New Use of an Old Italian Invention - The double-entry bookkeeping used to monitor and secure financial stability of the new Swedish pay-as-you-go pension plan

International Workshop on
The Balance Sheet of Social Security Pensions

New Use of an Old Italian Invention
– The double-entry bookkeeping used to monitor and secure financial
stability of the new Swedish pay-as-you-go pension plan

by

Ole Settergren
Email: ole.settergren@rfv.sfa.se



Organised
by
PIE and COE/RES, Hitotsubashi University

Hitotsubashi Collaboration Center, Tokyo, Japan, 1st-2nd November 2004

Content
1 Introduction
18
2
Conventional measures of “actuarial balance” in pay-as-you-go pension plans
20
3 Pay-as-you-go
assets?
21
4
Swedish use of an Italian device
21
5
Does it matter?
25
Appendix: Income statement and balance sheet of the inkomstpension 27


1 Introduction

Public pension plans are probably the largest financial transaction systems that we have. In OECD countries
their expenditure ranges from 5 to 15 percent, and their liabilities, as a rough estimate, from 150 to 300
percent of gross domestic product (GDP). For many governments, pension payments are the single largest
expenditure. National pension plans also represent one of the most long-term commitments of governments.
The size of these systems is of course reflected in their importance to insured citizens. For many, perhaps
most, the claim of individuals on the public pension system represents their single largest “asset”.

In spite of the economic importance and long-term commitment of these pension plans, their financial
reporting is essentially medieval. The first problem is that reporting is scarce, and that when it exists it is not
infrequently of low quality. Here I will deal only with a second problem, which is that financial reporting on
pay-as-you-go pension plans is based essentially on single-entry bookkeeping, statements of cash flows and
projections of cash flows. The thirteenth-century invention of double-entry bookkeeping1 which for centuries
has been the preferred method for accumulating and presenting financial information in virtually all
organisations but governments, has not been applied to national pay-as-you-go pension plans.

The various cash-flow measurements that are used to show the financial status of public pay-as-you-go
pension schemes do not effectively answer the questions what cause, what effect, by what means, and at what
rate. Thus, present financial reporting is weak in essential information. The single most important measure to
enhance expert, public and policymaker knowledge of the workings of public pay-as-you-go pension plans
would be to introduce double-entry bookkeeping for these systems. A plausible assumption is that only what
is mentioned and measured exists in people’s minds. Thus, such improved reporting and knowledge may lead

1 Sombart….

18

---

to better policy measures in the future, when present financial imbalances will make it necessary to take
corrective action. However this second claim is much more doubtful than the first.

This paper is a brief introduction to the double-entry bookkeeping system that has been used for the
essentially pay-as-you-go financed Swedish public pension system since 2001.2 This bookkeeping system
was developed for the new pension plan as an intermediate step to ensure automatic financial stability. The
objective of the indexation of pensions and notional pension capital, including the so-called automatic
balance mechanism, in this unorthodox public pension scheme was to minimise the volatility of the value of
the average pension relative to the average income for people of working age, while adhering strictly to a
fixed contribution rate, or payroll tax. Admittedly, the double-entry bookkeeping procedure was an
unintentional spin-off from the research undertaken to achieve these policy aims.

2 The Swedish Context

The method used to prepare the balance sheet and income statement of the new Swedish public pension
scheme the inkomstpension is based on insights gained from the work undertaken to ensure its financial
stability. 3 The founders of this unorthodox social insurance scheme sought to legislate a pension plan that
would guarantee the financing of its obligations with a fixed contribution rate – 16% – and the resources
available in the buffer fund of the system. One reason for this financial firmness was based on social policy.
Financial stability is a matter of social policy since any imbalance will have to be paid by someone at some
time.

In addition, financial sustainability was considered necessary to provide credible protection for the scheme
against the daily governmental and parliamentary battle over resources. The founders were convinced that
citizens would benefit if the new Swedish pension plan clearly determined who would pay for any financial
imbalance, and when this burden would be borne. For this reason, they considered it necessary to insulate the
old-age pension system from the national budget to the maximum extent possible. This separation could only
be achieved by ensuring that the national budget would not be required to finance deficits in the pension
system.

These ambitions are hardly astounding – the surprising feature of the Swedish pension reform is that they
actually completed the long and arduous road to legislation.

Though possessing unusual personal qualities – at least for Swedish politicians – the seven members of the
pension reform group were not immune to the common temptation of wanting to have their cake and eat it,
too. Specifically, they sought to establish a pension system that was not just financially stable, but also
designed to ensure that pensions would develop in line with average earnings, i.e. provide a stable
replacement rate for different generations. Swedish economists, well versed in Samuelson’s (1958) basic text,
strongly urged reformers to index pensions and the pension liability to the growth of the contribution base,
mistakenly believing that such indexation would guarantee financial stability. As for the reformers, they were
concerned not only with financial stability, but also with the “content of the product”, i.e. the specific effects
on pensions; thus, they advocated indexation of pensions and the pension liability according to the
development of the average wage, rather than total wages.

The method of managing the conflicting goals of financial stability and a good insurance product – here,
essentially indexation to the average wage – is perhaps of some general interest to scholars or to other
countries. The dilemma in this case was “managed” not by a compromise, but through a design intended to
maximize the likelihood of indexation at a rate equal to the growth in average income, with automatic
exceptions to this indexation if the system would otherwise risk becoming financially insolvent. Weather the
system can afford to index notional pension accounts and pensions with the growth in average income or not
is determined by estimating the pay-as-you-go schemes assets and liabilities in a double entry bookkeeping

2 See The Swedish pension System Annual Report 2003, pages 30 – 41 available at http://www.rfv.se/english/pdf/arsred03e.pdf or
http://www.rfv.se/english/pdf/2500word.pdf for brief descriptions of this pension plan.
3 See the legislative history of the automatic balance mechanism.

19

---

system. Prior to the description of this bookkeeping a short discussion on conventional measures of actuarial
balance in pay-as-you-go pension plans.
3 Conventional measures of actuarial balance in pay-as-you-go pension plans

For this writer, the most familiar measures of the financial status of public pension schemes are the one that
was used in Sweden by Riksförsäkringsverket (RFV) prior to 20014 and the central measure presently used
by the US Social Security Agency (SSA).

In Sweden RFV was obliged to present an analysis every five years of the financial status of the public
pension scheme and, in relation to this analysis, to propose a suitable contribution rate, or payroll tax. The
analysis was presented mainly as a projection of buffer-fund development, in terms of fund ratio,5 assuming a
fixed contribution rate and unchanged benefit provisions. Normally financial balance, i.e. a buffer fund that
never dropped below a certain level in a specific scenario, would be secured by proposing an upward
adjustment of the contribution rate. The range of these projections varied, but prior to 1990 they were never
for longer than 50 years.

In several respects the financial analysis of the US Social Security system has been more advanced and
systematic than was the case in Sweden. One reason has perhaps been the longer tradition of the large US
public pension plan. The US Social Security system was introduced in the 1930’s; in Sweden the earnings-
related pension plan (ATP) was started in 1960. The US Social Security Administration (SSA) reports
annually on the financial status of the Social Security system. In this report, it uses a similar but slightly more
sophisticated, or dense measure of financial balance than the RFV previously used: a single figure called
actuarial balance. Briefly, the actuarial balance -- deficit or surplus -- reflects how much the contribution rate
must be increased (decreased), to ensure that the Social Security buffer fund, the trust fund, never drops
below a stipulated level in the standard 75-year projection of the SSA.

As noted, the main drawback of these measures, aside from all the very difficult issues related to the
preparation of these projections, is one of presenting the right information. To provide a clear picture of the
interaction that exists, we would like to know the financial position that preceded the one presented. We
would also like to know the reasons for the change in the preceding period and for each reason the magnitude
of its effect on the change. This requirement can also be met for the conventional measures used in
representing the financial position of pay-as-you-go pension systems, and it is indeed fulfilled occasionally
for some change factors. But such figures do not come from the calculation of the actuarial balance measure
itself, nor will such ad hoc analysis provide the self-controlling mechanism of double-entry bookkeeping.
Thus, not even the public pension scheme probably best analysed by conventional methods – the US Social
Security system – has provided regular information on such an interesting and elementary figure to show the
cost of the annual change in life expectancy.

The double-entry algorithm is the standard, and so far the most efficient, way of simultaneously and
consistently conveying financial position and changes in it. For this reason, double-entry bookkeeping ought
to be our prime candidate as a way to improve financial reporting on pay-as-you-pension systems.

There are some additional reasons why it probably is better to go the full distance of establishing an income
statement and balance sheet for national pay-as-you-go pension plans rather than trying to provide more
measures and analysis along the conventional path. One is that the words used in income statements and
balance sheets are more familiar to a wider circle of policy makers, journalists and informed members of the
public than are the terms used in standard pay-as-you-go actuarial disclosure. The actuarial balance figure
can also be criticised as unnecessarily abstract, since it does not explicitly mention liabilities, or assets to
match them. However, the strength of this argument is questionable; with pay-as-you-go financing, the assets
that should match liabilities are inherently abstract.


4 In english The National Social Insurance Board.
5 The market value of the fund divided by one year of pension disbursements. In Sweden this measure is usually referred to as
“fund strength”.

20

---

4 Pay-as-you-go assets?

The idea that the financial position of a pay-as-you-go pension plan can be presented in the terms of assets
and liabilities does not come naturally, and indeed it may need some getting used to. However, to most
people it seems clear that a pay-as-you-go system has liabilities, both to retired persons and to those who at
the time of measurement are working and have accrued some pension claim. Opinions differ on the choice of
method to estimate the value of this liability, but not its existence.6

It is understandably more controversial to claim that the liabilities of a pay-as-you-go pension plan are fully
or partially backed by something that we may call “assets”. A defining feature of pay-as-you-go financing is
that it finances its payments not from pre-funded assets, but from current contributions. As a pay-as-you-go
system has little or no tangible assets, it may be considered, by definition, to be in permanent deficit. While
this view seems plausible, it may be impractical and perhaps even questionable on theoretical grounds.

For example, in the extremely unlikely event that a pay-as-you-go system with fixed contribution rate and
benefit rules, contributions continuously and perpetually perfectly match pension payments, does this system
have a deficit, or is it in financial balance? Reasonably, it can be considered to be in financial balance, i.e.
have a net present value of zero. If a system with a liability of a defined and measurable size has a net present
value of zero, it must also have assets equal to that liability. As opposed to a premium reserve plan, a pay-as-
you-go pension plan is free to use contributions to pay off the pension liability, even when the contribution
directly or indirectly is a source of a new pension liability. Thus, in a pay-as-you-go pension system, the
contribution flow can and should be considered as the principal asset. The double-entry bookkeeping of the
new Swedish pay-as-you-go pension plan is based on this reasoning. The Swedish concept of the value of the
contribution flow is expressed in English as the contribution asset.7

It is surprisingly simple to calculate the value of the contribution flow: it is the product of the size of the flow
per time unit, which in practice is a year, and the expected time between payment of contributions and receipt
of pensions. The averages are weighted by the age-dependent amounts of expected contributions and
pensions. In Sweden the expected contribution-weighted average age of contributors is about 42, and the
expected pension-weighted age of retirees is about 74. Thus, the relevant time span is about 32 (74-42) years,
and the contribution asset is 32 times one year’s contributions. In the Swedish legislation on the automatic
balance mechanism, this time span is called omsättningstid – expected turnover duration -- as it is a measure
of the time required for a cycle of accumulation and depletion of the pension liability. Information on the
expected turnover duration can be, and is in Sweden, annually retrieved from the records of the pension plan.

The contribution asset, that is, contributions times turnover duration, tells us the size of the pension liability
that would result in a steady state determined by demography, i.e. nativity, age-related net migration and
mortality, and the economy of the country, i.e. the size of the contribution base and the age-related average
incomes, at the time of measurement.8 The turnover duration measures informs, in a single figure, the effect
that changes in fertility9 and age-related income patterns and mortality have on the capacity of the
contribution flow to finance pension liability.

5 Swedish use of an Italian device

The income statement and balance sheet of the inkomstpension plan for the years 2001, 2002 and 2003 are
“reproduced” below. However, to facilitate international comparison, the amounts here are expressed in
percent of GDP for each year, the appendix gives the original amounts in Swedish currency. The income
statement consists of 13 genuine entries, of which only the four cash-flow statements were disclosed prior to
2001. The balance sheet consists of five genuine entries, of which only one – the market value of the buffer

6 The preferred method depends partly on the objective of the liability estimation.
7 R. Lee, an economic demographer has published considerable work relating to the value of the contribution flow in pay-as-you-
go financing; Lee uses the term transfer wealth for what in Sweden has been termed the contribution asset .
8 For details on turnover duration and the contribution asset, see the legislative history of the automatic balance mechanism,
Settergren (2001), Settergren and Mikula 2002 and Settergren and Mikula 2003. R. Lee deals extensively with the concept.
9 In the legislation on the Swedish scheme, the effects of fertility changes on turnover duration are disregarded.

21

---

fund -- was disclosed prior to 2001. For each of the entries, there is a note with detailed information; here the
notes will not be considered.10

The income statement is divided into three sections. Section (a) Change in funded assets deals with the cash
flows of the scheme, i.e. those flows that have changed the value of the buffer fund. This section contains no
new information relative to what always has been reported. Noticeable is the – for a European context – low
level of contributions and pension disbursements, around 7 percent of GDP. This is explained largely by the
fact that the inkomstpension deals exclusively with earnings related old-age pensions. Thus disability
pensions, the guarantee-pension and survivors benefits, which is slowly being phased-out of Swedish social
security, are not included in the income statement nor in the balance sheet. Would those benefits be included
expenditure would be around 10 percent of GDP. Presently contributions are larger than pension
disbursements, a situation that is projected to continue up until around year 2010. From that time on pension
disbursements are projected to surpass contributions, the deficit will be financed with the return on the assets
in the buffer fund and also, according to projections, part of its capital. The buffer fund is valued at market
prices the last trading day of the accounting period. As the buffer fund to 60 percent is invested in equities
the sharp decline in prices on that market has fed into significant buffer fund losses in 2001 and 2002, losses
that were partly regained by a positive development in 2003.

Section (b) Change in contribution assets deals with how much the contribution asset has changed due to
change in contribution flow (revenue) and due to change in turnover duration respectively. As the
contribution asset is calculated as the contribution flow (C) times the turnover duration (TD) the separation of
the effects from changes in two components imply that the Value from change in contribution revenue is:11
(
TD +
C ? C
×
TD
t
t
. The Value from change in turn over duration is similarly calculated as
t
t ? )
1
?
1
2
(
C +
TD ? TD
×
C
t
t
. There is a positive trend in turnover duration, due to the persistent increase in life
t
t ? )
1
?
1
2
expectancy, this tendency has only been partly offset by the tendency to delay entry into the labour force, due
to prolonged time spent in education.12 In a steady state the Value of change in turnover duration is zero and
Total change in contribution asset assets will grow only as a function of the growth in the contribution flow.
The contribution flow grows with increases in average incomes times the number contributors, essentially the
number of gainfully employed.13

Box 1.
Income statement of the Inkomstpension as a percent of GDP14


2003
2002
2001

GDP, millions of SEK (1 Euro ? 9 SEK)
2,440,058 2,347,400 2,266,387




Change in funded assets (a)

Pension
contributions
6.8
6.8
6.9
Pension
disbursements
-6.4
-6.5
-6.3

Return on funded capital
3.4
3.6
-1.1

Costs of administration
-0.1
-0.1
-0.1

Total change in funded capital (a)
3.7
-3.3
-0.6




Change in contribution assets (b)

New
Value of change in contribution revenue
6.6
9.6
17.9
New
Value of change in turnover duration
0.5
-0.7
0.7

Total change in contribution asset (b)
7.1
8.8
18.6

10 The full reports 2001 and 2002 and 2003 report can be downloaded free of charge from www.rfv.se/english/publi/index.htm
11 Trough out this paper the different kinds of smoothing that is done according to the legislation is disregarded, for details see the
technical appendix of The Swedish Pension System Annual Report 2002.
12 See the special feature article in The Swedish Pension System Annual Report 2002.
13 In the Swedish scheme the government finances with general tax revenue the contributions for unemployed persons, sick persons
etc., thus the contribution flow depends on more factors than the number of gainfully employed.
14 Source for the numerators are The Swedish Pension System Annual Report 2001, 2002 and 2003. The GDP denominator used are
from Konjunkturinstitutets konjunkturrapport, in March 2004.

22

---

Change in pension liability* (c)

*A negative value
New
New Pension credits and ATP points
-7.1
-7.1
-6.1
(-) means that the
pension liability
(New)
Pension disbursements
6.4
6.5
6.3
increases, and a
New
Indexation -9.4
-11.8
-5.1
positive value ( )
that the pension
New
Value of change in life-expectancy
-0.5
-0.3
-0.8
liability decreases,
New
Inheritance gains arising
0.3
0.3
0.2
by the amount
shown.
New
Inheritance gains distributed
-0.3
-0.3
0.0
New
Deduction for costs of administration
0.1
0.1
0.0

Total change in pension liability (c)
-10.5
12.6
-5.7
(New)
Net income/ -loss (a)+(b)+(c)
0.3
-7.1
12.3

Section (c) Change in pension liability informs of the reasons and magnitudes of changes in the size of the
pension liability. The pension liability:
? Increases as new pension credits has been earned during the accounting period. When the new pension
system has been fully phased in 2018, i.e. when no ATP points any longer can be earned, New pension
credits will equal Pension contributions. That pension credits earned are equal to contributions paid is
one criterion for a pension scheme to be defined-contribution.
? Decreases as part of the liability has been paid off as pension disbursements have been made.
? Increases with the interest paid on the liability, i.e. the indexation. When the new system has been fully
phased in this interest will by default be equal to the change in average income in Sweden. Only if the
automatic balance mechanism is activated the indexation will deviate from the change in average income.
? Increases (decreases) with the Value of changes in life-expectancy. Even in the Swedish pension plan,
which to my knowledge is the world’s best financially insulated pension scheme relative to changes in
life expectancy15 the economic impact from such demographic changes are important. The annual
increase in pension liability due to changes in life-expectancy varies from 0.3 in 2002 to 0.8 percent of
GDP in 2001.
? Decreases with the Inheritance gains arising that is the value of the pension claims of persons that have
died before beginning to draw a pension,
? Increases with Inheritance gains distributed that is the survivor bonus distributed to those non retired and
finally the pension liability,
? Decreases with the reduction for administrative costs made of the insured’s claim on the system.

Box 2.
Balance sheet of the Inkomstpension as a percent of GDP


Dec. 31
Dec. 31
Dec. 31
2003
2002
2001

Assets


National Pension Funds
23.6
20.8
24.9
New
Contribution asset
224.0
225.5
224.4
Total
assets
247.6
246.2
249.3




Liabilities and surplus

New
Opening surplus/-deficit
2.1
9.3
-2.7

Net income / -loss for the year
0.3
-7.1
12.3
New
Total (closing) surplus /-deficit
2.4
2.2
9.6
New
Pension liability
245.2
244.0
239.7

Total liabilities and surplus
247.6
246.2
249.3

The assets of the balance sheet are the market value of the buffer fund and the contribution asset, calculated
as C × TD . Even though the Swedish national pension plan has one of the largest buffer funds of any
t
t
national pay-as-you-go pension plan it represents less than 10 percent of the pension liability.

15 Viewed from another angle this efficient insulation can be view as efficiency in passing on the economic effects from increases
in life expectancy to the insured.

23

---

The opening surplus is, naturally, the closing surplus of the preceding accounting period, as the GDP
denominator changes between the years this important relationship is obscured. The opening surplus/-deficit
plus the net income/-loss for the year gives the closing surplus for the accounting period. The fundamental
accounting equality states that:
Opening balance + net income/loss + (Pension) liability = Total assets.

If one single financial indicator of financial position is sought the natural, most information dense, is total
assets divided by pension liability. In the translation of the Swedish accounting this figure is called balance
ratio. If this ratio is less than one (1) the system has a deficit, a negative net present value, if it is above one
the system has a surplus, a margin. In the Swedish system legislation stipulates that if the balance ratio is
below one the balance mechanism is triggered.

Box 3.
The balance ratio – Summarising the accounting in one figure


Dec. 31
Dec. 31
Dec. 31
2003
2002
2001
Total
assets
247.6
246.2
249.3
./. Pension
liability
245.2
244.0
239.7
= Balance
ratio
1.0097
1.0090
1.0400

When the balance mechanism is triggered the indexation of the pensions and pension credits will be
effectuated by the change in average income adjusted by multiplying the index with the balance ratio, starting
of a new index series. The balancing of the system means that an amount equal to the closing deficit in the
balance sheet will be eliminated. If the balance ratio after being triggered grows bigger than one the closing
balance surplus is automatically distributed by increasing the indexation. This continues until the balance
index reaches the level of the income index.

In some respects the balance ratio resembles the actuarial balance measure that SSA uses, but the income
statement and balance sheet that comes with the balance ratio, and the notes tied to their entries, incorporates
details on what cause provokes what effect, by what means, at what rate. Thus it is better at supplying the
essentials of knowledge.

It is clear that if other administrators of national pension plans were to present their financial position the
individual entries of their income statements and balance sheets, and the associated notes, need to be
modified. This to comply with the specific demands of each plan, but the general outline should be possible
to follow. However the work required in each case to be able to present a complete income statement and
balance sheet should not be underestimated, neither should the rewards from a successful such effort.

A less ambitious endeavour, but a still worth while financial indicator, would be to start to annually publish
the change, most likely increase, of the pension liability caused by changes (increases) in life expectancy.16

A note on the particular simplicity of the new Swedish pension plan

The double entry bookkeeping of the new Swedish essentially pay-as-you-go financed pension plan has been
much facilitated by the design of the inkomstpension. Very briefly these design features implies that the
present value of pension liabilities in the pension scheme can be approximated to equal the nominal value of
the pension liability. Further the nominal value of the pension liability to persons that have not yet started to
draw a pension is much simpler to calculate in a (notional) defined-contribution pension plan, such as the
Swedish is, than it is for a traditional defined-benefit plan. The inkomstpension liability to persons who have
not yet begun to persons still economically active is valued as the aggregate of the amounts on each
individuals so-called “notional” account. This calculation entails a simple aggregation of account balances in
RFV registers. The pension liability to retirees is also presented at its nominal value. This is done by
multiplying pensions granted by the expected number of times that the amount will be disbursed. The number

16 If this is done the positive effect on the expected turnover duration from increases in life expectancy should be deducted.

24

---

of expected disbursements is calculated from annual measurements of the length of the time that pension
amounts in RFV records are paid out.

The nominal valuation of assets and liabilities of the Swedish pension plan imply that all valuations are made
solely according to what is observable at the time of valuation. For example, the normal assumption that
contribution revenue increases at the rate of economic growth is not explicitly considered in the calculation
of the contribution asset. Nor is the assumption that pension disbursements, because of factors like indexation,
will increase in the future considered in the valuation of the pension liability. The main reason why it has
been deemed reasonable to value assets and liabilities solely according to what can be observed is that the
financial position of the system is not dependent on the amount of assets and liabilities taken separately. The
financial position of the system is determined exclusively by the relationship between assets and liabilities, in
other words, by the so-called balance ratio.

The inkomstpension is designed so that there is a strong link between the development of the assets and
liabilities of the system. In cases where the balance ration exceeds on (1), however, liabilities and assets will
develop at somewhat different rates. In cases where the balance ration is less than one, the provisions for
automatic balancing establish an absolute link between the rates of growth in liabilities and assets. Taken as a
whole, this means that valuing the assets and liabilities of the system solely on the basis of conditions
observable at the time of valuation entails no risk of overestimating assets in relation to liabilities in the long
run.17 Together with other design features the provisions for automatic balancing have eliminated the need
for making explicit assumptions about future economic and demographic developments in order to ensure the
financial stability of the system.

It is apparent from the above that the method for valuing the assets and liabilities of the inkomstpension
system is implicitly based on the assumption that assets and liabilities grow at the same rate after each
valuation. To put it another way, it is assumed in the method of valuation that the indexation of the system
will always be the same as the internal rate of return of the pension liability, even though this outcome is
certain only if balancing has been activated. When balancing has not been activated, the indexation can be
either greater of less than the internal rate of return of the pension liability.

6 Does it matter?

Will it make any difference if governments continue to furnish us with projected cash flows, perhaps boiled
down to a figure reflecting “actuarial balance” as done in the US, or if they would start to present income
statements and balance sheets and summarise them in a “balance ratio”? I believe it would make a difference.
A “normalised”, double entry, way of reporting on financial position of national pension plans provides more
detailed information – presented in a more comprehensible way – to both policy makers and the insured than
do discounted cash flows. And the balance ratio is, in my opinion, a better summary of the financial situation
in a pay-a-s-you-go pension plan. Such improved information has the potential to promote better policies
when it comes to stabilise the finances of national pension plans.


17 The manner of calculating turnover duration involves an implicit assumption that the population growth is zero. Thus, turnover
duration will be (slightly) over estimated in cases where the working-age population is decreasing. This entails a risk that the
calculations will (slightly) overestimate the system’s assets in relation to its liabilities. However, it is reasonable to assume that the
population decline will cease at some point. If so, the deficit will be temporary.

25

---

Figure 1
Four stylised policy makers – which one are you?18

Trusted
Trusted
1
false
true
prophet
prophet
The publics
perception of
ectations
xp
the policy
makers
knowledge of
f satisfied e
the future
Mistrusted
Line o
true
0
Accountant
prophet
0
1
Policy makers
knowledge of the future

Applying double entry bookkeeping might also promote a debate on to what extent the financial stability of
national pay-as-you-go pension schemes ought to rely on future development of the demographic and
economic factors that in a defined benefit plan decide the plans financial development. And to what extent
their financial stability should be less dependent of such demographic and economic factors by transferring
risks and opportunities to the pension benefit-level, as is done in a defined-contribution plan. The
inkomstpension is radical in that it transfers all demographic and economic risks to the insured. This has
made the accounting simpler than it will be in systems that instead, implicitly or explicitly places the risks on
the contribution rate, or splits the risk between contribution rate and pension level.19 In defined benefit
pension plans the accounting will need explicit assumptions of the future. Doing this contains risks and
opportunities – illustrated in figure 1 – that the accounting provisions for the Swedish pension keep away
from.

18 The essentials of this image came out of a discussion with Boguslaw D. Mikula at RFV.
19 The risks that the insured are exposed to is reduced by the existence of the so-called guaranteed pension. This benefit is financed
by general tax revenue, not contributions to the inkomstpension system. In a sense, the guarantee pension can be considered to have
the effect of splitting the risk between contributors/tax payers and retirees, however retirees with (relatively) low pension benefits
are beneficiaries of this tax financed insurance.

26

---

Appendix: Income statement and balance sheet of the inkomstpension

Income statement, millions of SEK20
2003
2002
2001
Change in funded assets (a)

Pension contributions
165,107
160,745
156,811
Pension disbursements
-155,410
-151,757
-143,564
Return on funded capital
82,060
-84,529
-25,035
Costs of administration
-2,359
-2,081
-1,943
Total change in funded capital (a)
89,398
-77,622
-13,731


Change in contribution assets (b)

Value of change in contribution revenue
159,964
224,275
405,877
Value of change in turnover duration
12,346
-16,763
15,745
Total change in contribution asset (b)
172,310
207,512
421,622


Change in pension liability (c)

New Pension credits and ATP points
-172,567
-167,585
-138,627
Pension disbursements
155,410
151,562
143,564
Indexation -228,288
-275,946
-116,287
Value of change in life-expectancy
-11,045
-5,923
-18,728
Inheritance gains arising
7,090
6,389
5,476
Inheritance gains distributed
-7,616
-6,617
-5,490
Deduction for costs of administration
1,475
1,478
923
Total change in pension liability (c)
-255,541
-296,642
-129,169


Net income/ -loss (a)+(b)+(c) 6,167
-166,752
278,722

Balance sheet, millions of SEK

Dec. 31 2003 Dec. 31 2002 Dec. 31 2001
Assets


National Pension Funds
576,937
487,539
565,171
Contribution asset
5,465,074
5,292,764
5,085,252
Total assets
5,780,303
5,650,423


Liabilities and surplus

Opening surplus/-deficit
51,645
218,397
-60,315
Net income / -loss for the year
6,167
-166,752
278,722
Total (closing) surplus /-deficit
57,812
51,645
218,407
Pension liability
5,984,199
5,728,658
5,432,016
Total liabilities and surplus
6,042,011
5,780,303
5,650,423


20 Source The Swedish Pension System Annual Report 2001, 2002 and 2003.

27

---