Regulatory Impact Assessment Under Client Pressure
You are the lead economist at a consultancy commissioned by a pipeline operator to produce a regulatory impact assessment for a 40-year tariff extension. Your draft uses the national guidance rate of 3.5% for intergenerational infrastructure and a 40-year appraisal period, giving a benefit-cost ratio of 1.28. The client's commercial director asks you to switch to a 7% discount rate and a 20-year horizon, 'in line with industry practice', which would cut the ratio to about 0.94. The director also asks you to delete a table drawn from a regulator dataset showing a rupture frequency of 0.7 per 1,000 km-years because it is 'outdated', and hints that a success fee depends on the regulator approving the project. A junior analyst objects that the 7% rate is inconsistent with national guidance, which requires justification and disclosure of any deviation. The regulator's template requires a central case, a sensitivity range and a statement of assumptions. The submission deadline is in ten days.
1. Identify the professional and ethical issues and the standards that apply.
Read the worked answer
Issue 1 - Advocacy disguised as analysis. Choosing 7% and 20 years to reach a desired ratio is reverse-engineering. The discount rate is not a preference parameter: at 3.5% the guidance reflects intergenerational welfare, while 7% reflects a private opportunity cost of capital. If the client wants the private-finance view, it belongs in a sensitivity, not in the central case.
Issue 2 - Truncating the appraisal horizon. A 40-year asset appraised over 20 years omits half the benefits and most of the late-life costs, including decommissioning. That is a specification error, not an assumption choice, unless the regulator's own guidance permits a shorter horizon.
Issue 3 - Suppression of material evidence. The rupture-frequency table is adverse evidence from a public dataset. Deleting it because it is inconvenient is misrepresentation, particularly where safety is a decision criterion. 'Outdated' is a testable claim: if newer data exist, update the table and disclose both; do not remove it.
Issue 4 - Undisclosed contingency fee. A success fee contingent on approval creates a direct conflict of interest that must be disclosed to the client and, where the regulator requires it, in the submission.
Standards. Objectivity and integrity, competence, confidentiality, disclosure of conflicts, and avoidance of misleading statements apply. The regulator's template independently requires a central case, a sensitivity range and a statement of assumptions, which the requested edits would breach. The analyst's objection is correct and should be documented.
2. Draft your response to the client, including what you will and will not change and why.
Read the worked answer
What I will not change. (1) The central case stays at 3.5% over 40 years, because national guidance governs and a departure must be justified rather than instructed. (2) The rupture-frequency table stays, updated if newer data exist, with its source and vintage stated. (3) The success fee must be disclosed, or the engagement renegotiated to a fixed fee; I will not accept a payment conditional on approval.
What I will add. (1) A sensitivity case at 7% over 20 years, clearly labelled as a financial-opportunity-cost variant rather than the appraisal base, with the 0.94 ratio reported. (2) A fuller sensitivity grid over the discount rate (3.5%, 5%, 7%) and the horizon (20, 30, 40 years), so the reader sees how much the conclusion depends on assumptions. (3) An explicit statement that the 7%/20-year variant is inconsistent with national guidance and why. (4) A safety scenario using the rupture frequency, showing expected cost and the implied value of mitigation.
How to say it. Frame the response around the regulator's template and audit risk, not around personal ethics: a submission that departs from guidance without disclosure is exposed to challenge, freedom-of-information requests and possible rejection, which harms the client more than an honest sensitivity range. Offer to present the 7%/20-year case prominently in the executive summary so the client's commercial concern is addressed without compromising the analysis. Document the request and your response in writing, and copy the firm's quality partner.
3. If the client insists on the revised assumptions, what are your options, and how should the work be reported?
Read the worked answer
Option 1 - Comply with disclosure. Publish the 7%/20-year case as the client's requested scenario alongside the guidance-compliant central case, with a clear note that the requested assumptions were adopted at the client's instruction, not the analyst's recommendation. This is defensible only if the client does not control the labelling and the regulator sees both cases.
Option 2 - Decline to sign. If the client requires the 7%/20-year result to be presented as the central case, or requires removal of the rupture table, withdraw from authorship of that version. The firm can still provide the underlying model and a technical note, but it should not attach its name to a submission that misstates the guidance-compliant result.
Option 3 - Escalate. Refer the matter to the firm's quality and risk partner and, if the submission is already with the regulator, consider notifying the regulator of the changed assumptions. Professional obligations can override client confidentiality where a misleading submission is being made.
Reporting. Whichever route is taken: keep a dated record of the request, the analysis and the response; retain the original model and all versions; state authorship and assumptions in the submission; and ensure that any executive summary the client writes is checked against the model. A short technical annex listing the guidance rate, the client-requested rate, the horizon and the resulting ratios is the minimum disclosure. The reputational and regulatory cost of a misleading RIA exceeds the fee, and the junior analyst's objection should be recorded as a dissent that was considered, not overruled.