Local politicians’ perceptions of financial conditions – do they align with financial indicators?

ABSTRACT

Throughout history, critics of democracy have doubted people’s ability to govern themselves. For example, are local politicians – who often lack any formal training in accounting or economics – really able to comprehend complex financial information? Previous studies addressing this issue are in short supply, and their results are generally pessimistic. The aim of this paper is to test which – if any – financial indicators are related to local politicians’ perceptions of financial conditions in the setting of Swedish municipalities, thus extending earlier studies to a European context. Our analysis shows strong and significant correlations between most of the financial indicators tested and politicians’ perception of their municipality’s financial condition, including financial indicators capturing the levels of net income, equity ratio and tax rates. These findings provide optimism and strengthen arguments for the defenders of democratic government.

KEYWORDS:

• financial condition
• politicians
• local government
• local democracy
• Sweden

1. Introduction

Since political decision-making is constrained by the scarcity of resources, politicians at all tiers of government must ultimately appraise the financial situation in order to determine what political goals can be pursued. However, as most politicians have no formal training in accounting or economics, they may find it challenging to fully comprehend complex financial information (Ezzamel et al. 2005; Hyndman 2016). Considering this, van Helden and Reichard (2019, 485) categorise most politicians – along with a substantial number of public sector managers and citizens – as laymen users of financial information. In this respect, politicians’ perceptions of financial conditions relate to the much broader issue of the role of knowledge and expertise in politics, which is an inherent challenge in all democratic systems.

Since antiquity, meritocrats have questioned whether people are qualified to govern themselves; in modern societies based on a representative democracy, a related question is whether the people’s representatives possess the necessary knowledge to make decisions in a complicated world (Ober 2008; Bagg 2018). Several solutions have been proposed to resolve this dilemma. One is to elect political representatives from a highly educated elite, with the expectation that their learning will ensure high-quality decisions. However, this elitism risks endangering other democratic values, such as politicians’ understanding of the interests and concerns of ordinary people who are not part of the elite (Bovens and Wille 2017). Another solution is to establish systems wherein politicians can rely on experts for guidance, such as accounting bureaucrats. However, when politicians become heavily dependent upon experts that assume the role of information brokers, there is a tangible risk of real power sliding into the hands of bureaucrats (Jorge, de Jesus, and Nogueira 2016), who are unelected and unaccountable to voters (Heazle and Kane 2015). In light of these flawed solutions, pessimists might conclude that this dilemma cannot be resolved and that democratic systems are doomed to prioritise either true democracy or informed decision-making, while optimists may have more trust in the abilities of citizens and hope that laypeople can comprehend complicated information such as financial reporting.

The local tier of government is an ideal arena in which to test these presuppositions, as most local politicians are laypeople with limited support from experts. We achieve this task by analysing the relationship between two different ways of measuring the financial conditions of municipalities: one based on financial indicators, and the other on local politicians’ perceptions. If the pessimistic view described above is correct, the politicians will be unable to comprehend the financial information, and their perception of the financial condition of their municipality will not correlate with the financial indicators. If the optimistic view is correct and if the financial reporting is of a high quality – that is, if it provides complete, reliable and timely information – then a significant correlation will be found between at least some of the financial indicators and politicians’ perception of the financial condition of their municipality.

A relatively large body of literature exists that is centred on analysing the financial conditions of local governments (e.g., Zafra-Gómez, López-Hernández, and Hernández-Bastida 2009; Brusca, Manes Rossi, and Aversano 2015; Rodríguez Bolívar et al. 2018; Cuadrado-Ballesteros and Bisogno 2019). This literature covers a broad range of issues; however, politicians’ perception of financial condition is not one of those issues. In fact, aside from the setting of local governments in the United States, about which a small number of studies have been published (Maher and Deller 2011, 2013; Leiser and Mills 2019; Leiser, Wang, and Kargman 2021), to the best of our knowledge, there is no literature that addresses the relationship between financial indicators and politicians’ perceptions of financial conditions in a European context.

In many European welfare states, municipalities are responsible for the production of extensive and essential public services. As a result, locally elected political leaders are in control of a substantial share of public money. The aim of this paper is to test which – if any – financial indicators are related to local politicians’ perceptions of financial conditions in the setting of Swedish municipalities. Local politicians need to be able to assess complex financial information in order to make apt political decisions. The extent to which local politicians’ perception of financial conditions correspond with financial information is thus potentially crucial for the legitimacy of welfare state models built on self-governance and local democracy.

Swedish local governments are arguably a critical case for testing the knowledge of local politicians in Europe, as Swedish local politicians have a relatively low educational level compared with their counterparts in most other countries (Verhelst, Reynaert, and Steyvers 2013). According to data available to us from Statistics Sweden, about 53% of Swedish local councillors had some form of university education in 2017, while less than 4% had specialised in accounting or economics.

The rest of the paper is structured in six main sections. Section 2 presents the research context and Section 3 discusses the literature related to financial conditions at the local tier of government. Section 4 includes our hypotheses, and Section 5 describes the method used to test these hypotheses. Section 6 analyses the relationship between financial indicators and local politicians’ perceptions of the financial condition of their municipality. Finally, Section 7 offers some conclusions.

2. Research context – the Swedish case

Compared with many other countries, Sweden has large municipalities with extensive political organisations, which deal with issues across the entire political spectrum (Montin 2015). All major decisions regarding the budget and annual report are made collectively by the local council and its 21–101 elected councillors (the size of the council depends on the size of the municipality). In terms of representing social groups in society, elected local politicians in Sweden are relatively more representative of the population than their colleagues in most other European countries. Although the educational level of these local politicians is somewhat higher than that of the average Swedish citizen, it is considerably lower than that of local politicians in other European countries, and the proportion of politicians with a working-class background is higher in Sweden (Verhelst, Reynaert, and Steyvers 2013; Karlsson 2013). In addition, local politicians in Sweden are responsible for making decisions on taxation and fees, as well as holding the responsibility for a local budget that is larger than that of municipalities in most other countries. The main source of revenue in a Swedish municipality is municipal taxes, which are levied on the personal income of municipal residents. According to The Instrument of Government, Swedish municipalities have extensive freedom to set their own taxes and decide on how they want to organise their service delivery. Another specific feature is Sweden’s far-reaching equalisation system, which distributes and redistributes resources in order to ensure that municipalities with very different conditions have a similar quality of services without large differences in tax and fees. In terms of budget and budget execution, the central government has imposed several mandatory requirements that are tied to the information produced by the currently used financial reporting system.

Financial reporting in the Swedish municipal sector was harmonised during the 1980s and 1990s. Accrual-based accounting was implemented through a set of voluntary generally accepted accounting principles (GAAP) established in 1986; from 1998 onwards, accrual-based municipal accounting legalisation and mandatory GAAP were applied (Tagesson and Grossi 2015). This reporting framework relies on what Oulasvirta (2021, 443) has labelled as ‘(…) the transaction-based, historical costs and income-statement-first approach’. This approach is also reflected in the regulatory constraints imposed on municipalities’ budgets.

As of 1992, when a new Local Government Act (no. 1991:900) took effect, Swedish municipalities are required by law to enact sound economic behaviour (god ekonomisk hushållning). This requirement includes both an operational and financial perspective. The latter is closely tied to the concept of intergenerational equity, meaning that every generation should carry the costs of its own services (government bill no. 1990/91:117). Later, in the bill preparing the amended Local Government Act that took effect in 2005, it was also emphasised that sound economic behaviour means that municipalities must be financially prepared to adapt to events that decrease revenues and/or increase expenses, without having to drastically reduce the quality of the services provided (government bill no. 2003/04:105).

As part of the amended Act of 2005, it became mandatory for municipalities to adopt financial targets, in which they define the financial performance necessary to achieve the legal requirement of sound economic behaviour. Considering that property, plants and equipment are valued at historical cost (Tagesson and Grossi 2015), there will normally be a gap over time between depreciation expenses and the capital expenditure required to replace these assets after their useful life has ended. Depending on the required renewal or replacement of assets, higher or lower levels of net income may be necessary to safeguard intergenerational equity, keep borrowing under control, and thus maintain a financial condition that allows the municipality to adapt to changes in economic conditions. A rule of thumb that is widely used in practice, as demonstrated by the adopted financial targets, is that a positive net income of 2% of tax revenues and grants is consistent with sound economic behaviour (Donatella et al. 2022).

The Local Government Act also includes a balanced-budget requirement, which has been applied since 2000. This requirement is tied to the operating statement: after adjusting for capital gains and losses and for unrealised losses on securities and reversals from such unrealised losses, revenues should cover expenses. Unless it is due to an exceptional situation, a negative net income must be restored by an equivalent positive net income within the next three years. As was made clear in the bills preparing the legalisation, the balanced-budget requirement is merely a short-term minimum threshold. Sound economic behaviour normally requires net income levels to be on a significant higher level than breaking even (government bills no. 1996/97:52; no. 2003/04:105).

3. Previous literature on financial condition in local government

Local governments’ financial conditions have attracted a considerable amount of academic attention. One stream of research focuses on identifying the factors that affect the financial conditions of local governments. The variables typically included in this research are economic and demographic factors (e.g., Rodríguez Bolívar et al. 2014; Drew and Dollery 2016; Navarro Galera et al. 2016). Other analyses are broader in nature; in addition to economic and demographic factors, they include the impact of political factors (García-Sánchez, Mordán, and Cuadrado-Ballesteros 2014; Rodríguez Bolívar et al. 2018) or how the institutional setting moderates the impact of certain factors (Brusca, Manes Rossi, and Aversano 2015; Cuadrado-Ballesteros and Bisogno 2019; Rodríguez Bolívar et al. 2021). Yet another stream of research is centred on developing and validating methodologies to identify entities on an unhealthy financial path (e.g., Kloha, Weissert, and Kleine 2005; Zafra-Gómez, López-Hernández, and Hernández-Bastida 2009; Clark 2015; Navarro Galera et al. 2017; Gorina, Maher, and Joffe 2018).

In contrast to the above-mentioned topics, little attention has been paid to empirical investigations of the relationship between financial indicators and elected and administrative officials’ perception of financial conditions. As previously mentioned, the few prior studies that do exist are based on data from local governments in the United States. Except for a more recent study by Leiser, Wang, and Kargman (2021), all these studies produced results indicating that financial indicators are unable to offer a robust explanatory and predictive power regarding elected and administrative officials’ perception of the financial condition of their local government (Maher and Deller 2011, 2013; Leiser and Mills 2019). When interpreting these results, much of the discussion has focused on whether ‘ … the limited amount of association … is a function of measurement error or bias in the manner in which county officers respond to the survey questions’ (Maher and Deller 2013, 134).

Such concerns over measurement errors relate to the possibility of the financial indicators being unable to capture the financial conditions of local governments. Two kinds of biases are debated in the literature, both of which are related to politicians’ perception of the financial condition of their local government. The first bias comprises the existence of intentional biases brought about by local government officials being unmotivated to provide their actual opinion in surveys and instead perceiving an opportunity to convey a message to resource providers (Maher and Deller 2011). Considering that the previous literature relies on survey data administered by higher levels of government as part of their surveillance of local governments within their jurisdictions, this bias is obviously a very valid concern. The second bias, which was first suggested by Maher and Deller (2011), comprises the possible existence of unintentional biases in the survey data due to local government officials having difficulty fully comprehending the financial condition of their local government. However, Leiser and Mills (2019, 80) disagree with concerns regarding this bias, depicting it as an ‘ … unnecessarily pessimistic view because local officials are uniquely positioned to shed light on areas where objective measures are known to fall short’. For example, there might be an asset maintenance and renewal backlog (Dollery and Grant 2011) or some other type of issue that is not recognised in the financial reporting and that is therefore not inevitably captured by financial indicators calculated based on information from financial statements.

As many developed countries, including Sweden, have shifted from reporting public money on a cash basis to reporting it on an accrual basis, whether or not unintentional biases enter politicians’ perceptions of the financial conditions of their local government may be dependent on other contextual factors as well. After all, the literature tends to conclude that most politicians have neither the time nor the training to use and properly understand complex financial information (van Helden and Reichard 2019). Furthermore, the complexity of financial statements has certainly increased under accrual-based reporting, which may very well have had a negative impact on politicians’ ability to fully comprehend the financial situation (Ezzamel et al. 2005; Hyndman 2016). However, over time, the public financial management system itself (as well as the users of the information it produces) is likely to see a change in terms of the commitment to rely on accrual-based information (Bergmann 2012). This suggests that a strong relationship between accrual-based financial indicators and politicians’ perceptions of the financial condition of their local government is more likely to exist in mature settings, rather than in settings in which accrual-based accounting was recently adopted.

4. Hypotheses

On the one hand, most prior studies directly (Maher and Deller 2011, 2013; Leiser and Mills 2019) or indirectly (van Helden and Reichard 2019) suggest that financial indicators explain little of the variation in politicians’ perception of the financial condition of their local government. This is a pessimistic stance for the advocates of local democracy, as it questions the abilities of the people’s elected representatives to make sound political decisions based on accurate information on available resources. On the other hand, the results produced in previous studies may be a function of issues that are country-specific or related to the survey instruments used. This perspective suggests that there are some grounds for optimism and that caution should be employed in generalising U.S. findings to other settings. More research is needed to provide further insight into whether a weak association between financial indicators and politicians’ perceptions of financial condition is due to features that are specific to previous U.S. studies, or whether this is also the situation in other country settings.

In order to test whether the conclusions of U.S. studies are valid in a European context, we will analyse the correlations between Swedish local politicians’ perception of the financial conditions of their municipality and seven financial indicators. As the expectations from earlier studies are pessimistic, we have formulated seven hypotheses – one for each indicator – predicting that the indicators and politicians’ perceptions are unrelated to one another. If all the hypotheses are retained, the pessimistic view of local politicians’ ability to assess financial information will be confirmed. For each hypothesis that falls, a more optimistic perspective will become tenable.

In relation to the short-term financial performance of local governments, the literature tends to focus on the income statement and, more specifically, on the extent to which revenues cover expenses (e.g., Carmeli 2002; Dollery and Grant 2011; Turley, Geraldine, and McNena 2015; Navarro Galera et al. 2016; Rodríguez Bolívar et al. 2018). Entities that do not achieve a balance between revenues and expenses over time are, by definition, on an unhealthy financial path. It is also necessary to consider the legal system that applies to Swedish municipalities. As previously explained, a reported deficit would probably constitute a breach of the requirement for a balanced budget. Taking this into account, we hypothesise that:

H1:

Local politicians’ perception of their municipality’s financial condition is unrelated to reported deficits.

Moreover, it is important to consider the extent to which the reported net income generates a cash flow that covers capital expenditures (Dollery and Grant 2011; Bergmann 2012). In particular, if property, plants and equipment are valued at historical cost, which is the case in Swedish local government, there will be a gap over time between depreciation expenses and the capital expenditure required to replace these assets after their useful life has ended. Depending on the necessary renewal or replacement of assets, higher or lower levels of net income may be necessary to keep borrowing under control. As previously explained, a widely used rule of thumb in Sweden is that a positive net income of at least 2% of tax revenues and grants are consistent with a municipality being on a healthy financial path. Taking this into account, we hypothesise that:

H2:

Local politicians’ perception of their municipality’s financial condition is unrelated to whether or not the municipality has a reported net income equal to or greater than 2% of its tax revenues and grants.

A more direct measure of the extent to which a municipality self-finances its investments is the self-financing of investments ratio (Bergmann 2012). Like the level of reported net income, a consistently low self-financing of investments ratio signals that an increase in borrowing is required, which in turn may shift the proportion of future revenue that will be required to cover interest expenses, thus leaving fewer resources available for service delivery. Taking this into account, we hypothesise that:

H3:

Local politicians’ perception of their municipality’s financial condition is unrelated to the municipality’s self-financing of investments ratio.

Regarding the long-term financial performance of local governments, the literature tends to rely on financial indicators deducted from the balance sheet rather than the income statement (Brusca, Manes Rossi, and Aversano 2015; Turley, Geraldine, and McNena 2015; Navarro Galera et al. 2017). The relationship between assets and liabilities is one such widely used financial indicator (Iacuzzi 2021). Considering the reporting framework applied in Swedish municipalities, if an entity has performed well historically in terms of reported net income, and therefore has been able to self-finance a large proportion of its investments, such an entity will have a relatively high equity ratio and thus be less vulnerable financially. In general, there will then be more room for a temporary change in debt policy to, for example, adapt to changes in macroeconomic conditions or other types of economic events that decrease revenues and/or increase expenses. Taking this into account, we hypothesise that:

H4:

Local politicians’ perception of their municipality’s financial condition is unrelated to the municipality’s equity ratio.

Another financial aspect to consider is the revenue base (Carmeli 2002; Turley, Geraldine, and McNena 2015), which partly relates to both the short- and the long-term financial performance. One argument is that dependence on external revenues may impede a municipality’s flexibility to adapt to changing economic circumstances (Zafra-Gómez, López-Hernández, and Hernández-Bastida 2009; Dollery and Grant 2011). More specifically, in the empirical setting of this study, Swedish municipalities have two main sources of revenues: taxation of their residents (own revenues) and general-purpose grants from the national government (external revenues). The general-purpose grants are integrated in an equalisation system, which the Swedish central government uses to distribute and redistribute resources. In such a system, all municipalities are exposed to the risk of changing revenues due to amended equalisation policies. However, it can be argued that municipalities that are more reliant on central government grants carry a higher risk and are thus more financially vulnerable. Taking this into account, we hypothesise that:

H5:

Local politicians’ perception of their municipality’s financial condition is unrelated to the municipality’s grant/own tax revenue ratio.

Another important contextual feature to consider when discussing the revenue base in Swedish municipalities is that they have the authority to set their own tax rate. Historically, tax policies have been an ideological issue, often dividing right- and left-wing parties in Sweden. The left-wing parties have advocated higher taxes and a larger welfare state, with right-wing parties taking the opposite position and advocating for lower taxes and a smaller welfare state (Karlsson and Gilljam 2014 chapter 2; Gilljam and Karlsson 2015; Högström and Lidén 2022). Despite these different ideological stances on tax policy issues, a comparatively low municipal tax levied can be regarded as a capacity to increase revenues to address a change in economic circumstances. In the Swedish context, municipal tax rates are therefore often considered to be a financial indicator per se (Donatella et al. 2022). Taking this into account, we hypothesise that:

H6:

Local politicians’ perception of their municipality’s financial condition is unrelated to the municipal tax rate.

Considering the ideological nature of tax policies, a municipal tax raise is likely to spur internal political debate within local political assemblies, as well as external debate in the local mass media and among the general public. When discussing the revenue base, the actual amount of tax levied may therefore not be the only factor to consider. If a change in the municipal tax rate occurred recently, it could affect the political viability of future changes. Taking this into account, we hypothesise that:

H7:

Local politicians’ perception of their municipality’s financial condition is unrelated to changes in municipal tax rate.

5. Method

The hypotheses will be tested using multiple ordinary least-squares regression (OLS) models with one main dependent variable and several independent variables, including financial indicators and control variables.

5.1 Dependent variable measurements

Local politicians’ perception of the financial condition of their municipality was measured by a question in the KOLFU survey of 2017, which was directed at all elected councillors in the 20 regions and 290 municipalities¹ of Sweden (Karlsson, 2017). Unlike prior studies mentioned earlier, this was a survey instrument developed and used by researchers, rather than by resource providers. The issue of intentional response biases should therefore be less of a concern.

Considering the legal requirement of sound economic behaviour and the language commonly used in the Swedish municipal context, the following survey question was formulated: ‘How would you, on the whole, assess the state of the following factors in your municipality? – The financial situation of the municipality’. The politicians responded on an eleven-point scale, from 0 (very bad) to 10 (very good). A total of 6,573 out of 11,870 councillors answered the question, resulting in a response rate of 55%. We received answers from more than half of the councillors in a majority of the municipalities (56%), and we had a response rate of 40% or more in almost all municipalities (93%).

As previously explained, all political decisions in Swedish local governments are made collectively, and the most important decisions – including those related to the annual budget and the annual report – are always made by the council. Even if there is a variation in perception among the councillors within the councils, an appraisal of the average councillor is an apt indicator for the overall position of this political body.² Therefore, all analyses in this paper are carried out at the municipal level, with the average perception of politicians in each municipality as the dependent variable.

Descriptive statistics covering the perception of politicians are reported in Table 1. More detailed information regarding the survey and this variable is presented in the online appendix.

Table 1. The dependent variable: perception of the financial condition.

	Min	Q1	Median (Q2)	Q3	Max	Mean	SD
(A1) Politicians’ average perception of the financial situation 2017 (N = 289 municipalities)	1.27	5.40	6.54	7.67	9.50	6.38	1.62
(A2) Politicians’ average perception of the financial situation 2017 (N = 176 municipalities)	1.27	5.70	6.64	7.74	9.50	6.49	1.62

The table presents results of how local politicians in Sweden assess the financial condition of their municipality on a scale from 0 (very bad) to 10 (very good). Line A1 presents the results for politicians in all 289 municipalities, as the municipality of Gotland is excluded due to its status as both a municipality and a region, while line A2 presents the same results in the 176 municipalities where responses from the CFOs are available.

For all the municipalities in Sweden, the mean value among the surveyed politicians on the scale from 0 (very bad financial situation) to 10 (very good financial situation) was 6.38 (standard deviation 1.62), indicating a generally positive view of the municipalities’ condition. As 5 represents a neutral position on the scale, the condition was regarded as ‘bad’ in less than 25% of the municipalities. The municipality with the highest value had 9.5 and the lowest had 1.27. Half of the municipalities had values between 5.40 and 7.67.

5.2 Independent and control variable measurements

We apply seven financial indicators as independent variables in our models, to test the proposed hypotheses. In addition, we control for eight potential confounders. Descriptive information regarding the independent and control variables is presented in Table 2.

Table 2. Descriptive information regarding independent variables.

	Original form							Transformed: 0=min, 1=max
	Min	Q1	Median (Q2)	Q3	Max	Mean	SD	Min	Q1	Median (Q2)	Q3	Max	Mean	SD
Financial indicators
(1) Number of deficits 2014–2016	0	0	0	0	3	.31	.60	0	0	0	.33	1	.10	.20
(2) Number of net income ≥2% 2014–2016	0	1	2	2	3	1.64	1.01	0	.33	.67	.67	1	.55	.34
(3) Self-financing investment ratio 2014–2016 (%)	−41	66	90	126	299	104	60	0	.31	.39	.49	1	.43	.18
(4) Equity ratio 2016 (%)	−34	2	14	27	70	14	20	0	.46	.55	.66	1	.56	.16
(5) Tax rate 2017 (%)	17.12	21.08	21.84	22.45	23.85	21.62	1.23	0	.59	.70	.79	1	.67	.18
(6) Changes in tax rate 2014–2017 (percentage points)	−.92	0	0	.25	1.25	.12	.28	0	.42	.42	.54	1	.48	.13
(7) Grant dependency 2016 (%)	−24	21	32	43	84	32	17	0	.42	.51	.62	1	.52	.16
Control variables
(8) Population size 2017 (1,000s)	3	10	16	35	950	35	72	0	.01	.01	.03	1	.03	.08
(9) Population change 2014–2017 (%)	−5.3	+1.6	+2.8	+4.2	+12.2	+2.9	+2.3	0	.39	.47	.54	1	.47	.13
(10) Unemployment rate 2017 (%)	5.5	10.5	13.3	15.7	23.4	13.3	3.8	0	.16	.43	.57	1	.43	.21
(11) Population density 2017 (inh/km^2)	.22	13	28	78	5700	154	559	0	.002	.005	.37	1	.03	.10
(12) Share of population 65 years and older 2017 (%)	13	20	24	33	34	23	4	0	.36	.54	.65	1	.50	.20
(13) Share of population born abroad 2017 (%)	6	11	13	18	42	15	6	0	.13	.21	.33	1	.25	.17
(14) Change in taxable income levels in population 2016–2017 (%)	+1.1	+3.1	+3.6	+4.2	+7.7	+3.6	+1.0	+/-0	.36	.44	.52	1	.44	.14
(15) CFO perception of the financial situation 2017 (N = 176 municipalities)	1	5	7	8	10	6.51	1.92	0	.44	.67	.78	1	.61	.21

The data refers to the statistics of independent variables in both their original form and a transformed form, where all variables are recoded so that 0 represents the minimum value and 1 the maximum value. In the regression models presented in Table 3, the transformed version of the variables is used in order to make the size of B-coefficients easier to compare. For financial indicators, the number of cases is 289, as the municipality of Gotland is excluded due to its status as both a municipality and a region. The CFO variable is based on 176 observations.

Table 3. Factors explaining politicians’ perception of the financial condition in their municipality (OLS regression, B-values and [SE]).

Dependent variable:	Politicians’ perception(N = 289)				Politicians’ perception(N = 176)
Independent variables:	POLbiv	POL1	POL2	POL3a	POL3b	POL4
Number of deficits (2014–2016)	−3.38***	−1.57***		−1.42**	−1.19*	−0.92
	[0.44]	[0.41]		[0.41]	[0.59]	[0.56]
Number of net income ≥2% (2014–2016)	+2.40***	+1.76***		+1.80***	+1.65***	+1.24***
	[0.25]	[0.27]		[0.27]	[0.34]	[0.33]
Self-financing investment ratio (2014–2016)	+0.66	−0.30		−0.24	0.00	+0.34
	[0.54]	[0.50]		[0.51]	[0.66]	[0.63]
Equity ratio (2016)	+4.19***	+2.78***		+3.29***	+3.52***	+2.73***
	[0.55]	[0.52]		[0.53]	[0.71]	[0.70]
Tax rate (2017)	−3.61***	−1.09*		−1.53*	−2.44**	−1.81*
	[0.48]	[0.55]		[0.64]	[0.92]	[0.88]
Changes in tax rate (2014–2017)	−1.46*	−0.46		−0.64	−1.01	−0.89
	[0.74]	[0.58]		[0.58]	[0.85]	[0.80]
Grant dependency (2016)	−2.66***	−0.64		−1.74*	−1.00	−0.94
	[0.59]	[0.58]		[0.76]	[1.10]	[1.04]
Population size (2017)	+1.83		−0.29	−0.46	−0.41	−1.11
	[1.25]		[1.57]	[1.23]	[2.09]	[1.99]
Population change (2014–2017)	+3.20***		2.45*	+1.61	+2.39*	+1.91
	[0.69]		[1.00]	[0.82]	[1.06]	[1.01]
Unemployment rate (2017)	−1.08*		−0.71	+1.01	+0.92	+0.75
	[0.45]		[0.62]	[0.56]	[0.73]	[0.69]
Population density (2017)	+2.58**		+1.09	−0.66	−1.97	−1.89
	[0.96]		[1.25]	[1.00]	[2.71]	[2.57]
Share of population 65 years and older (2017)	−1.75***		−0.05	+1.41*	+1.99*	+1.53
	[0.48]		[0.76]	[0.64]	[0.89]	[0.85]
Share of population born abroad (2017)	+0.96		+0.44	−0.61	−0.86	−0.81
	[0.55]		[0.78]	[0.64]	[0.83]	[0.79]
Change in taxable income levels in population 2016–2017	+0.11		+0.09	−0.61	−0.06	+0.09
	[0.70]		[0.74]	[0.58]	[0.71]	[0.67]
CFO perception (N = 176)	+4.30***					+2.21***
	[0.48]					[0.50]
Constant		5.45***	5.40***	4.60***	4.12***	3.24***
		[0.58]	[0.90]	[1.02]	[1.37]	[1.31]
Adj. R^2		.43	.06	.45	.43	.48

All dependent variables refer to politicians’ average perception of the financial condition of their municipality on a scale of 0 (very bad) to 10 (very good). All independent variables are transformed to a scale from 0 (minimum value) to 1 (maximum value). In models POL1, POL2 and POL3a, all municipalities but one is included (N = 289). Gotland has been excluded because of its co-status as a municipality and a region, which means that its financial indicators are not comparable with those of other municipalities. CFO models are based on 176 observations due to non-responses in the CFO survey. POLbiv (bivariate effects) is based on 289 observations, except for the CFO variable. P-values: *= P < 0.05; **= P < 0.01; ***= P < 0.001.

The number of deficits is measured over the 2014–2016 period – that is, the years from the term of office that were available at the time the survey was conducted, using net income before extraordinary items (indicator 1 in Table 2). During this three-year period, 19% of the municipalities reported deficits on one occasion, and 6% on two or three occasions (.31 on average). The number of reported net income that are equal to or greater than 2% of the tax revenues and grants is also measured over the 2014–2016 period, using net income before extraordinary items (indicator 2). On average, municipalities reported such surpluses 1.6 times during the 3-year period.

The self-financing of investments ratio is defined as the total of net income and deprecation expenses in relation to the net investments over the 2014–2016 period (indicator 3). The average ratio over this period was 104. The equity ratio (indicator 4) includes both on- and off-balance sheet pension liabilities. On average, the equity ratio of Swedish municipalities in 2016 was 14%. The municipal tax rate was measured based on the tax levied for 2017 (indicator 5) and the tax rate changes throughout the 2014–2017 period (indicator 6). The average municipal tax rate in 2017 was 21.62%, and the average tax rate increased by 0.12 percentage points during the four-year period. Dependence on external revenues was measured for 2016 by scaling the grants each municipality received with their own tax revenues (indicator 7). The average grant/own tax revenue ratio in 2016 was 32%.

In addition to the financial indicators, we control for several socioeconomic and other population characteristic factors. Following the prior literature on financial conditions in local governments (e.g., Zafra-Gómez, López-Hernández, and Hernández-Bastida 2009; Capalbo and Grossi 2014; Brusca, Manes Rossi, and Aversano 2015; Clark 2015; Drew and Dollery 2016; Navarro Galera et al. 2017), the relevant factors include: population size (indicator 8, average 35,000 people); population change (indicator 9, average +2.9%); unemployment rate (indicator 10, average 13%); population density (indicator 11, average 154 inh/km²); share of population 65 years and older (indicator 12, average 23%);  share of population born abroad (indicator 13, average 15%); and change in taxable income levels in the population (indicator 14, average +3.6%). Although an equalisation system is used in Sweden, these factors may still affect the revenues and expenses of a municipality. It is also relevant to check whether local politicians are more prone to basing their perception of the financial condition of their municipality on factors related to the local society, rather than on the financial ratios discussed above.

Finally, as the senior local bureaucrats in relation to financial matters, the local chief financial officers (CFOs) are likely to be important information brokers in the context of Swedish municipalities. Therefore, in the same year as the KOLFU survey, the same question about the financial condition was included in a separate survey of CFOs in Swedish municipalities. The results from this study allow us to present models in which the CFOs’ perception of the financial condition of each municipality is used as a control variable in order to appraise the impact of expert advice on politicians’ perceptions of financial condition.

We received responses from 177 CFOs/municipalities out of 290 (or 176 of 289 after excluding Gotland); in these cases, we were able to compare the perception of the CFO with the perception of the average politician in the same municipality. The perceived value reported by the average CFO was 6.51, indicating a marginally more positive view of the financial condition than the average politician in the same municipalities (6.49), although the difference is not statistically significant (t-statistic = −0.16, p = .87). In 40 cases (23% of the 176 municipalities), the average politician’s response was within 0.5 of the CFO’s response. In 63 municipalities (36%), the CFO had a more positive appraisal, and in 74 municipalities (42%), the politicians were more positive.

All independent and control variables (except the CFOs’ perceptions) are based on publicly available data from Statistics Sweden. However, when no values for financial indicators were available, data was collected manually from the annual report of that municipality.

6. Results

In this section, we will test our seven hypotheses by analysing the relationship between financial ratios and Swedish local politicians’ perception of the financial condition of their municipality. The regression results are presented in Table 3. We start by presenting the bivariate effects (B-values) of all independent and control variables (model POLbiv). Multivariate effects are subsequently presented separately for financial indicators (model POL1) and for a set of control variables (model POL2), and are then presented jointly (models POL3a/POL3b). In the final model, CFOs’ perceptions are introduced as an explanatory variable (model POL4).

The bivariate results in model POLbiv show that six out of the seven financial indicators are significantly correlated with politicians’ perception (with the self-financing of investments ratio being the exception). It should also be noted that model POLbiv shows that there is a strong positive correlation between the perceptions of CFOs and those of politicians (B = +4.30, Adj. R² = .31). This finding means that the perceptions of politicians and CFOs are generally aligned.

In the multiple models POL1 and POL3a/b, the effects of four out of seven indicators remain significant in all models, and one more (grant dependency) is significant in POL3a. The indicator most strongly related to politicians’ perception is the equity ratio. Indicators concerning deficits, surpluses and tax rates are also significantly correlated with politicians’ perception, but the sizes of the coefficients for these variables are smaller. Model POL1 produces an impressive Adj. R² of .43.

In model POL2, the explanatory power of the control variables is much weaker than that of the financial indicators in POL1. Of the control variables, only the population change variables are significant across several models. These results further strengthen the conclusion that there is a direct relationship between several of the financial indicators tested and politicians’ perceptions.

In model POL4, in which CFOs’ perceptions are introduced as a final control variable, the effects of three out of seven financial indicators remain significant. The deficit variable falls just outside the threshold of significance at the 5% level. A comparison of the coefficient estimates and standard errors for POL3a and POL3b suggests that this result is partly driven by the 113 observations excluded from POL3b. The explanatory power of the model POL4 is .48, which is .17 stronger than the bivariate explanatory power of the CFOs’ perceptions. This result indicates that, even though the perceptions of politicians coincide overall with those of the CFOs, the politicians’ perceptions are even more strongly related to financial indicators – mainly the equity ratio and tax rate.

Overall, the regression results lead us to reject H1, H2, H4, H6 and H7 and only retain H3 and H5. The five rejected hypotheses suggest that local politicians’ perception of the financial condition of their municipality is indeed – against pessimistic expectations – related to their municipality’s deficits, net income ≥2%, equity ratio, tax rate and grant dependency.

In the online appendix, we provide an additional analysis based on an extended period for some of the financial indicators. This analysis shows that the results are not sensitive to the number of deficits, number of net income ≥2%, average self-financing of investments and tax rate changes being calculated based on the last two terms of office, instead of the last term. We also analyse the deficit and surplus variables calculated after extraordinary items instead of before them, and find that the different approaches yield the same basic results. Moreover, as the CFOs’ perceptions provide a relevant benchmark for evaluating politicians’ perceptions, in the additional analysis, we systematically compare the bivariate and multivariate regression results using both the CFOs’ perception and the politicians’ perception of financial condition as independent variables. We find that the CFOs’ perceptions are also strongly related to several of the financial indicators, but to a lesser extent than the politicians’ perceptions.

7. Conclusions

Many scholars have expressed their concern that, as a user group, politicians may have neither the time nor the training to use and properly comprehend complex financial information (e.g., Ezzamel et al. 2005; Hyndman 2016; van Helden and Reichard 2019). However, unlike several studies in the context of U.S. local governments (Maher and Deller 2011, 2013; Leiser and Mills 2019), our analysis shows strong and significant correlations between several of the financial indicators tested and politicians’ perception of the financial condition of their municipality. These financial indicators include those capturing the levels of net income, equity ratio, tax rates and grant dependency.

On the one hand, our findings can be understood as a validation of the financial indicators used in the analysis. On the other hand, if one is willing to accept that these financial indicators are measures that accurately capture the financial condition of Swedish municipalities, then these findings suggest that, over time, harmonised accrual-based reporting can serve to make the average politician reasonably well-informed – that is, although detailed matters may not be fully understood, at least the overall financial condition of a local government can become common knowledge among members of political bodies. In this sense, our results offer a very different and much more optimistic view of the situation compared with those of previous studies.

Following the 2008 financial crisis, the financial conditions of public-sector organisations have been on the agenda of practitioners, standard setters and academics alike (Navarro Galera et al. 2016; Rodríguez Bolívar et al. 2018; Caruana et al. 2019). As the financial condition of most tiers of government was exacerbated by the Covid-19 pandemic, financial health initiatives are likely to remain on the agenda for the foreseeable future. Moving forward, a potentially wide audience could benefit from more empirical research. Expanding this research to other country settings would be helpful in offering a better understanding of whether a strong or weak relationship between financial indicators and politicians’ perceptions of financial conditions is the anomaly. Indeed, it would certainly be a cause for concern if local politicians’ perceptions were unrelated to the most important information produced by the financial reporting system used in a specific setting.

Finally, from a democratic perspective, our findings are very reassuring. Fundamental democratic values are at risk if the public financial reporting system fails to provide politicians with comprehensible information, as this will hinder an informed debate on the allocation and use of public resources. Critics of democracy throughout history have been pessimistic about the abilities of laypeople to grasp the complex realities of society, arguing in favour of meritocracy and rule by experts. From this perspective, our result – namely, that local politicians’ perceptions of the financial conditions of their municipality are closely related to financial indicators – provides optimism and strengthens arguments for the defenders of democratic government.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/03003930.2022.2158184.

Additional information

Funding

The work was supported by the Vetenskapsrådet [2017-02169].

Notes

1. One municipality, the island of Gotland, was included in the survey but excluded in these analyses due to its extraordinary status as both a municipality and a region. The results are thus based on 289 municipalities. See the Appendix for further explanation.

2. A multi-level regression analysis (null model) reveals that 34% of the variation in responses is at the municipal level and 66% is at the individual level, which is an exceptionally high proportion at the municipal level compared with other questions in the survey. Perception of the financial condition is thus something on which local politicians agree more than they do on most other matters. That said, there is still considerable variation within municipalities at the individual level. This variation could not be explained by financial indicators and may instead be explained by the divergent characteristics of members within a council, at the individual level. However, the individual level is not within the scope of the analysis in this paper.