How and why do political parties differ in their spending patterns? An empirical analysis of expenditure on party payroll staff

Abstract

Despite the pivotal role of payroll staff within political parties’ central offices, research on the staff expenses of parties remains scarce. In this article, we study the relative staff expenses of political parties, that is staff costs as a percentage of total annual expenses. We analyse which factors explain the differences between parties’ relative staff expenses, based on a dataset of 590 individual observations representing 52 parties from seven European countries. The multivariate model shows that relative staff expenses are higher among left-oriented parties and that they increase with party age, party membership figures and the number of years a party has been in government, while they decrease with party income. Relative staff expenses also decrease with the effective number of parties in the party system, and they are lower in election years.

Keywords

Party staff staff expenses party organisation professionalisation generalised linear mixed models

Introduction

There is a growing consensus among scholars that staff constitutes one of the most important resources for modern political parties. As parties have turned into professional organisations, they often rely heavily on a large and permanent workforce. These staff members not only support the party in its basic organisational functions, such as administration and intra-party coordination, but also help to guarantee the party’s organisational survival. In sharp contrast to the consensus on the importance of party staff, however, it remains a remarkably under-researched subject.

Thus, despite its importance, we know hardly anything about the personnel of political parties. This general absence of research on party staff can mainly be attributed to a lack of data: such studies are highly dependent on the willingness of parties to share information on their employees, yet they appear to be reluctant to provide such figures (Webb and Keith, 2017). As a result, staff data are generally highly fragmented and only partially available – for only a specific group of parties or over a certain time period. We consequently hardly know how staff levels differ between parties and how these differences can be explained.

In this article, we propose an alternative approach. Instead of analysing the number of party employees, we rely on more robust data, namely the level of relative staff expenses, that is, staff expenses expressed as a share of total annual expenses. These data are retrieved from the parties’ official annual financial statements, which have become widely available in recent decades. We particularly analyse the determinants of parties’ relative staff expenses based on financial information from 52 parties in seven European countries.

The first part of this article introduces the topics of party staff and staff expenses. The subsequent part offers an overview of the central variables and hypotheses of this study. After a discussion of our case selection and methodology, we focus on data and operationalisation. Next, the results of the analyses are presented. Finally, the discussion puts the main findings into broader perspective.

Political party staff and staff expenses

Political parties need resources, most notably money, members and staff, to maintain their organisational structure and to achieve their electoral and policy goals. Without financial and human resources, a party would be ‘an empty shell, a hollow device lacking the means of political action’ (Webb and Keith, 2017). When it comes to the relative importance of these resources, however, we find that parties have changed considerably during the past century, gradually evolving towards more professional and capital-intensive organisations.

The traditional mass parties were described as mass membership organisations, in the sense that they were equipped and financed by their members. In the second half of the 20th century, however, a decline in party membership emerged, as electorates became more volatile and partisan dealignment occurred (Kölln, 2014). This not only implied a drop of membership fees, but also a decrease of human resources. Since members served as voluntary workers, parties lost an important share of their human capital, compelling them to hire paid professionals instead (Krouwel, 2012). At the same time, the professionalisation of political communication, with the development of modern campaigning and marketing techniques, also required parties to attract additional resources and professional skills (Farrell and Webb, 2000).

Following these evolutions, parties gradually transformed into professional organisations, as described by the evolution towards catch-all, electoral-professional and cartel parties (Krouwel, 2012). They no longer relied on a strong membership base and voluntary activism, but instead became professional, capital-intensive and state-funded organisations aided by paid and skilled employees (Kölln, 2014; Mair et al., 2004; Tan, 2000). Parties thus increasingly hired staff as one of their main resources to support their professionalised organisation.

As a result of this trend, party staff are now more important than ever before. Parties generally rely on three types of staff workers (Webb and Kolodny, 2006; also see Farrell and Webb, 2000; Schutz, 1964). The first category includes the personnel within the party in public office; this is state-paid staff, such as parliamentary party workers and ministerial advisors. Second, professional consultants are hired by parties on a temporary basis for specific services, for instance marketing, accounting and legal advice. Finally, the party payroll staff refers to the personnel employed and paid by political parties, primarily active within the party in central office. These extra-parliamentary staff members constitute a party’s full-time and permanent workforce.

In recent decades, the relative importance of these three types of staff has been altered, with professional consultants and public office staff gaining importance. Nevertheless, the party payroll staff still have a pivotal position within the parties’ central offices. While political consultants have become highly relevant in the candidate-centred context of US politics, European parties have tended to gather sufficient in-house expertise in the form of permanent payroll workers (Farrell et al., 2001; Karlsen, 2010). Next, even though the party in public office has gained an ascendancy over the central office (Katz and Mair, 2002), it is the party payroll staff that guarantee the continuity of the party organisation; as these staff members live from their work for the party, their main goal is to preserve the survival of the party as an organisation (Appleton and Ward, 1997). In addition, the central office staff fulfils substantial functions for the party as a whole, including administration, finance and personnel management, intra-party coordination and communication (Webb and Kolodny, 2006).

In sum, central office payroll staff continue to play an important role within political parties. This is also reflected in the finding that the number of staff members on parties’ payrolls increased substantially over the second half of the 20th century in almost all political parties (e.g. Farrell and Webb, 2000; Krouwel, 2012; Mair, 1994). Given that ‘increasing central party staff is one aspect of the professionalization of the party’ (Kölln, 2014), figures on extra-parliamentary party staff are often applied in scholarly research as a proxy for the level of professionalisation of a party’s organisation or a party’s organisational strength (e.g. Krouwel, 2012; Scarrow and Webb, 2017; Tan, 2000).

However, this particular field of study suffers from a lack of empirical data, with staff figures often being only partially available. We therefore suggest, partly in line with Nassmacher (2009: 177–183), to apply an alternative indicator for a party’s organisational strength or professionalisation, namely staff expenses as a share of total annual expenses. This allows us to rely on hard data that are widely available, as an increasing number of countries have now introduced transparency regulations requiring parties to publicly release their financial reports.

Despite the pivotal role of party staff and the available financial data, research on parties’ staff expenses remains scarce. Nevertheless, we already know that staff costs represent one of the most important sources of expenditure for political parties. Nassmacher (2009), for instance, has shown that parties on average spend ‘more than twenty and less than fifty percent’ of their total expenses on personnel costs, a finding that is largely confirmed on the basis of data from a range of other countries (e.g. Blumenberg, 2013; Cordes, 2002; Gunzinger, 2008; Moens and Smulders, 2017). Moreover, he has also found that the average share of staff costs has grown from about 25% to 35% between 1974 and 1989 (Nassmacher, 1992), at least partially illustrating the professionalisation of party organisations.

One of the more implicit findings revealed in such studies is that strong differences exist between parties regarding the share of staff expenses. Hence the central research question of this article: Which factors explain the differences between political parties’ respective staff expenses expressed as a share of total annual expenses? In order to answer this question, we analyse parties’ relative staff expenses by means of multivariate techniques. As we argue that these expenses are a valid indicator for a party’s organisational strength, our analysis consequently contributes to the existing literature not only on party staff, but also on party professionalisation.

Variables and hypotheses

The explanatory variables of our analysis are clustered into four categories: institutionalisation, resources, ideology and system-level factors. These variables will be discussed in the following sections.

Institutionalisation

The first category of explanatory variables relates to party institutionalisation. Parties that do not yet take an institutionalised position in the party system do not have firm grassroots support. Hence they will primarily devote their financial resources to attracting supporters and voters (Appleton and Ward, 1997). Such campaigning efforts already call for a certain degree of organisational structure, including staff, but most resources will arguably be spent on marketing activities in order to gain visibility. The development of a large organisation only becomes useful when electoral survival is guaranteed and parties take a consolidated position. Moreover, it is only when they are institutionalised that parties are able to commit to more long-term engagements, such as employing permanent staff. We therefore expect the share of staff expenses to increase as parties have consolidated their position within the party system.

In the context of this study, party institutionalisation is operationalised in two ways. First, we expect an effect of party age. While younger parties will primarily invest in communication, older parties can be considered as taking an institutionalised position in the party system, having a larger party infrastructure (Blumenberg, 2013). Moreover, once the organisation has been set up, it has to be continuously maintained, requiring a permanent workforce. We consequently hypothesise that relative staff expenses increase with party age. Still, party age represents an infinite value, in theory at least, in the sense that it is not bound by an upper limit, while an organisation’s workforce cannot infinitely expand. So while we expect a positive relation between age and staff costs, we nonetheless expect this relation to gradually decrease.

A second element associated with a party’s institutionalisation is its membership base. Although membership figures have been dropping for many years now and parties rely on staff to perform their basic functions, members nevertheless remain important. They still constitute a loyal core of supporters who feel attached to a party and on whom a party can rely on election day (Arter and Kestilä-Kekkonen, 2014). In other words, a party can be considered to have a more secure position within the party system to the extent that it has a larger membership base. Consequently, these parties are able to spend more financial resources on the development of the party organisation instead of campaigning. Additionally, these parties also require a professional administration, and thus staff, to organise their membership records (Blumenberg, 2013; Webb and Keith, 2017). We therefore hypothesise that the share of staff expenses increases as a party’s membership base grows larger.

Resources

The second category of explanatory variables is related to a party’s resources. Presumably, parties enjoy a certain degree of organisational security when they rely on a large amount of stable resources. This then allows them to engage in long-term commitments, among others with respect to staff.

We first expect relative staff expenses to be impacted by a party’s total annual income. Wealthy parties have considerable financial leeway and can consequently recruit a large workforce, while their less affluent counterparts lack this budgetary leeway. Their organisational complexity will likely be more modest (Farrell and Webb, 2000; Mair, 1994). It can thus be expected that wealthy parties devote more money to their organisational development, including staff. We therefore hypothesise that relative staff expenses increase with revenues. Analogously to party age, we assume that a party’s workforce will not infinitely expand. While we expect a positive relation between income and relative staff expenses, we predict this relation to gradually decrease.

Not only the size of the total income is expected to play a role, but also the size of public funding as a share of total revenues. Public funding represents a stable resource for parties: while income from membership fees and donations can fluctuate strongly from one year to the next, the amount of subsidies a party receives is more or less predictable as this figure is usually based on a straightforward formula (Nassmacher, 1992). As a result, parties arguably are able to make predictions about future revenue streams when they rely heavily on state funding, allowing them to engage in long-term commitments with respect to payroll staff. When these subsidies constitute only a limited share of the income, parties are not inclined to recruit a large staff as there are no guarantees that they will be able to keep these employees on their payrolls for a long period of time. We consequently expect relative staff expenses to increase with the share of public funding.

A last variable connected to the resources of parties, yet more indirectly, relates to government participation. Parties in power enjoy a number of benefits that opposition parties do not (Bolleyer, 2009). Ministers can for instance rely on a group of personal staff members and political advisers whose salaries are paid by the government. Although they are officially hired to work for a minister, the party in central office undoubtedly also benefits from their services. The knowledge transfer is of particular importance: while opposition parties have to employ their own personnel to gather knowledge, ministerial staff members, ‘though notionally “civil servants”, are in effect party servants’ (Webb and Kolodny, 2006). We consequently expect government participation to be a cost-reducing factor when it comes to staff. However, our expectation is that government parties initially are cautious and only scale down on payroll staff once they have been in office for a considerable period of time. We thus hypothesise that relative staff expenses decrease the longer a party remains part of government.

Ideology

Our next category concerns party ideology. Although it is sometimes argued that ideology is losing its significance since the emergence of the catch-all and cartel party types (Katz and Mair, 2009), it cannot be denied that parties still differ from one another ideologically (Krouwel, 2012). These ideological differences moreover appear to have an impact on the way in which parties are organised. Several empirical studies have demonstrated that centre-right parties are characterised by lighter and less bureaucratic organisational structures with fewer staff members, while left-oriented parties employ high numbers of staff (Enyedi and Linek, 2008; Webb and Keith, 2017; Wilson, 1998). We consequently hypothesise relative staff expenses to be higher for the highly organised left-oriented parties.

System-level factors

Finally, we expect relative staff expenses also to be affected by system-level determinants. We first assume the level of the electoral competitiveness between parties to play a role. A high level of competitiveness implies electoral uncertainty for political parties. We again argue that parties in an insecure position will first attempt to guarantee their electoral survival, while they only develop a professionalised organisational structure when they establish a solid position with clear future prospects.

We use two proxies for electoral competitiveness, namely the effective number of parties at the electoral level and electoral volatility.¹ We expect competitiveness not only to increase with the effective number of parties as parties face more direct electoral opponents, but also with volatility as parties have less certainty about their future electoral results. These two trends automatically result in higher degrees of uncertainty, making it harder for parties to commit to long-term engagements with regard to personnel. We hence hypothesise that relative staff expenses will decrease as the effective number of parties and the level of electoral volatility increase. Similar to the influence of party age, we again expect these effects to gradually decrease. Even in party systems with a high level of electoral competitiveness, parties still incur a minimum level of staff expenditure.

The next system-level factor relates to electoral cycles. Parties’ expenditure levels substantially increase during election years, mainly due to a rise in communication costs, while office costs, including staff costs, remain relatively stable (Blumenberg, 2013; Nassmacher, 2009). As a result, the share of communication costs relative to a party’s total expenditure increases during election years at the expense of office costs. We therefore expect relative staff expenses to decrease in election years. In this context, we additionally consider the impact of campaign spending limits. We do not expect this to influence relative staff expenses directly, but we assume that it will mitigate the effect of election years: the share of communication costs will arguably be lower when campaign spending limits apply. We thus expect relative staff expenses to decrease in election years, yet less strongly when campaign spending limits are in place.

To conclude, we take the general level of labour costs in a country into account, as those levels strongly differ between countries. We expect these costs to have a mechanical effect on parties’ relative staff expenses as parties will have to spend more financial resources on staff when labour costs in their country are higher. As a result, we hypothesise that the higher the labour cost levels, the higher relative staff expenses.

Case selection

We test our hypotheses on the basis of data from seven European countries between 2000 and 2014. We do not go back further in time, as the central aim of this study is not to get an insight into the long-term evolution of the parties’ staff expenses, but rather to examine their determinants for present-day political parties. The selected countries are Austria, Belgium, Denmark, Finland, Germany, Portugal and Spain.

We took three elements into account in the selection of these cases. First, we focus on only Northern, Western and Southern European countries. These countries have a more or less stable political system, while Central and Eastern European countries are generally subject to specific dynamics related to their democratisation processes and their entry into the European Union (Walecki, 2007). Second, we restrict our analysis to parliamentary democracies with proportional and mixed-member electoral systems. Even though majoritarian and plurality systems in Europe, most notably France and the United Kingdom, are not as candidate-centred as the American system, they are nevertheless characterised by a stronger focus on individual candidates and by a stronger tendency towards individual campaigning than PR systems (Karlsen and Skogerbø, 2015). As this presumably influences party organisations and central office spending patterns, these countries are not included in our dataset. Third, a crucial criterion in case selection was the availability of data, regarding both the dependent variable and the explanatory variables (see also Data and operationalisation). Only the seven countries listed above fit these criteria.

In a final step, we collected information on all the relevant parties in the selected countries. We included all parties that have been represented in the national or federal parliament or lower house for at least two parliamentary terms between 2000 and 2014. An additional condition was again the availability of data, as for some parties no information on the dependent or one or more explanatory variables was available. We thus collected data on 52 parties from seven countries with a total of 590 individual observations (see Appendix 1).

Methodology

Because we consider 52 parties in seven countries, our data have a multilevel structure. Seven countries, however, is often considered as insufficient to perform multilevel analyses (Maas and Hox, 2005). We resolve this by interpreting our data in an alternative manner. We do not consider all observations for party P in country C as a separate cluster, but instead distinguish between different parliamentary terms. Our assumption is that observations of a party in a certain parliamentary term are more alike than observations in another term, due to the party’s different position in the changed party system. We therefore consider observations for party P in country C during term T as a separate cluster, as well as the observations for, for example, term T+1 and T+2. In other words, our observations are not clustered within parties in countries, but within parties over a particular term, in their turn clustered in countries over that term. These numbers prove considerably higher: our 570 data points are clustered in 197 parliamentary terms at the party level, in their turn clustered in 28 terms at the country level. Although the number of units at the highest level is still small, this nevertheless allows us to perform a multilevel analysis.

In addition, the dependent variable is a proportion and consequently only ranges from 0 to 1, causing the assumptions of normality and homoscedasticity to be violated. Moreover, a normal multilevel model can produce predicted values outside the logical range of proportions. We address this by running a generalised linear mixed model with a logit link and a beta distribution (Ferrari and Cribari-Neto, 2004).² As our sample sizes are rather small, we apply the MBN estimator for sample size correction (Morel et al., 2003). Although this method is most suitable for our data, the disadvantage is that the parameter estimates of the model are not interpretable as response proportions.

Data and operationalisation

Before turning to the analyses, we briefly discuss our data and the operationalisation of the variables. The dependent variable is a political party’s staff expenditure expressed as a share of its total annual expenses (see Supplementary Material online). Data are retrieved from the parties’ official financial statements. These annual statements often differ substantially with regard to the level of detail and the presentation of the data, yet when it comes to expenditure, one specific item can be clearly distinguished in the reports of nearly all parties, namely staff costs. These are the costs related to salaries and wages of payroll staff, ‘including all welfare taxes, social security charges and individual benefits provided by the employer’ (Nassmacher, 2009). Although in some cases the statements explicitly differentiate between these separate items related to staff costs, mostly only one aggregate figure is presented as the total sum spent on salaries and wages in a given year. We apply this total sum and calculate its ratio vis-à-vis the total amount of expenses, including routine expenses and expenses for communication and policy research. It should additionally be noted that while we rely on national-level central office data for Austria, Denmark, Finland and Germany, thus not taking into account staff expenditure at subnational (e.g. regional or provincial) levels, only consolidated financial statements are available for Belgium, Spain and Portugal. Figures for these countries thus also include the expenses of subnational party-related entities.³

The operationalisation of the explanatory variables is shown in Table 1, together with data sources and the predicted direction of the relationship with relative staff expenses. Three variables require particular attention. For party membership, to start with, we apply a relative measure: as is commonly done, we calculate the ratio between the number of party members and the number of registered voters. Additionally, we argue that not all parties address the nation-wide electorate, as regionalist parties for instance are directed towards only one specific region or electoral district. We therefore consider the number of registered voters for each party separately by examining whether a party is active in the entire country or only in one or more districts.

Table 1.

Overview of explanatory variables, with operationalisation, data sources and expected relationship.

Variable	Operationalisation	Expected relationship
Institutionalisation
Party age	Year of observation – year of foundation (Döring and Manow, 2016)	+
Membership	Membership figure/number of registered voters (Van Haute and Paulis, 2016)	+
Resources
Party income	Index of total annual revenues (Official financial statements)	+
Share of public funding	Public subsidies/total revenues (Official financial statements)	+
Government participation	Number of consecutive years in government (Döring and Manow, 2016)	−
Ideology
Party ideology	Continuous scale from 0 (left) to 10 (right) (Bakker et al., 2015; Polk et al., 2017)	−
System-level factors
Effective number of parties	Laakso and Taagepera measure (Gallagher, 2017)	−
Electoral volatility	Pedersen index (Emanuele, 2015)	−
Electoral cycle	Dummy variable: 1 = election year; 0 = off year (Döring and Manow, 2016)	−
Campaign spending limit	Dummy variable: 1 = limit; 0 = no limit (International IDEA, 2017)	0
Labour costs	Labour cost levels (Eurostat, 2015)	+

Second, party income should be expressed in a relative manner to control for cross-country and cross-time differences regarding size of the electorate and of the national economy. In the context of party spending, Nassmacher (2009) developed an index to control for these elements. This index can also be applied with respect to party income. The numerator of the index is composed of the relative income, that is, party income divided by the number of registered voters a party addresses. The denominator consists of the per capita GDP divided by 2000, assuming that a year is composed of 50 working weeks of 40 working hours each. As such, Nassmacher states that the denominator represents the hourly rate of national income per average person, allowing to control for inflation and economic growth.

Third, the dummy variable related to the electoral cycle takes into account only national or federal lower house elections. We exclude regional or provincial elections as including these would make the analysis unnecessarily complex. This is especially the case for countries where the various regional or provincial elections are not organised concurrently.

To conclude, Table 1 shows that we rely on existing datasets for information on the explanatory variables. While the advantage is that we can make use of reliable and readily available data previously collected by specialised scholars or institutions, these data were not gathered with the aim of this particular study. Hence we should adopt a critical attitude towards them. In particular, these data are sometimes incomplete or not available for all countries, parties or years. Where possible, we used linear interpolation to calculate missing values. If data were missing in a systematic manner, however, we dropped the corresponding observations or parties.

Results

This section analyses the parties’ relative staff expenses. The overall mean of parties’ relative staff expenses in our dataset is 26.3% (Table 2). However, the mean share of staff expenses strongly differs between countries. While parties spend up to 30% of their expenses on staff in Germany and Spain on average, for Austria, Portugal and Finland this figure is only around 22.5%. Denmark is situated around the grand mean. The mean value for Belgium (23.6%) is not exceptional, yet this country contains both the lowest and the highest value of relative staff expenses in the entire dataset.

Table 2.

Descriptive data on relative staff expenses per country.

Country	N	Mean	SD	Minimum	Maximum
Spain	120	31.3	9.7	9.3	52.7
Germany	90	29.6	10.6	10.0	50.4
Denmark	101	26.2	10.6	6.5	50.6
Belgium	156	23.6	12.2	0.1	67.9
Finland	39	22.6	9.6	8.4	43.5
Portugal	32	22.5	11.0	2.4	39.1
Austria	52	22.1	12.8	3.2	47.9
Total	590	26.3	11.5	0.1	67.9

In general, these averages are in line with Nassmacher’s finding (2009) that staff costs account for 20% to 50% of total annual expenditure. As is illustrated in Figure 1, 68.3% of all observations in the dataset is indeed characterised by relative staff expenses between 20% and 50%, while staff expenditure accounts for more than half of total annual expenses only in a marginal number of cases (0.7%). On the other hand, we see that relative staff expenses are below 20% in almost one-third of the cases in our dataset (31.0%). We should of course keep in mind that we gathered data on a specific set of countries and parties; Nassmacher’s study was based on different cases and, for instance, included Anglo-Saxon countries.

Figure 1.

Distribution of the dependent variable.

It goes without saying that these values conceal considerable differences between parties and over time. The scope of this article does not allow us to discuss each party separately, but we nevertheless present in Figure 2 the evolution of average relative staff expenses in all countries but Finland.⁴ We only average the parties for which data were available for the full time period presented in the graphs. By having a fixed number of parties we avoid possible artificial fluctuations related to the appearance and disintegration of often smaller parties.

Figure 2.

Mean share of staff expenses per year and per country.

Two conclusions can be drawn from Figure 2. First, the graphs illustrate that the share of staff expenses has remained fairly stable over the past 10 to 15 years. Although this does not rule out the possibility that staff costs have increased in absolute numbers during this period, in relative terms staff expenses have not gained importance compared with other expenditures. In other words, the importance of payroll staff in the parties’ central offices has not increased over the last decade. While previous findings demonstrated that the increase of party professionalisation took place in the second half of the 20th century (Farrell and Webb, 2000; Mair, 1994), our data suggest that this trend did not continue during recent decades. Second, relative staff expenditure levels are lower in election years, with the exception of Belgium. This finding is in line with our hypothesis that the relative staff expenses are lower in election years due to higher communication expenses. Relative staff expenses during election years average 22.1% when examining all 590 observations, while this figure increases to 27.8% in ‘off years’. A t-test reveals this difference to be significant (p<0.001).

Having explored the parties’ relative staff expenses, we now turn to the multivariate model to examine which factors influence these expenses (Table 3). Contrary to simple multilevel models, no clear measures exist for the present generalised linear mixed model to calculate the intraclass correlation (ICC) or the explained variance.⁵ In order to assess the model fit, we present minus twice the residual log pseudo-likelihood (being 892.03 for the unconditional model). As we expect the effects of party age, party income, the effective number of parties and electoral volatility to gradually decrease, we include a quadratic term for these variables in the model. To test whether campaign spending limits mitigate the effect of the electoral cycle, we include the interaction effect between both variables.

Table 3.

Result of the generalised linear mixed model, with the relative staff expenses as dependent variable.

Fixed effects (standardised coefficients)	Estimate	SE
Intercept	−1.1619***	0.052
Institutionalisation
Party age	20.3842**	6.851
Party age ²	−16.5993*	6.502
Membership	3.2344^†	1.658
Resources
Party income	−8.4233**	2.921
Party income ²	3.0891	2.118
Share of public funding	−0.5416	1.101
Government participation	2.1528*	1.058
Ideology
Party ideology	−7.1112***	1.499
System-level trends
Effective number of parties	−35.2118*	15.211
Effective number of parties ²	33.2420*	14.194
Electoral volatility	8.2427	8.478
Electoral volatility ²	−9.6696	8.499
Election year	−4.4873***	0.668
Campaign spending limits	−0.6205	1.415
Election * spending limits	2.0467**	0.620
Labour costs	1.8251	2.269
Variance components	Estimate	SE
Intercept (level-2)	0.3048	0.048
Intercept (level-3)	0.0078	0.024
–2 res log pseudo-likelihood:	769.69
N observations:	590
N level-2 groups:	197
N level-3 groups:	28

†

p<0.10.

p<0.05.

p<0.01.

***

p<0.001.

SE: standard error.

Our hypothesis regarding the effect of party age is confirmed: the older a political party, the higher the proportion of staff expenses in its total annual expenditure. The quadratic term for age is significant as well, illustrating the non-linear effect of the variable: the positive effect between age and staff costs gradually decreases. The effect of membership is only significant at the 0.10 level, albeit in the expected direction: the larger a party’s membership base, the higher that party’s relative staff expenses. These results illustrate the effect of a party’s institutionalisation: the more secure the position of a party in the party system, the more it will spend on personnel.

With respect to resources, the results are not in line with our expectations. We hypothesised party income to have a positive effect on relative staff expenses, yet the model shows the opposite: the higher a party’s revenue, the less it will spend on personnel. This effect does not decrease, as the parameter of the quadratic term is not significant. This finding does not rule out the possibility that a party will spend more on staff in absolute terms as revenues increase, but it does illustrate that other expenses will become more important. Higher revenues, for instance, allow a party to engage in permanent campaigning, expanding the share of communication costs. Second, the effect of the share of public funding is not significant. Access to this type of reliable income apparently does not cause parties to spend more money on staff. Finally, the duration of government participation does have an impact on staff expenses, but again contrary to our expectations: the longer a party has been in office, the higher its relative staff expenses. We suspect that we captured another characteristic of parties by this variable, namely party institutionalisation. Analogously to age and membership, we assume that as parties are longer in office, they become more consolidated and are more inclined to develop a large party infrastructure, hence the positive effect.

One of the most significant effects relates to party ideology. The hypothesis regarding this variable is confirmed: relative staff expenses are lower for right-oriented parties, these parties generally being characterised by lighter and less bureaucratic organisational structures. Stated otherwise, the highly organised left-oriented parties spend more on staff.

As concerns the system-level variables, the model shows that the effective number of parties plays a role in line with our expectations. The parameter estimate is negative, illustrating that the higher the number of parties, the lower the share of staff expenses. Moreover, the effect of the quadratic term is significant, indicating that the negative effect gradually decreases. The effect of electoral cycles is also convincingly confirmed by our findings: in relative terms, parties spend significantly less money on staff in election years than in ‘off years’. Yet, this negative effect is mitigated when campaign spending limits apply, given the significant positive interaction effect. Two variables do not show a significant effect. First, electoral volatility does not yield the expected effect. The level of competitiveness between parties is probably better captured by the effective number of parties as this is immediately visible for the parties, while electoral volatility represents a less visible characteristic of a party system. The expected positive effect of labour costs also fails to materialise. This might be the result of the fact that if labour costs are high, other goods and services are also more expensive. As a result, increased staff expenses for parties may be levelled out by an increase of other costs.

Discussion

In this article, we investigated which factors explain the differences in relative staff expenses between parties, based on data from 52 parties in seven European countries. Staff expenditure on average represented 26.3% of the parties’ total annual expenses. We found that the level of relative staff expenses remained largely stable over the past 10 to 15 years. As we argued that relative staff expenses can serve as an alternative indicator for a party’s organisational strength, this finding suggests that parties have not professionalised their organisation any further during the past decade.

Next, we conducted a multivariate analysis to explain parties’ relative staff expense levels, or alternatively their level of professionalisation. We found that older and leftist parties with a large membership base, operating in an electoral system with few parties, tend to have high levels of staff expenditure, thus being more professionalised. On the other hand, younger and right-oriented parties with few members, active in a system with a high number of competing parties, are generally characterised by low levels of staff expenditure, indicating only limited organisational strength. Based on these findings, we can distinguish, following Blumenberg (2013), between apparatus-oriented parties and election-oriented parties respectively. While the first category mainly spends money on the development and maintenance of the party organisation, including staff, the second type is characterised by high expenditure levels on communication and marketing.

Following Scarrow and Webb (2017), three implications of a party’s apparatus-oriented versus election-oriented character are worth mentioning. First, the organisational strength of the central office can influence the frequency of party splits: disloyal members may lose considerable resources if they establish their own party, leading to a competitive disadvantage. Next, while election-oriented parties generally apply a short-term strategy aimed at immediate electoral success, apparatus-oriented parties arguably value long-term strategies directed towards organisational survival. Finally, whether a party relies on a professional workforce instead of on volunteers can influence intra-party dynamics, among others with respect to internal decision-making processes. In line with Michels (1915), it could for instance be suggested that the more professionalised organisations are, the less they are democratically structured internally.

To conclude, our analysis of parties’ relative staff expenses has revealed a number of interesting results that pave the way for a distinction between party types on the basis of their expenses. Party types have traditionally been based on, for instance, ideology, organisational structure or income streams (Krouwel, 2012), but the present study suggests that it might also be possible to distinguish between types based on spending patterns. Such a typology would ideally be based on a more fine-grained distinction between the various cost items in parties’ financial statements. Unfortunately, this type of data is difficult to obtain, in particular because the level of detail found in party accounts greatly differs between countries and parties. This is one of the reasons why we settled for the rather simple binary distinction between staff costs and other costs for the purpose of this study. The good news is that this variable, however crude, proved to relate to relevant explanatory variables in a clear and largely predictable manner. Using this variable to study the spending patterns of parties and to construct party typologies appears to offer a promising path for future research as party finance data will increasingly become publicly available.

Footnotes

Appendix 1. Data

Table 4.

Overview of countries, parties and time period covered in the dataset.

Austria	2003–2014 BZÖ – FPÖ – GRÜNE – ÖVP – SPÖ
Belgium	2000–2014 Agalev/Groen – CVP/CD&V – Ecolo – LDD – PRL/MR – PS – PSC/cdH –SP/sp.a – VB – VLD/Open Vld – VU/N-VA
Denmark	2001–2014 A – B – C – D – F – I – O – Ø – V
Finland	2009–2014 KD – Kesk. – Kok. – PS – RKP – SDP – Vas. – Vihr.
Germany	2000–2014 Bündnis 90/Die Grünen – CDU – CSU – FDP – PDS/Die Linke.PDS/Die Linke – SPD
Portugal	2003–2012 BE – PCP – PS – PSD
Spain	2000–2014 BNG – CDC – EAJ-PNV – ERC – ICV – IU – PP – PSOE – UDC

Acknowledgements

A previous version of this article was presented at the Düsseldorf Graduate Conference on Party Research (12–13 February 2016). The authors would like to thank all participants for their comments and insights. The authors would also like to express appreciation for the invaluable suggestions received from Gert-Jan Put and Pieter Moens, as well as from the anonymous reviewers of the International Political Science Review.

Funding

This work was supported by the Research Foundation – Flanders (FWO).

Supplemental material

Supplemental material is available at .

Notes

ORCID iD

Jef Smulders

Author biographies

Jef Smulders is a PhD Research Fellow of the Research Foundation – Flanders (FWO), affiliated to the KU Leuven, Public Governance Institute. His research interests include political finance, party organisation and elections. His work has been published in journals such as Government and Opposition, Local Government Studies and Election Law Journal.

Bart Maddens is a Professor of Political Science at the KU Leuven, Public Governance Institute. His research interests include political finance, elections and multi-level systems. His work has been published in journals such as West European Politics, Electoral Studies and Acta Politica.

References

Appleton

Andrew M

Ward

Daniel S

(1997) Party Response to Environmental Change: A model of organizational innovation. Party Politics 3(3): 341–362.

Arter

David

Kestilä-Kekkonen

Elina

(2014) Measuring the Extent of Party Institutionalisation: The Case of a Populist Entrepreneur Party. West European Politics 37(5): 932–956.

Bakker

Ryan

de Vries

Catherine

Edwards

Erica

Hooghe

Liesbet

Jolly

Seth

Marks

Gary

Polk

Jonathan

Rovny

Jan

Steenbergen

Marco

Vachudova

Milada

(2015) Measuring Party Positions in Europe: The Chapel Hill Expert Survey Trend File, 1999–2010. Party Politics 21(1): 143–152.

Blumenberg

Manuela S

(2013) Ausgabenstrukturen Demokratischer Parteien im Internationalen Vergleich. Hamburg: Verlag Dr. Kovač.

Bolleyer

Nicole

(2009) Inside the Cartel Party: Party Organisation in Government and Opposition. Political Studies 57(3): 559–579.

Cordes

Doris

(2002) Die Finanzierung der politischen Parteien Deutschlands, Österreichs und der Niederlande. Oldenburg: Carl von Ossietzky Universität Oldenburg.

Döring

Holger

Manow

Philip

(2016) Parliaments and Governments Database (ParlGov): Information on parties, elections and cabinets in modern democracies. Development Version. Available at: http://www.parlgov.org

Emanuele

Vincenzo

(2015) Dataset of Electoral Volatility and its Internal Components in Western Europe (1945–2015). Rome: Italian Centre for Electoral Studies. DOI: 10.7802/1112.

Enyedi

Zsolt

Linek

Lukas

(2008) Searching for the Right Organization: Ideology and Party Structure in East-Central Europe. Party Politics 14(4): 455–477.

10.

Eurostat (2015) Labour cost levels by NACE Rev. 2 activity. Available at: http://ec.europa.eu/eurostat/product?code=lc_lci_lev

11.

Farrell

David M

Webb

Paul D

(2000) Political Parties as Campaign Organizations. In Dalton

Russell J

Wattenberg

Martin P

(eds) Parties Without Partisans: Political Change in Advanced Industrial Democracies. Oxford: Oxford University Press, 102–128.

12.

Farrell

David M

Robin

Kolodny

Stephen

Medvic

(2001) Parties and Campaign Professionals in a Digital Age. Political Consultants in the United States and Their Counterparts Overseas. The International Journal of Press/Politics 6(4): 11–30.

13.

Ferrari

Silvia

Cribari-Neto

Francisco

(2004) Beta Regression for Modelling Rates and Proportions. Journal of Applied Statistics 31(7): 799–815.

14.

Gallagher

Michael

(2015) Election indices dataset. Available at: http://www.tcd.ie/Political_Science/staff/michael_gallagher/ElSystems/Docts/ElectionIndices.pdf

15.

Gunzinger

Mathieu

(2008) Analyse comparative des ressources financières des partis politiques suisses. Cahier de l’IDHEAP 240/2008. Chavannes-Lausanne: Institut de hautes études en administration publique.

16.

International IDEA (2017) Political finance database. Available at: http://www.idea.int/themes/political-finance

17.

Karlsen

Rune

(2010) Fear of the Political Consultant: Campaign Professionals and New Technology in Norwegian Electoral Politics. Party Politics 16(2): 193–214.

18.

Karlsen

Rune

Skogerbø

Eli

(2015) Candidate Campaigning in Parliamentary Systems: Individualized vs. Localized Campaigning. Party Politics 21(3): 428–439.

19.

Katz

Richard S

Peter

Mair

(2002) The Ascendancy of the Party in Public Office. In Gunther

Richard

Montero

Jose R

Linz

Juan J

(eds) Political Parties: Old Concepts and New Challenges. Oxford: Oxford University Press, 113–135.

20.

Katz

Richard S

Peter

Mair

(2009) The Cartel Party Thesis: A Restatement. Perspectives on Politics 7(4): 753–766.

21.

Kölln

Ann-Kristin

(2014) Party Decline and Response: The Effects of Membership Decline on Party Organisations in Western Europe, 1960–2010. Enschede: University of Twente.

22.

Krouwel

André

(2012) Party Transformations in European Democracies. Albany: SUNY Press.

23.

Maas

Cora JM

Hox

Joop J

(2005) Sufficient Sample Sizes for Multilevel Modeling. Methodology 1(3): 86–92.

24.

Mair

Peter

(1994) Party Organizations: From Civil Society to the State. In Katz

Richard S

Mair

Peter

(eds) How Parties Organize: Change and Adaptation in Party Organizations in Western Democracies. London: SAGE Publications, 1–22.

25.

Mair

Peter

Müller

Wolfgang C

Plasser

Fritz

(eds) (2004) Political Parties and Electoral Change: Party Responses to Electoral Markets. London: SAGE Publications.

26.

Michels

Robert

(1915) Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracy. New York: Hearst’s International Library Company.

27.

Moens

Pieter

Smulders

Jef

(2017) The human capital of political parties. An analysis of Belgian Central Party Offices. Paper presented at the Politicologenetmaal, Leiden, 1–2 June. Available at: https://lirias.kuleuven.be/handle/123456789/584438

28.

Molenberghs

Geert

Verbeke

Geert

(2005) Models for Discrete Longitudinal Data. New York: Springer.

29.

Morel

Jorge G

Bokossa

Neerchal

Nagaraj K

(2003) Small Sample Correction for the Variance of GEE Estimators. Biometrical Journal 45(4): 395–409.

30.

Nassmacher

Karl-Heinz

(1992) Parteifinanzen im Westeuropäischen Vergleich. Zeitschrift für Parlamentsfragen 92(3): 462–488.

31.

Nassmacher

Karl-Heinz

(2009) The Funding of Party Competition: Political Finance in 25 Democracies. Baden-Baden: Nomos Verlagsgesellschaft.

32.

Polk

Jonathan

Rovny

Jan

Bakker

Ryan

Edwards

Erica

Hooghe

Liesbeth

Jolly

Seth

Koedam

Jelle

Kostelka

Filip

Marks

Gary

Schumacher

Gijs

Steenbergen

Marco

Vachudova

Milada

Zilovic

Marko

(2017) Explaining the Salience of Anti-elitism and Reducing Political Corruption for Political Parties in Europe with the 2014 Chapel Hill Expert Survey Data. Research & Politics 4(1): 1–9.

33.

Scarrow

Susan E

Webb

Paul D

(2017) Investigating Party Organization: Structures, Resources, and Representative Strategies. In Scarrow

Susan E

Webb

Paul D

Poguntke

Thomas

(eds) Organizing Political Parties. Representation, Participation, and Power. Oxford: Oxford University Press, 1–27.

34.

Schutz

Charles E

(1964) Bureaucratic Party Organization Through Professional Political Staffing. Midwest Journal of Political Science 8(2): 127–142.

35.

Tan

Alexander C

(2000) Members, Organization and Performance. An Empirical Analysis of the Impact of Party Membership Size. Hampshire: Ashgate Publishing.

36.

Van Haute

Emilie

Paulis

Emilien

(2016) MAPP Dataset. Zenodo. Available at: http://doi.org/10.5281/zenodo61234.

37.

Walecki

Marcin

(2007) The Europeanization of Political Parties: Influencing the Regulations on Political Finance. EUI Working Paper MWP 2007/29. San Domenico di Fiesole: European University Institute.

38.

Webb

Paul D

Dan

Keith

(2017) Assessing the Strength of Party Organizational Resources: A Survey of the Evidence from the Political Party Database. In Scarrow

Susan E

Webb

Paul D

Poguntke

Thomas

(eds) Organizing Political Parties: Representation, Participation, and Power. Oxford: Oxford University Press, 31–61.

39.

Webb

Paul D

Robin

Kolodny

(2006) Professional Staff in Political Parties. In Katz

Richard S

Crotty

William

(eds) Handbook of Party Politics. London: SAGE Publications, 337–347.

40.

Wilson

Frank L

(1998) The Center-Right at the End of the Century. In Wilson

Frank L

(ed.) The European Center-Right at the End of the Twentieth Century. London: Macmillan Press, 247–270.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.05 MB