From Lapdogs to Watchdogs: the value of independence in auditing

Local governments worldwide are charged with making spending decisions and delivering public goods and services, in part because local officials are seen as more informed and responsive to local citizens’ needs.

However, whatever benefits that local knowledge might confer may be outweighed by the costs of corruption: local officials, captured by local elites, may be unable or unwilling to pursue social welfare maximizing policies. Oversight by external monitors is a common tool to ensure that local agents behave in the interests of the community. Unfortunately, oversight mechanisms are frequently ineffective, as monitors themselves might be corruptible.

To be effective, monitors need to be independent from the audited party. But ensuring independence is not easy and potentially very costly. For example, Brazilian random anti-corruption audits have proven to be effective but are also very expensive: each audit costs more than $50,000 and less than 2% of municipalities are audited every year.

Ensuring effective independent audits becomes even more important at times of emergency (as highlighted in another recent World Bank post), when governments finance recovery programs using expedited procedures, which are faster but also pose higher risk of mismanagement or corruption.

In my job market paper, I provide evidence of an inexpensive yet effective way of ensuring independent oversight. I exploit a unique natural experiment, resulting from a reform to the system governing the appointment of auditors overseeing municipal governments’ budgets in Italy. The main feature of the reform was a switch from a discretionary system in which mayors could choose their auditors, to a random assignment mechanism, in which auditors are randomly selected from a list of participating certified public accountants.

Italian municipalities are required to draft budgets and balance sheets that need to be reviewed and approved by professional auditors, who act as an intermediate layer of oversight between municipalities and the central monitoring performed by the National Court of Accounts. The audit system aims to ensure responsible spending practices, in compliance with national fiscal rules, which dictate no deficits and require effort towardrepayment of outstanding debts. As distinct from both third-party and anti-corruption auditors, who merely certify documents or verify expenditures ex-post, municipal auditors in Italy assist and oversee the entire budgetary process of municipalities and can thus influence municipal financial decisions ex-ante and in a more continuous manner due to their three-year long term.

The new rules became effective in December 2012, but its implementation took place upon the expiration of the current auditor’s term, which occurred at different dates across municipalities for historical reasons unrelated to the reform and beyond a mayor’s control. For example, the municipality of Rome had last appointed an auditor in January 2010, so it was first treated in January 2013, while Florence last appointed an auditor in September 2012, so it received a new randomly appointed auditor only in September 2015. My identification exploits the reform’s staggered introduction across municipalities in a generalized difference-in-differences setting, comparing the evolution of financial outcomes of municipalities treated earlier to those treated later, before and after the arrival of an independently appointed auditor.

The reform became effective in December 2012, but its implementation took place upon the expiration of the current auditor’s term, which occurred at different dates across municipalities for historical reasons unrelated to the reform and beyond a mayor’s control. For example, the municipality of Rome had last appointed an auditor in January 2010, so it was first treated in January 2013, while Florence last appointed an auditor in September 2012, so it received a new randomly appointed auditor only in September 2015. My identification exploits the reform’s staggered introduction across municipalities in a generalized difference-in-differences setting, comparing the evolution of financial outcomes of municipalities treated earlier to those treated later, before and after the arrival of an independently appointed auditor.

Do independent auditors improve financial health?

Upon the arrival of a draft-appointed auditor, municipalities increase their expenditure in debt repayments by 9 % and their net budget surplus by 8 %. To give a sense of these magnitudes, the increase in the net surplus is about 1% of the overall annual budget, and 2% of the annual discretionary budget.

Effects appear immediately upon the arrival of the (randomly assigned) auditor and remain essentially constant over the auditor’s three-year term, suggesting an immediate and persistent shift in budgetary practices.

How do municipalities improve their fiscal stance?

Municipalities can either increase their revenues (e.g., by collecting more taxes) or cut spending (e.g., by cutting investments or current expenditures), or a combination of the two.

I show that they do not cut spending, but instead improve their fiscal position through revenue-based adjustment, and specifically by increasing revenues from the local property tax.

How does auditors’ random assignment work ?

I provide suggestive evidence that random assignment breaks up potential collusive ties between auditors and mayors.

The reform is more effective in:

• municipalities where collusion was more likely to happen under discretionary appointments (as proxied by worse financial conditions and corruption investigations)

• municipalities in which collusion is less likely to happen under random assignment (i.e., those that happen to receive an auditor coming from farther away or someone that is new to the profession).

What do we learn and why does it matter?

While considered a crucial tool for good governance, effective monitoring is frequently impaired by conflicts of interests. My results highlight that changes in the design of monitoring institutions can significantly improve governance outcomes. There are two important policy takeaways. One pertains to the cost-effectiveness of the reform. Rather than introducing a new auditing system (like the federal audits in Brazil) or an additional supervisory board (like the PCAOB for US audit firms), the reform operates by introducing frictions that make potential collusive agreements harder, while leaving features of institutional design unchanged. In this sense, the benefits come with zero cost.

The second lesson relates to the identity of the auditors. As distinct from other types of reforms, the auditors here are not central government’s bureaucrats, nor are they village members; they are certified public accountants hired directly by municipalities to carry out auditing duties. Auditing governments is only a limited part of their work activities. This suggests a potentially important role for the use of external professionals to carry out audit roles also in other settings.

While my research focuses on the Italian context, the findings may help to inform the broader debate about policies to ensure enforcement of rules and effective oversight more broadly, and may apply in both governmental and private settings.


by Silvia Vannutelli – Assistant professor of Economics at Northwestern University

Local Opportunities Lab

  • Milan
  • Turin
  • Rome
  • Paris
  • Nuremberg
  • Berkeley

Powered by

Algebris Policy & Research Forum

Contacts

Newsletter


© 2020 Local Opportunities Lab. All rights reserved.