the nature of recovery to a dynamic process of growth vividly displayed in the jump in

investment, the radical reorganisation of production processes and the introduction of

new products (Carlin 1989). By the time of the currency reform, it was clear that recovery

was to be encouraged, a market economy was to be restored and that private ownership of

firms would remain largely intact. The episode of large-scale barter was ended abruptly

by the introduction of functional money and price and wage liberalisation. The “normal”

incentives of a market economy took over.

This episode raises the question why barter has persisted and indeed expanded in

transition economies after prices were liberalised, and why it has continued even in the

context of reasonable macroeconomic stability. Presumably other characteristics of

transition economies have interfered with the rapid establishment of “normal” market

economy incentives and practices. Marin and Schnitzer’s (1999) analysis suggests that a

key differentiating characteristic may be that the nature of the output collapse in post-war

Germany and in the transition countries was different.

The degree of “disorganisation” in terms of the disjunction between the relationships

of suppliers and purchasers of inputs in the planned economy and those sustainable in a

market economy appears to have been much greater in the transition economies than in

post-war Germany. The pattern of trades in postwar Germany seems to have been

motivated by producers trying to maintain supplier and customer relationships (Stamp

1947). In transition economies major changes in supplier/customer relationships were

required. When planning collapses and leaves behind bilateral monopoly relationships




2

between input suppliers and purchasers, there is great scope for “hold-up” problems and

stalemates. As a result, production chains collapse (Blanchard & Kremer 1997). The

collapse is greater where new entrants and foreign suppliers are unable to play a

substantial role.

Marin & Schnitzer argue that trade credit expands in transition economies to help

offset the bargaining power of the input supplier. However, in the context of uncertain

contract enforcement trade credit is highly risky. Barter may therefore help the process of

output recovery by allowing trade credit to be collateralised in the form of the borrower’s

own output. This will allow output to be maintained in a world of disorganisation, though

it may have other more long-term costs. Normally these costs include the fact that firms

find themselves having to accept and re-sell products in the trading of which they have no

comparative advantage. But this particular cost may be lower when - as a symptom and

by-product of disorganisation - trading networks are limited and firms operate in

informational “islands” (Seabright, 1999); trading partners may be able to pool search

costs without sacrificing comparative advantage. In these conditions barter may have

fewer drawbacks than other responses to the problem of limited creditworthiness.

In post-war Germany output recovery in a barter-dominated economy before the

currency reform was often seen as remarkable (Abelshauser 1975). It is nevertheless very

clear that only when the functioning of market processes was fully restored did dynamic

future-oriented restructuring take place.

In understanding the prevalence of this expansion of barter in transition countries and

particularly in Russia, Ukraine and other CIS countries, it is important to bear in mind

that what is commonly referred to as “barter” in the Russian and Western literature on

these countries is not “barter” as conventionally defined. The New Palgrave Dictionary of

Economics, for example, defines barter as “a simultaneous exchange of commodities …

without using money.  It is thus a form of trade in which credit is absent or weak….”

(Hart 1987). The Russian term barter, however, encompasses not only the exchange of

goods for goods, but also the exchange of goods for debt.  If, for example, a firm pays for

a purchase of inputs with a bill of exchange (Russian veksel, from German Wechsel), then

this is barter (Russian), but it is certainly not barter as conventionally defined in the

English-language economics dictionaries.  Indeed, Commander & Mummsen (1999)

show that most of what Russian firms refer to as barter is not in fact what economists

would term barter, i.e., the exchange of goods for goods; it is rather payment for goods

using non-monetary methods and instruments, i.e., debt.

There is, however, an important difference between the use of bills of exchange and

other debt instruments in capitalist economies and the countries of the CIS. When a bill

of exchange is redeemed in the CIS, typically the holder of the claim on the issuing firm

is not the customer that initially accepted the bill as payment. It is a different firm that has

purchased or otherwise acquired the bill (though precisely how often this occurs is

unclear).  Furthermore, the bill of exchange may often be redeemed by the issuing firm

not in cash or equivalent, but in goods produced by the issuing firm.  It is this last feature

that most clearly distinguishes the use of bills of exchange in CIS countries from the way




3

they have been used in capitalist economies.  The use of debt offsets in CIS countries, the

third main form of barter (along with bills of exchange and “barter” in the standard sense

of goods-for-goods) is conceptually similar. In the multilateral debt swaps observed in

CIS countries, by contrast, debts are essentially redeemed in goods, not cash.  This is not

barter as conventionally defined, but it is a close cousin.

This paper analyses the transactions of firms conducted using non-monetary methods

and instruments: exchange of goods for goods, payment using bills of exchange, debt

swaps, redemption of debt in goods, etc. – “barter” as understood in the Russian sense of

the word.



2.

 

Empirical findings

2.1

The nature of the survey

A large survey of enterprises in twenty transition countries was conducted in the early

summer of 1999 by the EBRD and the World Bank, and its provisional findings have

been published in the EBRD Transition Report 1999. The aim of the survey was to

investigate how enterprise restructuring behaviour and performance were related to

competitive pressure, the quality of the business environment, and the relationship

between enterprises and the state.  The survey included approximately 125 firms from

each country, with the exceptions of Poland and Ukraine (over 200 firms) and Russia

(over 500 firms).

One question on barter was included: “What share of your firm’s sales are now (and

three years ago) conducted in barter, offsets or bills of exchange (money surrogates)?”

The six possible answers were one point (exactly zero) and five intervals of varying size

(1-10%, 11-25%, 26-50%, 51-75%, 76-100%).  The econometric technique we use when

this is our dependent variable is interval regression (StataCorp, 1997). The advantage of

interval regression is that the coefficients on the exogenous variables can be interpreted

as if ordinary least squares were being applied to a continuous dependent variable; e.g.,

the coefficient on a dummy variable will give the impact in percentage points on the

share of barter.

Since only this question was asked, we have no way of checking if “barter, offsets

and bills of exchange (money surrogates)” were interpreted in the same way by different

firms and in different countries. There may be substantial cross-country differences of

interpretation (for instance whether trade credit is included in the definition). While this

places some limits on the interpretation of the findings from the survey, the breadth of

other information collected presents an unparalleled opportunity for exploring the causes

and consequences of barter.

The full sample size was 3,125 firms. Sampling was random from the population of

firms in each country, with the exception of minimum quotas for state-owned firms and

large firms. We omitted from the analysis firms with missing information, leaving us with




4

3,079 firms.  The sample is dominated by small and medium-sized enterprises; almost

half the sampled firms employ fewer than 50 persons, and less than 10% employ more

than 500.  Half the firms in the sample are newly established private firms, 10% were

privatised to insiders, 25% were privatised to outsiders, and 15% remain state-owned.

Firms in the industrial and service sectors are roughly equally represented, each

accounting for 40-45% of the sample, with agricultural firms making up the remainder

(14%).  Most firms were located in either large cities or national capitals (30%) or in

medium-sized cities (32%), with the rest in towns and rural areas (38%).  Out of the full

sample of 3,000-odd firms, only 12 failed to answer the question on their current use of

barter, a response rate of over 99.5%.  The response rate for the use of barter three years

previously was significantly lower, at 85%.

For just under one-third of the firms in the survey, barter and non-monetary

transactions make up more than 10% of their “sales” and for nearly one-fifth of firms, it

accounts for over 25%. Barter is more prevalent in Russia and Ukraine than elsewhere:

just over one half of firms report using barter for 10% of their business transactions and

just over one-third conduct 25% of their business this way (see Table 1). Other studies of

barter and non-monetary transactions in Russia and Ukraine are in line with the order of

magnitude reported in the EBRD survey.

Here we explore the data in several stages. To begin with, we look at size, sectoral

and locational effects. Next, using these as controls, we then look at the extent to which

the level of barter and non-monetary transacting reported by firms is related to

ownership, to financing problems and arrears and to competition in the product market.

After looking for firm-level correlates of barter, we examine whether some country level

variables are relevant: inflation, a measure of the softness of the budget constraint and of

the quality of the business environment. Finally we examine the consequences of barter

and non-monetary transactions for firm restructuring and performance.

2.2

Where does barter happen?

Table 1 shows the distribution of reported levels of barter and non-monetary

transactions by country. Barter is widespread in Russia and Ukraine. Elsewhere in the

CIS its incidence varies greatly, with high levels in Belarus, Moldova and Kazakhstan

and very low levels in some other countries. More surprisingly, barter and non-monetary

transactions appear in the Central and Eastern European countries (where they have often

been assumed to be absent). While there are relatively small proportions of firms

reporting barter at the level of 25% of sales or more, barter is non-negligible except

perhaps in Hungary. Croatia and Slovenia look quite out of line with the other non-CIS

countries in terms of the proportion of firms reporting no involvement in barter. This

suggests that the question may have been interpreted differently in those countries from

elsewhere. In the rest of the non-CIS countries (Central, Eastern and South-Eastern

Europe plus the Baltic states), the proportion of firms reporting no barter transactions

ranges from 49% in Estonia to 90% in Hungary.




5

Across all countries large firms are more likely to be engaged in barter than are small

ones. This suggests that there are economies of scale in barter and non-monetary

transactions (see Guriev & Ickes, 1999). However, as Table 2 shows, there are both

sectoral and country variations to this pattern. This table provides a method of comparing

the likelihood of a firm being involved in barter (to the extent of at least 25% of sales)

across countries, sectors and size class of firm. To illustrate the patterns in the data, we

show the predicted probability that barter and non-monetary transactions account for

more than 25% of sales for a small firm (with less than 50 workers) and for a large firm

(with more than 500 workers). For many countries there are not enough agricultural firms

in the sample to form the basis for predicted probabilities. In these cases, the results for

industry and services only are shown in the table.



Table 1. Percentage of firms in sample reporting each level of barter and non-monetary transactions,

by country

(EBRD enterprise survey, 1999)



Country

Percentage of sales accounted for by

barter and non-monetary transactions

No. firms

None

1-9

10-25

26-50

51-75

76-100

Russia


28.4

19.1


16.8

15.3


12.6

7.8


524

Ukraine


28.6

21.4


16.1

11.6


11.2

11.2


224

Other CIS

Armenia

82.3


10.5

4.8


0.8

1.6


0

124


Azerbaijan

78.0


8.7

11.8


0.8

0

0.8



127

Belarus


32.8

30.4


21.6

8.8


5.6

0.8


125

Georgia


72.1

10.8


12.4

3.9


0.8

0

129



Kazakhstan

36.2


21.3

18.1


7.9

11.0


5.5

127


Kyrghistan

47.2


11.2

18.4


11.2

8.0


4.0

125


Moldova

23.2


16.8

16.0


20.8

16.0


7.2

125


Uzbekistan

68.0


8.0

10.4


7.2

3.2


3.2

125


Non CIS

Bulgaria


64.8

24.8


7.2

1.6


1.6

0

125



Croatia

9.5


18.2

23.0


23.0

19.8


6.3

126


Czech Rep.

74.8


17.0

5.2


3.0

0

0



135

Estonia


49.2

42.2


6.1

2.3


0

0

132



Hungary

89.8


8.6

1.6


0

0

0



128

Lithuania

75.7

17.1


4.5

2.7


0

0

111



Poland

65.8


21.6

8.1


3.6

0.9


0

222


Romania

72.8


10.4

8.8


4.0

0.8


3.2

125


Slovakia

56.6


13.2

7.8


3.9

3.9


14.7

129


Slovenia

13.6


40.8

26.4


15.2

3.2


0.8

125


From Table 2 the size and sectoral distribution of barter looks quite similar for

Russia and Ukraine. Firms in industry are more likely to be engaged in barter than service

sector firms and in both cases, it is large firms that are more heavily involved. It is clear

that in Russia, barter is much more prevalent in agriculture than in the rest of the

economy. Small enterprises in Russian agriculture are just as likely to be involved in

barter as large ones.




6

Table 2 shows that there are wide differences between the other CIS countries in

the size and sectoral patterns of barter and non-monetary transactions, as well as in their

prevalence. In Kazakhstan and Moldova there appears to a lot of barter in agriculture –

but this is not true of Uzbekistan, where barter seems to be disproportionately found in

the services sector. There is also no uniform finding of a higher prevalence of barter in

large than in small firms. Amongst the more advanced reformers in Central and Eastern

Europe including the Baltics, large firms in industry are more involved in barter but there

do not appear to be size effects for services firms.

The patterns in the group of CIS countries look too disparate for the analysis of

the pooled results to be very meaningful. We therefore omit the other CIS countries from

the more detailed examination of the correlates of barter, and we concentrate henceforth

on Russia and Ukraine.  For similar reasons, we limit our analysis of the non-CIS

countries to the more advanced CEE reformers, excluding Croatia and Slovenia because

of doubts about data comparability (see above).



7

Table 2. Prevalence of barter and non-monetary transactions: firm size and sectoral effects

This table shows the predicted probability that barter accounts for more than 25% of sales of a small firm

(with 50 employees) or a large firm (with 500 employees). In countries in which too few agricultural

enterprises were included in the survey, results are shown for industry and services only. The predicted

probabilities are calculated from an ordered logit regression for each country in which the regressors are the

log of employment, sector dummies and interactive terms in size and sector.

Country

Size of


firm

Industry


Services

Agriculture

Percentage of firms reporting

>25% sales as barter and

non-monetary transactions

Russia


small

15.6


10.6

51.4


large

22.9


13.1

53.3


35.7

Ukraine


small

18.7


7.9

large


27.1

10.6


34.0

Other CIS

Armenia

small


2.9

1.7


large

3.8


0.6

2.4


Azerbaijan

small


0

0.1


large

0.2


0.3

1.6


Belarus

small


5.5

13.2


5.1

large


7.7

11.5


7.1

15.2


Georgia

small


8.5

0.9


large

7.4


1.2

4.7


Kazakhstan

small


10.5

4.2


59.5

large


16.3

7.9


57.1

14.5


Kyrghistan

small


28.2

4.1


12.8

large


27.2

4.6


17.9

23.2


Moldova

small


9.6

8.1


68.6

large


17.4

12.4


68.6

44.0


Uzbekistan

small


9.1

25.2


13.6

large


10.6

19.5


6.2

66.0


CEE+Baltics 

(excl. Croatia & Slovenia)

Czech Rep.

small

0.8


2.0

large


1.2

2.1


3.0

Estonia


small

3.0


2.0

large


2.9

1.9


2.3

Hungary


small

0

0



large

0

0



0

Lithuania

small

0.1


2.9

large


0.2

1.8


2.7

Poland


small

1.4


0.3

0.2


large

2.4


0.5

0.2


4.5

Slovakia


small

10.1


17.0

large


17.8

16.6


22.5

In addition to the size of firm and the sector, we also check for any association

between location and tendency to barter. The barter variable is regressed on two location

dummies, “big city” and “town”, (small city is the omitted category). The size, sector and

size-sector interaction terms are included as controls. For Russia, the location dummies

are highly significant – barter is much more prevalent in the more rural locations. For

example, in a firm in a town (the most rural location), barter as a share of sales is



8

estimated to be 11 percentage points higher than in a small city. In turn, barter in a big

city is estimated to be 6 percentage points lower than in a small city (see Table 3). This is

consistent with the idea that barter and non-monetary transacting in Russia may be in part

a product of limited trading networks, or “informational islands” (Seabright, 1999).

The sample size for Ukraine is substantially smaller than that for Russia (205 as

compared with 524). This will tend to pull down the significance levels of the coefficients

in the Ukraine regressions. Even keeping this in mind, the clear location effects

characteristic of Russia do not seem to be present in Ukraine. The signs on both big city

and town are positive, and the coefficients insignificant. Location also does not appear to

play a part in barter in the advanced reform countries.

Table 3. Location effects on barter and non-monetary transactions

The table reports the coefficients on the location dummies (the omitted category is “small city”) in an

interval regression with the percentage share of barter in sales as the dependent variable (see text).

Coefficients can be interpreted as the effect of location in percentage points on the share of barter in sales.

Size, sector and size-sector interaction variables are included in all regressions. The standard error is shown

in parentheses. Significance levels are indicated as follows: * indicates significance at 10%, ** at 5% and

*** at 1%

Location effects

Russia

Ukraine


CEE excl. Croatia &

Slovenia


big city

-6.39 (2.54) **

.21 (4.42)

-1.02 (1.23)

Town

10.88 (3.47) ***



4.68 (3.97)

.10 (1.11)

Number of firms

524


205

840


 2.3

The causes of barter and non-monetary transactions

The next step is to analyse in turn a series of possible correlates of barter and non-

monetary transactions. For example, do state firms do more or less barter than new

private firms; is barter more prevalent where the product market is less competitive; is

barter higher in firms reporting financing problems? In the regression analysis, we control

for size, sector - and in Russia, also for location - and allow for country fixed effects

within the Central European region. The omitted ownership category is privatised firms

that are not insider-owned. In Russia and Ukraine, new entrants make less use of barter

than do other firms. There is a clear tendency for state-owned firms in the CEE region to

do less barter – there is no sign of this in Russia and the effect in Ukraine although large

and positive is not significant. There is no indication that insider versus outsider

ownership of privatised firms makes any difference to involvement in barter.

It might have been expected that a foreign ownership stake would make

involvement in barter less likely by providing access to the parent company’s suppliers

elsewhere. However this effect is only found in Ukraine – the presence of a foreign

owner reduces the share of barter in sales by just under one-fifth. In neither Russia nor

Ukraine, nor in the CEE group, was there a correlation between engagement in exporting

and the presence of barter.




9

Table 4.  Ownership and exporting effects on barter and non-monetary transactions

The regression results for three regressions for each region are reported. The first part of the table reports

the coefficients on the ownership dummies (the omitted category is “privatised but not insider-owned”).

The second part reports the coefficients on the dummy variable for whether or not the firm has a foreign

owner. The third part reports the coefficients on a dummy variable for whether the firm exports or not.

Interval regression with the percentage share of barter in sales as the dependent variable is used in all cases

(see text).  Size, sector and size-sector interaction variables are included in all regressions. For Russia,

location dummies are also included. The standard error is shown in parentheses. Significance levels are

indicated as follows: * indicates significance at 10%, ** at 5% and *** at 1%

Ownership effects

Russia

Ukraine


CEE excl. Croatia &

Slovenia


(1)Ownership type

insider ownership

4.36 (3.16)

-5.29 (5.99)

-2.59 (2.54)

State


1.01 (5.29)

-10.70 (7.28)

-5.64 (1.58) ***

ab initio private firm

-7.28 (2.87) **

-10.37 (5.77) *

.03 (1.29)

(2) Foreign stake

.96 (8.11)

-18.86 (8.89) **

-.59 (1.58)

(3) Export

-.14 (.10)

.10 (.11)

.03 (.02)

Number of firms

524

205


840

There is a strong relationship between perceived financing problems and the role

of barter in the firm. This is clearly true in Russia and CEE, and true for some measures

though not all in Ukraine. Managers were asked to give a score to the seriousness of

financing problems in general, problems of access to long term bank credit and

difficulties caused by high interest rates. In each region, there is a very strong positive

correlation between the seriousness with which financing problems are rated by managers

and their involvement in barter. When asked specifically about problems with accessing

long term bank credit, managers’ ratings again showed a strong correlation with barter

except in Ukraine. High interest rates seem to capture a feature of financing problems

relevant to barter although the effect is not significant in Ukraine (see Table 5).



10

Table 5. Financing problems and barter and non-monetary transactions

Each row in the table reports the results from a separate regression for each region. Interval regression with

the percentage share of barter in sales as the dependent variable is used in each case (see text). The scaling

of the independent variable measuring financing problems runs from 1 to 4, with the exception of “frozen

bank accounts” and “tax offsets”, which are 1/0 dummies.  Size, sector and size-sector interaction variables

are included in all regressions. For Russia, location dummies are also included. The standard error is shown

in parentheses. Significance levels are indicated as follows: * indicates significance at 10%, ** at 5% and

*** at 1%

Financing problems

Russia


Ukraine

CEE excl. Croatia &

Slovenia

Financing problems in

general

4.66 (1.12) ***



5.14 (1.97) ***

.78 (.43) *

Access to long term

bank credit

1.91 (.93) **

-.51 (1.53)

1.94 (.41) ***

High interest rates

3.17 (1.10) ***

2.63 (1.93)

1.51 (.45) ***

Payments overdue to

suppliers

7.08 (1.03) ***

6.86 (1.64) ***

1.27 (.46) ***

Receivables overdue

from customers

5.17 (1.03) ***

5.79 (1.59) ***

2.16 (.45) ***

Tax arrears

8.00 (.99) ***

8.11 (1.67) ***

1.84 (.52) ***

Frozen bank accounts

12.90 (2.67) ***

13.18 (4.88) ***

N/A

Tax offsets



12.79 (2.51) ***

21.94 (4.54) ***

N/A

Given the findings for the correlation between barter and financing problems, it is



not surprising that there is a strong correlation in all three regions between managers’

reports of the extent of barter and both payments overdue to suppliers and overdue

receivables from customers (see Table 5).

The usefulness of barter and non-monetary transactions as devices to avoid

taxation has been much discussed in the literature. In the survey, firms were asked about

their overdue tax payments and there was a strong positive correlation between this

measure and barter in all regions.

When the EBRD survey was implemented in Russia and Ukraine two specific

questions were asked concerning tax arrears. Managers were asked to respond to the

following: (i) “Did your firm have your primary bank account blocked for non-payment

of taxes at any time in 1998?” and (ii) “The Federal, oblast and municipal governments

sometimes pay for their purchases from enterprises by reducing the tax liabilities of the

selling firm. During 1998, did your firm receive such a tax offset from any level of

government?” As is clear from Table 5, there is a very large and significant connection.

Barter and non-monetary transactions go together with the presence of frozen bank

accounts and of tax offsets arising out of non-payment of taxes.




11

Table 6. Product market competition and barter and non-monetary transactions

The results of two regressions for each region are shown. Interval regression with the percentage share of

barter in sales as the dependent variable is used in each case (see text). The omitted category in the first

regression is ‘no competitors’ and in the second regression, ‘many customers would switch to our

competitors’. Size, sector and size-sector interaction variables are included in all regressions. For Russia,

location dummies are also included. The standard error is shown in parentheses. Significance levels are

indicated as follows: * indicates significance at 10%, ** at 5% and *** at 1%

Product Market

Competition

Russia

Ukraine

CEE excl. Croatia &

Slovenia

(1) No. of competitors:

one to three

-10.12 (5.28)*

-.64 (7.55)

1.76 (2.07)

more than three

-9.85 (4.39)**

-2.67 (5.92)

4.98 (1.67)***

(2) Response to 10%

increase in own price

demand lower

.82 (2.91)

-4.97 (4.48)

-.65 (1.22)

demand slightly lower

 -.051 (2.91)

-2.35 (4.16)

-2.42 (1.15) **

no change in demand

 -2.32 (3.47)

-1.72 (6.71)

-1.98 (1.42)

The questionnaire used two approaches to eliciting information about market

power. Managers were asked whether the firm faces no competitors, one to three or more

than three competitors in the market for its main product. They were also asked to predict

what would happen to demand for their main product if they raised their price by 10%

(relative to inflation and to the prices of their competitors).

The correlation between each of these measures and the extent of barter and non-

monetary transactions is reported in Table 6. There is no uniform pattern across the three

regions when the relationship between competition and barter is examined. In Ukraine,

there seems to be no particular link between competition in the product market and barter.

We therefore concentrate on Russia on the one hand, and the CEE countries on the other.

In Russia, firms facing competitors in the product market were engaged in less barter than

were monopolists. The indicators of monopoly power from the 10% price test were not

significant.

But in the countries of Central and Eastern Europe, competition and barter are

related in the opposite way: firms with more than three competitors report more barter

and non-monetary transactions than do monopolists. There is some support for this kind

of effect from the second regression – i.e. using the 10% price test. Compared with the

omitted category in which customers switch to alternative suppliers if the firm puts its

price up by 10%, it seems that firms with market power do less barter.

The top panel of table 7 brings together the correlates of barter and non-monetary

transactions into one regression. One variable is used to reflect financing constraints

(arrears to suppliers) and the number of competitors is used to reflect competitive

conditions. The sample sizes are somewhat smaller here because we want to compare this

baseline regression with a regression that includes lagged barter. It seems that whilst

there are a number of common features of firms engaged in barter across the three regions

(size, financing problems and arrears, including tax arrears), there are also important




12

differences. In Russia, barter is a rural phenomenon but there is no locational aspect in

Ukraine or the CEE. In Ukraine, product market competition and barter are not related

whereas there are effects going in opposite directions in Russia as compared with the

CEE. Ownership effects are also quite different across region.

A fairly similar picture emerges when the change in barter over the past three

years is investigated. In all three regions, the presence of liquidity problems is strongly

correlated with the growth of barter. In Ukraine, there was a sharp increase in the use of

barter in outsider-owned privatised firms that is reflected in the highly significant and

large negative coefficients on the other ownership types.



Table 7.  Correlates of barter and of the change in the use of  barter

The results of two regressions for each region are shown. Interval regression with the percentage share of

barter in sales as the dependent variable is used in each case (see text). In (1), the right hand side variables

are a measure of financing problems (arrears to suppliers), product market competition (number of

competitors) and ownership dummies. In (2), the level of barter three years ago and a performance measure

(sales growth) are added. Size, sector and size-sector interaction variables are included in all regressions.

For Russia, location dummies are also included. The standard error is shown in parentheses. Significance

levels are indicated as follows: * indicates significance at 10%, ** at 5% and *** at 1%. ns mean not

significant at the 10% level.

Russia

Ukraine

CEE excl. Croatia &

Slovenia

(1) Benchmark

insider ownership

5.57 (3.27) *

-11.13 (6.46) *

ns

state-owned



ns

ns

-4.94 (1.67) ***



ab initio firm

-7.53 (3.15) **

ns

ns

payments overdue to



suppliers

7.66 (1.14) ***

7.06 (1.88) ***

1.39 (.49) ***

one to three competitors

-9.22 (5.48) *

ns

ns

more than three



competitors

 -11.73 (4.49) ***

ns

2.84 (1.78) (sign. at



11%)

Number of firms

404

174


741

(2) Change in barter

lagged barter

.64 (.04) ***

.77 (.06) ***

.75 (.02) ***

sales growth

Ns

ns



ns

insider ownership

4.18 (2.54) *

-14.99 (4.50) ***

ns

state


ns

-16.64 (5.18) ***

-2.65 (.81) ***

ab initio firm

ns

-10.21 (4.40) **



ns

payments overdue to

suppliers

4.83 (.90) ***

4.43 (1.35) ***

.62 (.23) ***

one to three competitors

ns

ns



ns

more than three

competitors

ns

ns



ns

Number of firms

404

174


741

Country-level correlates of barter and non-monetary transactions were also

investigated. For this exercise a pool comprising all twenty countries in the EBRD survey

was used. Only industrial firms were included. There was no correlation at all between

inflation (over the preceding three years) and the extent of barter. A country-level

measure of the softness of budget constraints constructed from managers’ responses to




13

the question on tax arrears was also insignificant. However, a country-level composite

measure of managers’ perceptions of the investment climate indicated that in countries

with a poor investment climate, there was more barter.



Table 8. Country-level correlates of barter and non-monetary transactions in industry (20 countries)

The results of three regressions for the pool of 20 countries are shown. In each case, the dependent variable

is the midpoint of the barter/sales interval; interpretation is the same as for interval regression. Estimation is

by a generalised least squares random effects model. The size variable is included in all regressions. The

standard error is shown in parentheses. Significance levels are indicated as follows: * indicates significance

at 10%, ** at 5% and *** at 1%

Inflation

 -5.30 (9.45)

Softness of budget constraint

10.41 (8.75)

Investment climate

 -19.70 (6.96) ***

Number of firms

1239


To sum up, there is substantial variation between countries in the correlates of

barter and non-monetary transactions. Nevertheless a few clear messages emerge from

the survey:

 



Financing problems are strongly linked to the presence of barter and non-monetary

transactions in all countries

 

Large firms are more likely to engage in barter than small firms



 

Difficulties with tax payments are strongly associated with barter



 

There is a strong degree of persistence of barter over time



 

There is more barter in countries with a poor investment climate



 

State-owned firms are less likely to engage in barter in Central and Eastern Europe,



but there is no systematic relationship with ownership in Russia and Ukraine

 



The phenomenon of barter in Russia is different from that in other countries in a

number of respects, and is much more linked to monopoly power and rural location

than it is elsewhere.

2.5

The consequences of barter

Is there a connection between barter, restructuring actions taken by managers and

performance at the level of the firm? The survey allows us to examine several dimensions

of restructuring: the introduction of new products, upgrading of existing ones and

changing the organisational structure of the firm. For each type of restructuring in turn,

we examine the probability that the firm has undertaken restructuring of that type. As

explanatory variables, we include the usual controls for size, industry and location. In

addition, we include ownership variables and the two kinds of competition variables

discussed above. To avoid problems of two-way causation, our barter variable is the

firm’s level of barter sales three years prior to the survey. This allows us to interpret the

results as indicating whether involvement in barter three years ago was a significant

predictor of subsequent restructuring actions.




14

In most countries there is no significant relationship between restructuring and

barter. But Russia is different. Table 9 illustrates. Here there is a strong tendency for

firms engaging in barter and non-monetary transactions to undertake more organisational

change than other firms. However, they are less likely to upgrade their existing products

than other firms (a firm 50% of whose sales are conducted in barter is about 50% less

likely to upgrade its existing products than one that does no barter at all). The

introduction of new products is unrelated to barter.



Table 9. The influence of barter and non-monetary transactions on restructuring in Russia

The table shows the results of logit regressions that control for size and industry effects as well as

ownership, urbanisation and the degree of competition faced by firms. The third equation is an ordered logit

where the dependent variable takes the values 1-4. Significance levels are indicated as follows: * indicates

significance at 10%, ** at 5% and *** at 1%

(1) New Product

Development

(2) Product

Upgrading

(3) Firm Re-organisation

(Scale 1-4)

% of barter in sales

0.002

-0.013**


0.011***

Standard error

0.005

-0.005


0.004

No. of observations

425

425


424

Without more information, the interpretation of these findings remains somewhat

speculative. But the results on organisational change are consistent with the evidence

from Russian microstudies (Ledeneva and Seabright, 1999) that barter deals frequently

involve significant diversion of managerial effort and the construction of ingenious

chains of transactions. The negative impact of barter on actions to upgrade existing

products in Russia fits the notion that barter and non-monetary transactions allow

otherwise unsaleable goods to be traded. The fact that there is no observed negative

impact on decisions to introduce new products may be due to the need for firms engaged

in barter to bring in new products purely to satisfy the needs of the barter chain (a

possibility documented also in the microstudies); this offsets what would otherwise be a

negative impact of barter on new product development.

If this interpretation of the results on organisational change is accurate, it implies

something about the kind of organisational change that is undertaken by firms heavily

involved in barter and non-monetary transactions - namely that it will be less effective

than similar reorganisations undertaken by firms who do not do barter. Table 10 tests this

suggestion by examining to what extent barter is associated with firm performance

(interpreted in terms of growth in sales over the three years prior to the survey). We have

included a direct effect of barter, a direct effect of reorganisation on performance, plus an

effect of barter for those firms that have reorganised. The results confirm the hypothesis

to a striking degree. Firm reorganisation is very strongly associated with improved

performance. In the basic equations, firms that have reorganised have 46 percentage

points higher sales growth over three years than those that have not. But this effect is

much weaker for firms that barter significantly, to the extent that a restructured firm

bartering 46% of its output performs no better than a non-bartering firm that has not

restructured at all. The results are even stronger when we control for the endogeneity of




15

the restructuring decision and for the fact that other kinds of restructuring take place

simultaneously. Here restructuring is associated with a 98 percentage point improvement

in performance over three years, an effect that disappears once a firm barters 44% or

more of its output.

Table 10. The influence of barter and non-monetary transactions on firm sales growth in Russia

The table shows the results of ordinary least squares regressions that control for size and industry effects as

well as ownership, urbanisation, the degree of competition faced by firms and the perceived pressures from

competitors, shareholders and creditors. The barter variable is the level of barter three years ago.

Significance levels are indicated as follows: * indicates significance at 10%, ** at 5% and *** at 1%

Basic equation

Using instrumental

variables and

controlling for other

restructuring forms

Major reallocation of

responsibilities among

departments

0.46***

0.98**


Standard error

0.13


0.38

Lagged % barter in sales

0.0004

0.003


Standard error

0.002


0.004

Lagged % barter

(firms with major

reallocation only)

-0.010***

-0.023**


Standard error

-0.004


-0.009

No. of observations

324

324


Overall, therefore, the conclusions of the survey are clear:

 



Outside Russia there is no clear link between the tendency to engage in barter

and non-monetary transactions and either firm restructuring activity or

subsequent sales performance

 



In Russia there is strong evidence that barter retards the improvement of

existing product lines, presumably by enabling firms to trade otherwise

unsaleable products

 



There is no clear link with the development of new products, probably

because what would otherwise be a negative impact on product development

is offset by a tendency to create new product lines to satisfy the demands of

partners in a barter chain

 

Barter is definitely associated with significant organisational change. But



unfortunately it also tends to make such change comparatively ineffective in

yielding performance improvements. This is consistent with the view of barter

as a significant diversion of managerial energy and initiative



16

3.

 

Conclusions

How much does the survey tell us about why barter is happening and whether it

matters? Not only the extent of barter but also its nature, its causes and effects, vary to an

important degree from country to country. What are the overall lessons to be learned?

 

Barter and non-monetary transactions are everywhere associated with



financing difficulties. This strongly supports the view that barter assures

liquidity in a trading environment in which credit is scarce and the

enforceability of loan contracts is uncertain.

 



The overall economic costs associated with using barter for this purpose vary

significantly from country to country. In Central and Eastern Europe the costs

appear to be low, perhaps because what firms are reporting as “barter and

other non-monetary transactions” refers principally to bills of exchange and

other debt instruments that do not distort firm behaviour to an important

degree (and may often be redeemed in cash).

 

In Russia, however, barter does distort firm behaviour, perhaps because firms



find it difficult or costly to redeem their bills of exchange and other debts in

cash. Firms that barter are less likely to devote their energies to improving

their products. They are more likely to engage in internal reorganisation

purely to keep their barter chains in being rather than to conquer new markets

and transform their future prospects.

 



In Russia but not elsewhere the findings are therefore consistent with the

hypothesis that economic disorganisation, in the sense of Blanchard & Kremer

(1997), means barter is both more likely to occur and more damaging when it

does occur.

Overall, the evidence suggests that barter and non-monetary transactions are often

highly inventive and resourceful responses of firms to difficult business conditions. It is

the conditions themselves rather than the responses that are the problem. In a better

business environment firms could direct the ingenuity and effort required by barter

transactions to more productive ends.



17

Bibliography

Abelshauser, A. (1975): Wirtschaft in Westdeutschland, 1945-1948 Deutsche Verlags-

Anstalt, Stuttgart.

Blanchard, O. & M. Kremer (1997): “Disorganization”, Quarterly Journal of Economics.

Carlin, W. (1989): “Economic reconstruction in Western Germany, 1945-55: the

displacement of ‘vegetative control’” in I. Turner (ed.) Reconstruction in Post-War



Germany: British Occupation Policy and the Western Zones 1945-1955. Berg

Commander, S. & C. Mummsen (1999): “Different Types of Non-Monetary Transaction

in Russia”, in Paul Seabright (ed): The Vanishing Rouble: Barter networks and non-

monetary transactions in former Soviet societies, Cambridge University Press,

forthcoming.

Guriev, S. & B. Ickes (1999): “Barter in Russia”, in Paul Seabright (ed): The Vanishing

Rouble: Barter networks and non-monetary transactions in former Soviet societies,

Cambridge University Press, forthcoming.

Hart, Keith (1987): “Barter”, in J. Eatwell, M. Milgate and P. Newman (eds.), The New

Palgrave: A Dictionary of Economics.  Macmillan Press, London and Basingstoke.

Ledeneva, A. & P. Seabright (1999): “Barter in post-Soviet societies: what does it look

like and why does it matter?”, in Paul Seabright (ed): The Vanishing Rouble: Barter

networks and non-monetary transactions in former Soviet societies, Cambridge

University Press, forthcoming.

Marin, D. and Schnitzer, M (1999), “Disorganization and Financial Collapse”, Centre for

Economic Policy Research (CEPR), Discussion Paper No. 2245, London

.

Mendershausen, H. (1949): “Prices, money and the distribution of goods”, American



Economic Review Vol. 39, No. 4. pp. 647-668.

Seabright, P. (1999): “Introduction”, in Paul Seabright (ed): The Vanishing Rouble:



Barter networks and non-monetary transactions in former Soviet societies, Cambridge

University Press, forthcoming.

Stamp, A.M. (1947): “Germany without incentive”, Lloyds Bank Review, Vol. 5.

StataCorp (1997): “cnreg – Censored-normal, tobit, and interval regression”, Stata



Reference Manual: Release Five, Stata Press, College Station, Texas.

Yüklə 152,43 Kb.

Dostları ilə paylaş:




Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur ©genderi.org 2024
rəhbərliyinə müraciət

    Ana səhifə