Number 12/March 1993

CAPITAL BUDGETING IN THE PUBLIC SECTOR:

LESSONS FROM ALBERTA’S HISTORY

Paul Boothe
Department of Economics

University of Alberta

Historical Lessons on Capital Budgeting

INTRODUCTION

Large and persistent government deficits at
both the federal and provincial level in Canada, have
sparked renewed interest in the treatment of capital
spending in the public sector. For example, in his
recent report1 Alberta’s Auditor General called for the
establishment of a capital budget in which capital
items would “not be charged to operations when they
are purchased. Rather, they should be expensed as
they are consumed.” (p.5) Proponents of capital
budgeting argue that it is useful to distinguish between
public expenditures whose benefits are mainly felt in
the current budgetary period and those which will
produce a stream of benefits long into the future. If the
benefits of some expenditures will accrue into the
future, so the argument goes, so should the costs.
Ideally the flow of benefits and costs would be
distributed over time so that in every period costs and
benefits to taxpayers would match.

Debate over the merits of capital budgeting in
the public sector is not new. Auld (1984)2 traces the
debate back at least to Adam Smith in 18th-century
England. Classical economists such as David Ricardo,
Jean-Baptiste Say and John Stuart Mill all contributed
to the literature in this area. Proponents have argued
along the lines suggested above, while opponents have
pointed to a number of perceived dangers including
macroeconomic effects and the loss of accountability.

The purpose of this note is to draw lessons from
a period in Alberta’s history when capital budgeting
played an important, perhaps even decisive role in the
province’s public finances. The period in question,
1905-1936, begins with the creation of Alberta as a
province and ends with the collapse of the province’s
finances and the default on Alberta government bonds.
Specifically, we will look at the effects of capital
budgeting in three main areas: political
accountability; future operating expenses and
accounting for depreciation of public capital. Our main
conclusion is that despite its attractiveness on a
conceptual level, the dangers inherent in capital
budgeting far outweigh the benefits. Indeed, it could
well be that capital budgeting actually exacerbates
the problem it is intended to solve — controlling the
size of the government deficit.

1Auditor General of Alberta (1993) Annual Report o f
the Auditor General 1991-92, Alberta Legislature,
Edmonton.
2D. Auld (1984) Should there be a Capital Budget for
the Public Sector? C.D. Howe Institute, Toronto.

Historical Lessons on Capital Budgeting

POLITICAL ACCOUNTABILITY

One of the most serious dangers in capital
budgeting in the public sector is the potential loss of
accountability. The reason accountability is lost is
that the existence of two separate accounts — one for
current expenditures and one for capital — obscures the
critical `bottom line’ for the public on which overall
affordability is judged. Evidence of the effects of
losing accountability from Alberta’s history is
presented in Figures 1 and 2. In Figure 1 we see the
Income Account deficit and the change in funded and
unfunded debt1. Budget Speeches and newspaper
articles from the time attest to the fact that the
Income Account deficit or surplus was the summary
measure most often used by politicians and journalists
when discussing the province’s finances. However, it is
the change in net funded and unfunded debt which
gives the consolidated measure of the province’s
financial performance for a given year. As is clear from
the Figure 1, by itself, the Income Account deficit gives
little warning of the province’s impending default in
1936.

The cumulative effect of reporting which
focused primarily on the Income Account is shown in
Figure 2. By the end of the period we see that Income
Account-related debt is only a small fraction of the
province’s total debt. However, the distinction
between `bad’ Income Account debt and `good’ capital
debt gave Albertans (both politicians and general
public) a false sense of security. This illusion was
destroyed only by fall of the United Farmers of
Alberta government and the default on provincial
bonds.

A second key problem arises in finding an
acceptable definition of `capital’. If capital spending
receives less scrutiny than current spending, politicians
will be tempted to expand the definition of capital to
include investments in `human capital’ such as current
expenditures for education or health care. Evidence on
this point is also to be found in Alberta’s history. In
the fiscal year 1930-31, the province was faced with
the burden of providing massive unemployment relief
because of the Depression. The definition of capital
spending was expanded to include relief, which
allowed $1.3 million to be charged to the Capital
Account rather than the Income Account. This
permitted the government to report an Income Account
deficit of $2.3 million (14.6 percent of Income Account
revenue) rather then $4 million (25.5 percent of Income
Account revenue). It is interesting to note that these

1Funded debt refers to debt which is issued as
government securities such as bonds on debentures.
Unfunded debt refers to debt without the same formal
security such as bank loans. The series in Figures 1 and
2 break in 1926 when the province moved from calendar
year to fiscal year accounting.

same issues are being discussed in the U.S. as the new
government searches for ways to come to grips with its
deficit.2

2See Karen Pennar (1993) “Beware of Accounting Magic
Tricks, Mr. Clinton.” Business Week, January 18, p.55.

Historical Lessons on Capital Budgeting

FUTURE OPERATING EXPENSES

Part of any capital-spending decision in the
private sector is a careful analysis of the impact on
future operating expenses. If capital and current
spending are to be segregated, it is especially
important to recognize the interaction between the two
kinds of spending. Simply put, capital spending today
contains an implicit commitment to operating expenses
in the future. Just as individuals include the cost of
taxes, utilities and ongoing maintenance in the decision
to purchase a home, so should the government take
account of the commitment to future operating
expenditures when they spend on capital today.

Historically, the province reported the book
value of all assets including physical assets such as
roads and hospitals. For example, by 1930-31, the
province had constructed roads, bridges and ferries
valued at $29.4 million. Operating expenses in that
year were $1.6 million or about 5.5 percent of their book
value1. Although most hospitals were privately
operated, provincially-owned hospitals in that year
were valued at $5.3 million and incurred operating
expenses of $0.871 million or about 16.4 percent of their
book value. Overall, provincially-owned capital
assets increased 16.1 percent in 1930-31 to reach a total
of $46.9 million. If we assumed an average ratio of 10
percent operating expenses to capital value, this
capital spending implied an additional future
commitment of $0.650 million (4.1 percent of Income
Account revenue).

Today, because of measurement difficulties, we
have no official estimate of the value of provincially-
owned capital. In addition, given that provincial
capital expenditures are mostly in the area of social
programs, ie. health care, post-secondary education
and recreation which are more labour-intensive than
roads and bridges, it is likely that the ratio of
operating expenses to capital costs is significantly
higher — probably at least in the 20-30 percent range.
If we assume for the sake of argument that the current
value of provincially-owned capital is of the order of
$20 billion, this implies that at least $4 billion of
current expenditure stems from past capital spending.
New capital spending of about $1.2 billion reported in
1990-91 committed the province to at least an
additional $240 million in future operating expenses.
In other words, almost 15 percent of the 1991-92 deficit
could be attributed to the capital spending of just the
previous year.

1Unfortunately, we have no way of knowing the
market value (after accounting for depreciation) of the
assets. However, the book value does permit us to
calculate a lower-bound estimate of the ratio of
operating expenses to current value.

Historical Lessons on Capital Budgeting

DEPRECIATION OF PUBLIC CAPITAL

In addition to accounting for operating
expenses, individuals and firms in the private sector
must also consider future depreciation as part of any
decision to purchase capital assets. Despite its long
history of capital budgeting, Alberta never
consistently reported depreciation and the public was
given a distorted view of the province’s finances by the
annual reports of the book value of capital assets.
Indeed, a proper accounting for depreciation might
well have given an early warning to citizens before the
province defaulted on its debt in 1936. For example,
assuming a 25-year life for capital assets and applying
`straight-line’ depreciation, an Income Account expense
of $1.6 million for depreciation should have been
reported in 1930-31. This would have increased the
Income Account deficit from $2.3 million to $3.9 million
or from 14.6 percent to 25 percent of Income Account
revenue.

Applying the same methodology to the current
situation is instructive. Again assuming a value of $20
billion for provincially-owned physical capital in
1990-91, an expense of $0.8 billion should have been
charged for depreciation. Thus, segregating new
capital expenditures of $1.2 billion would have had a
much smaller impact on the reported deficit in 1990-91

if proper account were taken of depreciation of capital
assets.

Properly accounting for the declining value of
capital assets is even more important in a jurisdiction
where provincially-owned capital is a major source of
revenue. Of course, in Alberta the principal assets in
question are oil and gas reserves. The implication is
that revenue from the depletion of oil and gas reserves
should not be treated as income without an offsetting
charge representing the decline in the value of those
assets. While measurement of the true value of the
decline in the asset is problematic, a first
approximation would simply be to charge the value of
net oil and gas royalties as an expense as well as a
revenue. The Province has made similar
representations to the federal government when it
comes to defining revenue for the purposes of
calculating Equalization payments. Revenue produced
by transforming a physical asset into a financial one —
in this case, the interest income from the AHSTF —
should continue to be counted as income. The net effect
of this accounting reform would be to increase the
provincial deficit by the amount of energy revenues —
about $2.7 billion in 1990-91.

SUMMARY

A number of practical issues arise when
considering whether capital budgeting should be
introduced by governments. The key issue is political
accountability, and it is clear from Alberta’s history
that capital budgeting can have the practical effect of
impairing the public’s ability to judge the
affordability of government spending plans. In
addition, it is difficult for politicians to resist the
temptation to define capital very broadly when
capital spending receives less scrutiny than its current
counterpart. It is clear that capital spending had an
important, perhaps determining role to play in the
collapse of the province’s finances in the 1930s and
Alberta’s debt default.

A second issue related to capital budgeting has
to do with the interaction between capital and current
spending — the ongoing operating expenses implied by
capital spending. It was shown that past capital
spending was an important contributor to default in the
1930s, and to mounting deficits and debt today.
Political accountability would be enhanced if the
public were fully aware of the future commitments
implied by enlarging the stock of provincial capital.

Proper accounting for depreciation was shown
to be a third important issue in capital budgeting.
Indeed, it was shown, based on reasonable assumptions,
that current depreciation charges probably largely

offset any reductions in the deficit that come from
segregating current capital spending. A related issue is

the treatment of Alberta’s most important capital
assets — oil and gas reserves. Properly accounting for
the depletion of those resources would have a large,
negative impact on the deficit.

Thus, while there are conceptual reasons for
distinguishing between current and capital
expenditure, in practice, capital budgeting has a
number of potential pitfalls which may hamper sound
policy making. If the goal of government is to make
the public better understand the difficult choices
which must be faced, capital budgeting may hinder
rather than help in difficult education process to come.