Fiscal cost of task-based IT contracting

Summary

Based on a request from the Standing Committee on Government Operations and Estimates, this report estimates the fiscal cost of task-based IT contracting.

The Parliamentary Budget Officer (PBO) learned that much of the contract data required to complete our analysis is not held in a centralized database within federal departments and agencies. Departments and agencies had to manually retrieve the requested information, which resulted in significant delays and a substantial decrease to the scope of our analysis.

Consequently, this report examines the fiscal cost of task-based IT contracting in the fiscal year 2022-23 only. PBO sought information from the top five organizations with the highest spending on informatics services in that year. However, only four organizations were able to provide sufficient data to conduct a high-level analysis.

PBO found that the cost of external contractors was between 22.0 per cent and 25.7 per cent higher than their public sector equivalents for the four departments analyzed. With the data that was provided, PBO is unable to determine the underlying reason for these premiums and whether these premiums were justified.

For a more in-depth analysis of the use of task-based IT contracting, parliamentarians may wish to ask the Office of the Auditor General to conduct an audit, as the PBO is limited to the data it receives through information requests submitted to departments and agencies.

Context

On March 18, the Standing Committee on Government Operations and Estimates adopted a motion requesting that the Parliamentary Budget Officer (PBO) complete a full report on the increased costs taxpayers pay due to the government using 635 staff augmentation companies.[^1]

Upon consultation with government departments and agencies, it was determined that the 635 companies referred to in the motion are those who qualify under the task-based informatics professional services (TBIPS) supply arrangements. As such, this analysis attempts to estimate the fiscal cost of the use of task-based IT contracting by the federal government.

Supply arrangements are a method of supply wherein departments can solicit bids from and award contracts to suppliers within a pre-qualified pool for specific requirements.

Task-based informatics professional services are used when a particular activity is required to address a specific information technology need. The tasks are finite assignments that are usually associated with a specified set of responsibilities.

Public Services and Procurement Canada.

Public Services and Procurement Canada.

PBO first sent an information request to the Minister of Public Services and Procurement seeking the full list of companies that qualify under the TBIPS supply arrangement. Public Services and Procurement Canada (PSPC) provided the full list, which includes companies for which supply arrangements are both active and inactive.

We then sent an information request to the President of the Treasury Board, seeking detailed information for all companies in the list from which the Government of Canada used resources on a per diem basis in 2022-23 and 2023-24. The President of the Treasury Board indicated that they are not in possession of the requested information beyond contracts issued within their departmental mandate and that PBO would need to request the information directly from each individual department and agency.

Through conversations with government departments and agencies, it was communicated that such data requests could represent a significant use of time and resources to fulfill. PBO was informed that the data required to complete its analysis, including the category of contractor, per diem rates, total days worked and number of contractors, is generally not held in a central financial system within departments. The generation of this data would be a manual exercise, whereby departments and agencies would review individual contracts to extract the requested information.

Unlike the Office of the Auditor General, whose staff can directly access information from departments and agencies to complete an audit, PBO can only receive information from federal departments and agencies through information requests sent to the minister.[^2] As such, our office relies on data compiled by and sent to us by departments and agencies. To provide timely analysis to the committee, PBO revised the scope of its analysis.

Scope of PBO analysis

PBO relied on data from the Public Accounts to narrow the scope of its analysis. At the time of analysis, the 2023-24 Public Accounts were not yet published. As such, PBO reduced the timeframe of its analysis from two fiscal years (2022-23 and 2023-24) to 2022-23 only.

In 2022-23, spending on professional and special services totalled $18.6 billion across fourteen categories (Table 1).[^3] PBO focused its analysis on the informatics services category of professional and special services. Informatics services are comprised of information technology and telecommunications consultants, information management services, information technology services and computer services.[^4]

PBO identified the top five organizations with the highest spending on informatics services in 2022-23. As shown in Table 2, this includes the Canada Border Services Agency (CBSA), the Department of National Defence (DND), Employment and Social Development Canada (ESDC), Public Services and Procurement Canada (PSPC), and Shared Services Canada (SSC). These five organizations represent over half of total spending on informatics services in 2022-23.

PBO sent information requests to each of the five organizations seeking information on relevant contracts.[^5] The content of these information requests was developed in consultation with PSPC. PBO also sought PSPC’s guidance on the time required for organizations to fulfill these requests, extending the typical deadline by over one month to allow for sufficient time for data collection.

The information requests required the organizations to identify the contracts with the relevant companies that were used for task-based informatics professional services, where the contractors were paid on a per diem basis, and where the contractors worked more than 200 days in fiscal year 2022-23.[^6][^7] For the identified contracts, PBO then requested the following details on the actual contractors:

• Category of consultant;
• Per diem rate per consultant;
• Total days worked per consultant;
• Number of consultants; and,
• Total contract value.

In subsequent correspondence with the departments and agencies, several confirmed that this information is not kept in a centralized database, requiring departments and agencies to manually search their contract documentation to get the requested data. The lack of easy access to this information prompted several organizations to seek extensions to the information requests beyond the already extended timelines. Additional follow-up was required in some instances to obtain a complete response that would facilitate PBO analysis.

Methodology and assumptions

To determine the fiscal cost of task-based IT contracting, PBO estimated the price differential between private contractors and their public service full-time equivalents. PBO assumed the public service equivalent to IT contractors to be those in the information technology (IT) occupational group, which includes “positions for which the application of comprehensive computer systems knowledge is the primary requirement to the development, implementation and/or maintenance of IT systems and infrastructure”.[^8]

To create a point of comparison against the current practice of hiring contractors, PBO assumed that all potential public service full time equivalents (FTEs) would be either the IT-04 classification (senior advisors, senior project managers, or managers), or the IT-05 classification (directors) if the contractor title referenced an executive or managerial role. This was done to select the senior most government position that could work on a contracted project, to reflect that less experienced workers may not have the skills to complete the work currently being done by external contractors.[^9]

For these public service FTEs, PBO assumed that they would be paid at the maximum amount for their position during the fiscal 2022-23 year. Because the IT collective agreement does not match the fiscal year, a weighted average was used to estimate 2022-23 salaries based on when the different pay scales went into effect. It was assumed that these employees would be paid for the full fiscal year, so long as the contractor they were replacing met the criteria we provided the departments. We then scaled the resulting salary by 27 per cent to account for Employee Benefit Plans to obtain total compensation.[^10][^11]

With total compensation per employee, we finally compared the total of the employees to the total paid to contractors based on their per diem rate and number of days worked. This notably excludes contractor and government fixed costs, with the focus of our analysis being on the replacement cost of the labour undertaken.[^12]

Several caveats must be taken into consideration. In our analysis we lacked the information needed to determine if a specific contract was completing a task that could be a one-off as far as the contracting department was concerned. We thus implicitly assumed that spending on public service personnel versus contractors reflects tasks that would happen on a regular basis. We also assumed that the current compensation that the government is offering is sufficient to attract workers with the skills needed to replace contractors.

Results

Table 3 presents the results for all departments except for CBSA, who was only able to provide a partial response to PBO’s information request. Details on the results from CBSA can be found in the subsequent section.

The four departments identified a total of 93 contracts in 2022-23 that were used for task-based informatics professional services, and where the contractors were paid on a per diem basis and worked more than 200 days for a single contract. ESDC had the fewest relevant contracts, at six, while SSC had the most, at 50. Under these 93 contracts, 330 contractors worked 200 or more days. ESDC had the fewest number of contractors with eleven meeting the PBO’s criteria, while SSC had 162.

A weighted average of number of days worked and cost per diem was calculated.[^13] There was broad consistency for the four departments across these two metrics. The average number of days worked ranged from 222 for ESDC to 238 for the DND. The average per diem rate ranged from $924.87 for SSC to $964.30 for ESDC.

PBO found that the price difference between external IT contractors and their public sector equivalents ranges from a low of 22.0 per cent (SSC) to a high of 25.7 per cent (DND) for the four departments analyzed. With the data that was provided, PBO is unable to determine the underlying reason for these premiums and whether these premiums were justified. [^14]

Canada Border Services Agency

PBO received only two months of data from CBSA following a lengthy information request process. In discussions with the agency, officials indicated that it was not possible to provide the full twelve months of data in a timely manner as this was not tracked in a central database, and that they would subsequently need to manually search each contract to retrieve the information. A smaller dataset was selected to allow for our analysis to proceed.

The two months provided were April 2022 and January 2023, which respectively represent the beginning of the fiscal year, and the month in which 200 working days have elapsed since the beginning of that fiscal year. Having only two months makes it difficult to compare the cost of using contractors versus FTEs. Compared to the other departments, it was not possible to screen out those contractors that worked less than 200 days in a fiscal year. For the contractors provided, only about 62 per cent had a non-zero number of days worked in the two months that were provided. It is also unknown how many days a contractor worked in the intervening period between those two months, as even for contractors that worked for both months there is a high degree of variability in the number of days worked. These uncertainties prevent any strong conclusions being made from the data, however it can still be used to provide some illustrative examples.

For the month of April, it was found that for contractors that billed seven or more days it would have been less expensive to have used employees instead. For the month of January several high per diem rates paid to contractors working that month resulted in it being less expensive in aggregate to use employees for the entire month than to have used contractors at any threshold of number of days worked. For contractors working in both months, this threshold was seven days. It must be noted that public service FTEs are not hired on a month-to-month basis, but generally for longer periods, along with the benefits that they receive as employees. Consequently, there may be differences between these illustrative examples, and what would have been estimated if the PBO had been provided the full year of data.