Why Toronto should transition to a new ‘cash allocation basis’ of accounting for budgets

For an example of the application of the ‘cash allocation basis’, see the companion article Example: The City of Toronto 2023 budget using the ‘cash allocation basis’ of budget accounting.

The new ‘cash allocation basis’ of budget accounting would allocate cash to three basic groupings: external revenue allocations, external expense allocations, and internal capital allocations (internal cash transfers). The current ‘modified cash basis’ allocates cash to only two basic groupings: revenue allocations and expenditure allocations. This causes conflations and distortions. Let me explain.

What’s the problem with ‘modified cash basis’, and what to do about it

There are four reasons to abandon the ‘modified cash basis’ of budgeting:

1. It’s confusing
2. It’s misleading
3. It’s inaccurate
4. It’s unnecessary

A couple of examples

1. Issues with Toronto’s stated budgets

Toronto’s published 2022 budget was ~$15.2B (from open data listing — formal presentation was $15B), with equal revenues and expenditures balancing to $0 (zero). I’ve seen senior reporters and civil society members report to the public and their constituents that this budget represents the day-to-day costs of running the City, when in fact it includes much more than that with unconventional inclusions. Following is a rough* listing of these inclusions.

For budgeted revenues if you take out

• contributions from reserves of $602M and contributions from capital of $207M (from open data: source; those aren’t revenues),
• reclassification between revenue and expense of $340M, and internal cost recoveries (double counted) of $312M (from auditor corrections: source, note 20),

then you instead have budgeted external revenues of $13.7B. This is a difference of $1.5B.

For budgeted expenses (“expenditures”) if you take out

• general contributions to reserves of $679M, Toronto Water, Waste and Parking contributions to reserves (surpluses) of $1B, and contributions to capital of $348M (from open data; those aren’t expenses),
• reclassification between revenues and expenses of $340M, debt principal repayments of $439M, and internal charges (double counted) of $312M (from auditor corrections),

then you instead have budgeted external expenses of ~$12.1B. This is a difference of $3.1B.

Those differences are 10% and 20% respectively from Toronto’s published revenue and expense budget figures.

So, first of all it’s impossible for operating budgets to net to $0, since they require surpluses to deal with capital needs (such as principal repayments). Second, the unconventional inclusions (and errors), corrected by the auditor, represent a substantial difference from Toronto’s published budget figures. It’s impossible to know on the face of it exactly what Toronto’s stated budget figures mean.

These contradictions and uncertainties undermine confidence in the budgets, and the ability of readers to make sense of them. The budgets, in other words, are confusing, misleading, and inaccurate. (See also the SOGR issue at the end of this article).

2. Misleading Operating Variance Reports

In its last 5 years, the Tory administration accumulated $8.5B in auidted surpluses, $7.3B of which was un-budgeted (in nominal $). In 2020, 21, and 22, Toronto received $3.9B in Covid support. Even without that there was a $4.6B 5yr surplus. But this was achieved mainly by massive underspending. See the clear trends in the following chart (in $M).

Figure 1. City of Toronto budget variances. Source: Henrik Bechmann. See spreadsheet

Here are the underlying audited surpluses during the same period ($M). (Note that accounting policies changed substantially in 2009).

Figure 2. City of Toronto audited operating surpluses. Source Henrik Bechmann. See spreadsheet.

These positive financial results (but negative operational results — underspending is bad) are confirmed in the 2022 audited financial statements. Toronto’s financial health is in fact quite good. For example 2/3 of Toronto’s financial assets are liquid (~$10B). There are many more positive financial indicators in the City’s audited statement presentation.

Figure 3. Source: Toronto staff presentation regarding 2022 audited financial statements

Meanwhile, the City’s Operating Variance Report for 2022 reported a net cash variance (shortfall!) of $395M for tax supported programs, and a surplus cash variance of $100M for rate supported programs. The City’s variance report is obviously completely different from the audited results ($1.8B surplus), as it does not take into account the massive underspending and surpluses reported by the auditor.

Moreover the detailed Operating Variance Report shows program “expenditure” underspending of $435M ($385M tax supported + $50M rate supported), again distorted by internal cash transfers, and nowhere near the $2.5B underspending reported by the auditor.

Given that the audited statements are the gold standard, then surely the Operating Variance Report should be seen as doubtful at best, or at the very least obscuring critical information.

The main underlying design problem

Accounting relies on grouping and summarizing of accounting transactions into clear, well understood categories.

The main conventional categories are

• revenues (money received from external sources),
• expenses (money paid to external parties),
• assets (owned items of value), and
• liabilities (debts).

Revenues and expenses change the value of the enterprise; changes to assets and liabilities (capital transactions) are ‘trades’ of one balance sheet item for another, and result in no net change in the enterprise value.

Toronto’s ‘modified cash basis’ deliberately breaks these conventions. (If you find this next part confusing, that’s because it is — don’t try this at home). The ‘modified cash basis’ uses unconventional classifications which

• include internal cash transfers from reserves (savings) with external ‘revenues’
• include internal cash transfers to reserves with external ‘expenses’ (‘expenses’ are renamed ‘expenditures’ to acknowledge the change in practice)
• include debt principal repayments (not just the interest) with ‘expenditures’
• include transfers from revenue intended for capital purchases with ‘expenditures’

The confusion comes from conflating these unconventional classifications with conventional categories. This breaks the reader’s mental model.

The solution in a nutshell

The solution is simple. Group the above unconventional classifications into a separate category (I use the term ‘capital allocations’), which protects the composition of the conventional groupings. Then the unconventional classifications are no longer unconventional, and become much easier to understand.

Where did this ‘modified cash basis’ come from?

History

The Province of Ontario authorizes municipalities to (optionally) use the ‘modified cash basis’ of accounting for budgets. (See here — they call the ‘modified cash basis’ the ‘modified accrual basis’ — same thing). I think the motivation came from the provincial directive for municipalities to not use debt for day-to-day operations. But municipalities can use reserves (savings) for operations, and they can save money to reserves for future operational use. Also they have to pay debt principal from operating surpluses (not from debt). This is the basis of the “balanced budget” requirement imposed on municipalities by Ontario.

So Ontario allowed municipalities to mangle all those things together under the title ‘modified cash basis’ (or ‘modified accrual basis’), with internal capital items conflated with external revenue and expense items. Maybe it seemed like a good idea at the time. It wasn’t.

It also wasn’t necessary. It’s easy to regroup the same line items included in the ‘modified cash basis’ so that they make more sense.

A simple design change to correct the ‘modified cash basis’

Study Figure 4 below. The first two columns are the reporting options mandated by Ontario. I’ve added the third column — ‘cash allocation basis’ — to demonstrate how easy it is to avoid the confusion of the second column.

Note that the cash allocation basis I’m proposing provides an important new piece of information — the cash operating surplus (revenue allocations less expense allocations, excluding amortization).

The ‘cash allocation basis’ has the same line items as the ‘modified cash basis’ data, but provides more information, and groupings that are much more natural. See how naturally the revenues and expenses sections read. Yet the zero cash balance is preserved at the bottom, and the important budgeting information in the capital allocations section is preserved. You could think of the new capital allocations section as internal cash transfers.

So to summarize, under the ‘cash allocation basis’ proposal the budget would have three basic sections, all involving cash:

• external revenues
• external expenses
• internal cash transfers (capital allocations of operating surplus)

This becomes in effect an operating cash flow model.

Figure 4. Improved Ontario Municipal Budget Accounting. Source: Henrik Bechmann

Note that amortization (an accrual, not a cash transaction) from the first column is left behind by the modified accrual basis of the second column. I have no objection to that, since a cash flow model is useful for budgeting.

What are the benefits of transitioning to the ‘cash allocation basis’ of accounting for budgets?

The ‘cash allocation basis’ for budget accounting is

• easy to understand, because it uses natural, well understood groupings. This supports greater engagement.
• insightful, because it separates internal and external transactions. This supports better decisions.
• accurate, because it doesn’t distort external revenues and expenses with internal transactions. This supports greater trust.
• easy to implement because it’s really just re-sorting and re-grouping existing budget line items. This supports feasibility.
• roughly comparable to accrual operating statements (such as audited statements), because mostly the same revenue and expense groupings are used (apart from amortization accruals). This broadens compatible resources available to stakeholders.
• a virtuous cycle, because as budgets become easier to understand, there will be pressure for more detailed, useful information. This supports constant improvement.
• supportive of analytics, because the standard groupings support standard comparisons and time lines. This supports precision in planning and learning.

Outstanding issues

I should note that although implementing the ‘cash allocation basis’ of budget accounting would make the variance reports more accurate, and bring the operating results of variance reports and audited statements closer together, there are a couple of outstanding issues.

State of Good Repair (SOGR)

The audited operating statements include State of Good Repair (SOGR) numbers (both budget and actual), because repair is an expense, not a capital item. However in Toronto budgets SOGR is included in the capital budget, not in the operating budget. This is a substantial amount, and the SOGR variance from budget to actual is historically large as well. So this discrepancy is an important barrier to achieving compatibility between the two sets of statements.

Revenue dedicated to capital

There’s another outstanding issue with the audited budget. Some revenues in the audited statements are dedicated to capital payments, and therefore presumably unavailable for operations, though the precise breakdown is unclear. This itself should result in a matching audited operating budget surplus, but apparently doesn’t. For example for the 2022 audit, government transfers recognized as revenue for capital purposes (not including SOGR?) was $648M (see note 22 of the 2022 audited statements). Development charges for capital projects in 2022 was $344M (from statement of operations and Note 9 — deferred revenue). Also water/wastewater “capital” collections for system improvements are posted to deferred revenue until earned (spent). Of the 2022 earned revenue from this fund (~$1B), $506M was listed as new capital construction in Schedule 1. So there could be a total of ~$1.5B 2022 revenue dedicated to capital purchases. Yet the statement of operations and accumulated surplus actually shows a budgeted deficit of $138M. Why there wouldn’t be a budgeted surplus for these capital revenues is a mystery, and leaves all aspects of the analysis of revenue dedicated to capital unclear.

However implementing the ‘cash allocation basis’ would greatly reduce the complexity of the differences between the budget and audited statements, making future resolution of outstanding issues more accessible.

*(there are some issues with the Toronto 2022 open data Summary by Expenditure Category budget listing, so the numbers offered may be approximate)

Henrik Bechmann is a retired software developer with an interest in all things related to the Toronto budget. From 2015 to 2018 he was the lead of the budgetpedia.ca project. Twitter: ‘@HenrikBechmann’