Are Taxes or User-Fees More Popular Among Politicians? The Case of Childcare

1 Introduction

How shall publicly provided excludable goods be financed – by general taxation, user fees (uniform or income-dependent) or a combination of both? And how do politicians decide which of these modes of financing they choose for which type of service? Prominent applications are in education, notably tertiary (universities) and early childcare.

The literature has hitherto concentrated on three types of issues: a) efficiency: under quite general assumptions, user fees are more efficient because of incentives on the demand side (Economides et al. 2017; Economides and Philippopoulos 2020); b) equity: Economides and Philippopoulos (2012) show that user-fees are not necessarily regressive; c) political economy: Swope and Janeba (2005) show that the distribution of preferences is decisive for the choice of instruments.

The general conclusion of this literature is that even accounting for positive externalities as in Economides et al. (2017), exclusive tax financing is neither efficient nor desirable under widely shared distributive goals. Yet, in the field of education in many countries exclusive tax financing prevails. Moreover, modest user fees, which were designed to cover a small share of the costs of university education, were introduced in Germany in seven states with conservative governments in the period 2007–2012 but abolished everywhere by mostly leftist state parliaments thereafter (Kauder and Potrafke 2013).

In this paper we shall focus on another, even more striking example, namely childcare: While tuition fees in universities are typically uniform, childcare fees can be (and are often) made dependent on parents’ income so that efficiency and equity goals do not have to be in conflict in this area.

In a recent paper, Koll et al. (2024) show in a dynamic structural model with heterogeneous households that the marginal efficiency costs of redistribution within the childcare subsidy programme are lower than those of the income tax schedule. Therefore, it is efficient to shift some share of the desired redistribution from the income tax system to the system of childcare financing. This result is based on the findings by Havnes and Mogstad (2015), Cornelissen et al. (2018), and García et al. (2020) that income-dependent fees (compared to flat fees) reduce the attendance of high-income children and increase the attendance of low-income children, who benefit more from attending childcare facilities in their development so that total human capital formation is increased.

It is sometimes argued that parental fees may still be an obstacle to the access of low-income children to public childcare (Abrassart and Bonoli 2015) so that the possible positive externalities cannot be reaped, but their analysis refers to Switzerland, a country in which 85 per cent of total costs are covered by parental fees, and is thus not relevant for the German case where the relation between taxes and fees is roughly the opposite (see Section 3).

Moreover, as Huebener et al. (2020) show, the abolition of parental fees for the last year before school starting age in some states increased the labor supply of mothers only by a small margin and was thus an inefficient way to reach this goal.

Given these rather clear arguments in favor of some (income-related) user fees for childcare, it is surprising to notice that some German states with leftist governments have abolished the existing (and rather modest) parental fees and replaced them with pure tax financing, even without raising additional (and perhaps progressive) taxes.^[1] In other states,^[2] leftist parties campaigned in recent elections with the promise to follow suit.

It is the purpose of this research to first, investigate the attitudes of politicians towards parental fees for childcare, including the reasons given by them for supporting or, for that matter, opposing parental fees for childcare, and second, to examine how the attitude towards these fees depends upon the politician’s ideological position and his/her demographic characteristics. To answer our research questions, we employ face-to-face online interviews with an embedded survey with members of eight federal state legislatures. Our final sample consists of 535 legislators across the party spectrum and is largely representative in terms of age, gender and political experience.

We find that our respondents are rather equally divided on the statewide abolition of childcare fees. Among the respondents who support fees, a clear majority favors income-dependent fees. Left-leaning politicians tend to be against childcare fees and indicate as a justification mainly that “education must be free for all”, although they admit that an abolition would help mainly the rich. Right-leaning politicians tend to support fees and consider an abolition as helping mainly the rich and a problem for the quality of childcare.

This paper is structured as follows: Section 2 summarizes previous literature on the topic, and Section 3 provides some background information on the system of childcare financing in Germany. In Section 4 we derive the theoretical hypotheses to be tested. In Section 5 we describe the survey as a whole and the questions relating to childcare fees. Section 6 is devoted to an analysis of the responses, and Section 7 to a discussion of the main findings. Finally, Section 8 concludes.

2 Previous Literature

To our knowledge, this is the first study which examines politicians’ preferences on childcare financing. The literature which comes closest to our topic comprises studies on citizens’ preferences on the same matter.

The first paper in this literature is Biel et al. (1997). In a survey of 1840 citizens of five Swedish municipalities they asked for their opinion on the fairness of the allocation of childcare places and of the mode of financing. The most interesting result related to our topic is summarized in the following statement (p. 76): “Preference for the Conservative political party increased willingness to pay by fees and decreased willingness to pay by taxes, whereas preference for the socialist or liberal political parties increased willingness to pay by taxes and decreased willingness to pay by fees.”

Neimanns and Busemeyer (2021) use data from the INVEDUC survey conducted in 2014 (see, Busemeyer et al. 2018), in which 8,905 individuals in 8 European countries were asked for their preferences with respect to public provision of childcare services and parental fees. The results for Germany, which the authors characterize in 2014 as a country with low availability of childcare places and income-dependent fees, show that low-income groups support more public provision, whereas middle and high-income groups have strong preferences for lowering fees.

An even more closely related study is the one by Busemeyer and Goerres (2020), which is, however, geographically restricted to citizens of one particular town, Konstanz, who were invited in 2014 to participate in the following vignette exercise: Each vignette consisted of a description of a family with a two-year old child who was supposed to be attending childcare in Konstanz and which had a number of characteristics which could attain different values (family income, employment status of parents, religion and a few others). It also contained a monthly fee level for childcare. The respondents were then asked to rate the fee level for this particular family on a scale from −5 (unfairly too low) to +5 (unfairly too high). The more than 6.000 ratings of the 1,255 participants were then analyzed by a regression model to determine the impact of the various given family characteristics on the fee level which is considered “fair” by the average participant. The two most striking results are first that the mean “fair” fee level is 191€ per month, which is only slightly lower than the actual (uniform) fee prevailing in 2014, and secondly, that the “fair” fee is income dependent with a rather steep slope: 1,000€ additional monthly income leads to a fee increase by 53€. Even more surprisingly, the “fair” fee level increases with respondents’ own income.^[3]

In a survey among German parents of children in childcare facilities, Camehl et al. (2015) asked for the satisfaction with various aspects of these facilities as well as for their willingness to pay for childcare. It turned out that overall satisfaction with the quality of care was high, that low-income parents were not quite as satisfied with the costs of care and that high-income parents would have been willing to pay more than they actually did.

Given the observed widespread agreement among voters that parents should pay for childcare, it seems puzzling that state parliaments with leftist majorities in fact abolished them in recent years. To shed light on this puzzle, Neimanns (2022) examined the INVEDUC dataset mentioned above to examine the relationship between preferences for public childcare and vote intentions stated by the survey participants. His results show that preferences for increasing public childcare spending are positively connected to electoral support for the left and that this correlation is stronger in high-income groups. Thus it pays for leftist parties to cater to the interests of these groups by reducing or even abolishing income-dependent parental fees. This insight aligns with the broader trend of left parties increasingly courting middle-class voters, often at the expense of their lower-income constituencies (see e.g. Gingrich and Häusermann 2015).

3 Some Facts on Childcare in Germany

The utilization of public childcare facilities in Germany has greatly increased in the past decade, especially since the legal claim to a place in such an institution was extended in 2013 from the age range 3–6 years to all children 12 months or older. In the year 2021, 34 per cent of all children below the age of 3 and 92 per cent in the age range 3–6 years attended these facilities (Autorengruppe Bildungsberichterstattung 2022), and public childcare expenditures increased in the period 2010 to 2021 from 14.7 to 40.3 bill. Euro, i.e. by 174 per cent (Statistisches Bundesamt 2022), while parental fees covered only between 5 and 25 per cent of total costs, depending on the federal state (Dohmen 2016, p. 10).

The procedures by which the structure and level of parental fees are determined are rather complicated and vary quite substantially among the 16 federal states (Geis-Thöne 2024, p. 8). In four of these states, the fee schedule is determined by the service provider, while in most other cases it is set by the municipality. However, many states have passed state laws that enforce free childcare either for all children or for certain age groups or for a number of hours per day. As a consequence, in recent years the state and the municipality each provide approximately one-half of the public subsidies (Statistisches Bundesamt 2022, p. 49).

In 2018 a federal law was passed that stipulated that parental fees have to be differentiated, but parents’ income was mentioned only as one of several possible criteria besides the number of siblings in childcare and the duration of attendance. However, as was valid before, recipients of social assistance and similar transfers have to be exempt from fees. Moreover, the law determined that the federal government would grant 5.5 bill. Euros over the period 2019–2022 to the 16 states, which the latter could use either to enhance the quality of childcare facilities or to reduce parental fees (even to zero). As a consequence, Berlin and Mecklenburg-Vorpommern completely abolished user fees and Sarre decided to abolish them by 2027,^[4] while the majority of the other states offer free childcare for the last (1, 2, 3 or 4) years before school starting age, and Hamburg and Hesse offer free childcare for all children for up to 5 or 6 h a day, respectively (Geis-Thöne 2024, p. 9).

Despite these additional funds and in contrast to the legal entitlement, places in childcare facilities are still scarce: According to the Bertelsmann Foundation, in 2023 there was a lack of 430,000 places.^[5] The large gaps between demand and supply of these places for different age groups of children are documented in detail by Huebener et al. (2023).

The regulations with respect to the income-dependence of parental fees are even more complicated. In some states, notably Hesse, fees are usually uniform except for subsidies to low-income families, whereas in other states (such as Bavaria and Baden-Württemberg), some municipalities have uniform fees and in others these depend on parental income (ibid., pp. 20ff.). In other states such as Hamburg, free childcare for the first 5 h a day is combined with income-dependent fees for the additional hours.

Where they (still) exist, parental fees can be quite sizable: According to a survey in 2019 (Hubert et al. 2021, p. 25) median fees for below-3 year old children in the states of Northrhine-Westphalia and Sarre amounted to 400 Euros, while in six other states they were at 300 Euros or more. However, childcare costs are to a large extent (i.e. two-thirds up to a maximum of 4,000 Euros per child per year) deductible from the income tax.

Given the institutional structure of decision-making, it is justified to conclude that the crucial decisions on the existence and scope of parental fees are mostly taken on the state level. Thus it seems appropriate to focus on state representatives when choosing which type of politicians to include in our survey study.

4 Deriving Theoretical Hypotheses

What would economic/political theory predict about the (stated) attitudes of politicians with respect to the use and structure of childcare fees? The answer to this question depends on the degree to which politicians view themselves as advocates of the “common good”, which can be identified with efficiency and/or equity goals, or as maximizing their own or their party’s utility, which means that they support policies that are popular with the electorate. Within both the “common good” framework and the vote-maximizer framework, some goals and strategies hold irrespective of political ideology, while others differ by political ideology and party affiliation. In the following, we provide an overview of the considerations in favor or against childcare fees and derive hypotheses based on the respective dominance of considerations in favor or against for right- and left-leaning politicians.

We start with motives that are independent of political ideology. Starting with the goal of efficiency of the tax-transfer system, it seems justified to conclude that parental fees can be used as a means of containing distortionary income taxes. Efficiency goals should in principle be universally shared.^[6] On the other hand, vote-maximizing behavior may suggest promoting policies that are characterized by concentrated benefits for specific groups, while the costs are dispersed over the universe of tax-payers who may not even be aware of the costs (Weingast et al. 1981). The abolition of parental fees falls into this category, because the beneficiaries (parents of children in the age range 1–6 years) are a relatively small and well-defined group.

We now turn to motives that discriminate by political ideology. There are several reasons to expect differences in support for parental fees between more right-leaning and more left-leaning politicians.

First, the relative importance of efficiency goals, when compared with equity goals should be larger for more conservative politicians. Another reason to expect support for parental fees from conservative politicians is adherence to the “benefit principle” of taxation, which is popular among conservative and liberal economists: if the state provides private goods (or public goods which benefit some parts of the population more than others), then the taxes or fees levied to finance these services should be proportional to these benefits.^[7] In the case of childcare, fees that are roughly proportional to parents’ wages are in line with the benefit principle because they reflect the opportunity costs of caring for their children at home.

In contrast, more left-leaning politicians often subscribe to the goal of reducing overall economic inequality in society and therefore favor redistribution through the tax system, including redistribution between parents and non-parents.

Furthermore, there is an additional rationale for politicians of the left to oppose parental fees: they consider childcare facilities as an important part of the education system of a country and maintain that education is one of the key fields of government responsibility. According to this view, free and unlimited access, potentially even with mandatory enrolment at least for the last year before school starting age, is more important than income redistribution using fees as an instrument, in particular given that the revenue from these fees covers only a small part of costs anyway. Furthermore, it would be hard to explain why the attendance to childcare should come with sizable costs to the parents, while the immediately following primary school education is completely free of charge. This difference in fees could be justified only if public childcare facilities were not fully considered as part of the education system, but their function of caring for the children was emphasized. This, e.g., may be the rationale of providing a limited number of hours per day (“the education time”) free of charge while levying fees for the additional “caring time”, as is the case in the state of Hamburg.

Moreover, if income-dependent fees constituted a sizeable share of the revenues of childcare facilities, there would be the danger that facilities in high-income neighborhoods could provide higher quality than those in poor areas so that – similar to school districts in the U.S. – the neighborhood would have an undesirably strong impact on child development.

While all of these motives of left-wing politicians fall into the category of following the ”common good”, based on the results in Neimanns (2022), the abolition of income-dependent parental fees may also be advocated by leftist politicians because they are most likely to attract the votes of young parents with above-average income.

Based on these considerations, we formulate the following hypotheses:

Based on hypotheses H1 and H2, we formulate the following hypothesis for the direct comparison between left- and right-leaning politicians:

Among the left-leaning politicians who do support parental fees, we would expect that they support a fee structure that is beneficial for lower-income parents. We thus formulate the following hypothesis:^[8]

For a similar reason as mentioned with respect to H2, left-leaning politicians might support state-wide regulation of fees because if every municipality could determine their own fee schedule, more affluent communities could afford higher quality and/or lower parental fees, which would contradict equity concerns. This is expressed in

5 Methods and Data

6 Results

6.1 General Attitudes towards Childcare Fees

As a first step, we present the attitudes towards the abolition of childcare fees, income-dependence of the fees and their mode of regulation for the full sample. All descriptive analyses are based on a weighted sample, with weights adjusted for party affiliation, to ensure representativeness. Tables and Figures for the unweighted sample are provided in Figure B1 and Table B1 in the Appendix.^[14]

We first turn to the crucial question of support for or rejection of parental fees altogether. Figure 1a shows that a narrow majority of 50:45 per cent pleaded against statewide abolition of childcare fees. We next turn to the support for income-dependent fees. Figure 1b shows that 60 per cent of respondents who indicated that they supported fees favored income-dependent fees.^[15] Around 23 per cent think that only recipients of social assistance should be exempt. Only 16 per cent voted for the same fee level irrespective of parents’ income. With respect to the mode of regulation, about two thirds of all participants voted for statewide uniform fees and one third favored decision making by the municipality (see Figure 1c).

Figure 1:

Attitudes towards childcare fees. (a) Fee abolition. (b) Income dependence. (c) Regulation. Notes:The left panel shows the answers to the question “Should childcare fees be abolished statewide?” The middle panel shows answers to the question “Should childcare fees depend on parents’ income?” The right panel shows the answers to the question “How should childcare fees be regulated in your federal state?” Percentages are based on a weighted sample, with weights adjusted for party affiliation, to ensure representativeness.

How do these attitudes towards childcare fees differ according to the left-right political orientation of the respondents? On this matter, Table 2 comprises descriptives with respect to the political orientation, where we make the split at the median of this variable in our sample. It turns out that leftist politicians favor the abolition of fees by a margin of over 2:1 (Chi-squared test, p < 0.001), while among their right-wing colleagues, the split is almost exactly opposite (Chi-squared test, p < 0.001). These results clearly confirms hypotheses H1, H2 and H2*.

Table 2:

Descriptives by political orientation.

Pol. Orientation	Abolition		Income dependence			Regulation
	No	Yes	No	Exem.	Yes	Uni.	Diff.
Left	69	153	2	9	58	182	43
Right	199	87	40	53	104	153	140

1. The table shows the number of respondents that agreed with the respective statements by political orientation. Counts are based on a weighted sample, with weights adjusted for party affiliation, to ensure representativeness. The weighted counts are rescaled to match the original sample size. Respondents are categorized as “Left” if their political self-placement is below or equal to the median self-placement of 4 on a scale from 0: left to 10: right.

Moreover, fee supporters who place themselves on the right of the median favor income-dependent fees by a margin of slightly over 2:1, while among those on the left of the political spectrum, nearly everyone is in favor of income dependent fees (Chi-squared test, p < 0.001), which is in line with hypothesis H3 that among those who accept parental fees, left-leaning politicians are more likely than right-leaning politicians to support income-dependent fees rather than uniform fees. The view that only recipients of social assistance should be exempted from fees is expressed by only 13 per cent of the left-leaning fee supporters while it is twice this percentage among right-leaning fee supporters. Finally, politicians who place themselves to the left of the median are strong advocates of a uniform regulation (by a margin of more than 4:1), while more right-wing politicians split almost evenly on the question (Chi-squared test, p < 0.001). This confirms hypothesis H4.

These observations are confirmed by estimating a linear probability model on the dummy variables “abolition”, “exemption”, “income dependence” and “differing fees by municipality” using the characteristics of the respondents as regressors (Table 3). For each outcome variable, the first regression model only includes political orientation and state fixed effects as predictors. The second regression model adds controls for age, education, gender and political experience.

Table 3:

Linear regressions: attitudes towards childcare fees.

	Dependent variable:
	Abolition: Yes		Exemption: Yes		Income dep: Yes		Regulation: Diff
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Left	0.35***	0.35***	−0.15**	−0.14*	0.34***	0.32***	−0.26***	−0.27***
	(0.04)	(0.04)	(0.06)	(0.06)	(0.06)	(0.07)	(0.04)	(0.04)
Young		−0.02		0.01		−0.02		−0.01
		(0.04)		(0.06)		(0.06)		(0.04)
Tertiary		−0.13*		0.11		−0.07		0.06
		(0.06)		(0.10)		(0.11)		(0.06)
Female		−0.02		−0.02		0.08		−0.01
		(0.04)		(0.07)		(0.07)		(0.04)
Newcomer		−0.04		0.08		−0.07		−0.005
		(0.04)		(0.06)		(0.07)		(0.04)
Constant	0.16***	0.31***	0.25***	0.10	0.54***	0.64***	0.62***	0.59***
	(0.04)	(0.08)	(0.05)	(0.11)	(0.05)	(0.12)	(0.04)	(0.07)
State FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Observations	509	488	242	238	242	238	519	499
R2	0.25	0.25	0.06	0.08	0.15	0.16	0.22	0.23
Adjusted R2	0.23	0.23	0.03	0.03	0.12	0.12	0.21	0.21

1. Linear regression models. Dependent variables: Columns (1) and (2): Binary outcome indicating a preference for not abolishing the fees nationally (0) versus abolishing the fees nationally (1). Columns (3) and (4): Binary outcome indicating a preference for uniform fees or fees that depend on the parents’ income (0) versus exempting only recipients of social assistance (1), restricted to respondents in favor of fees. Columns (5) and (6): Binary outcome indicating a preference for uniform fees or exempting only recipients of social assistance (0) versus fees that depend on the parents’ income (1), restricted to respondents in favor of fees. Columns (7) and (8): Binary outcome indicating a preference for fees that are uniform on the state level (0) versus vary on the municipality level (1). Robust standard errors in parentheses, where *p < 0.05, **p < 0.01, ***p < 0.001.

Columns (1) and (2) show that left-leaning increases support for statewide abolition of fees by 35 percentage points. Columns (3) and (4) show that left-leaning decreases support for only exempting recipients of social assistance among those who support fees. Columns (5) and (6) show that left-leaning increases support for income dependence among those who support fees. Furthermore, columns (7) and (8) show that left-leaning increases support for statewide uniform fees. With respect to the other respondent characteristics, while age and gender do not seem to play a substantive role for any of the attitudes towards childcare fees, having some tertiary education decreases the likelihood to reject parental fees by 14 percentage points.

Tables C2–C5 in the Appendix show that these results are robust to additionally controlling for whether a legislator has children,^[16] and for using continuous instead of categorical variables for age and political experience. The tables also show that the results hold for different specifications of political orientation. First, also for political orientation we can use a continuous instead of a categorical variable. Second, the survey included a second question to elicit political orientation, namely whether the respondent preferred lower taxes or higher government spending. The results also hold when using the continuous measure and when we classify a politician as left-leaning (right-leaning) only when their ideological self-placement and their preference for lower taxes versus higher government spending are classified as left (right) based on a median split.

With respect to differences between states, it turns out that relative to the omitted state Baden-Württemberg, politicians in five other states (Sarre, Hessia, Thuringia, Northrhine-Westphalia and Schleswig-Holstein) are considerably stronger in favor of fee abolition, whereas those in Bavaria and Berlin do not differ significantly from Baden-Württemberg (see Table C1).^[17] The highest percentage in favor of abolition is observed in the Sarre, in which the abolition has already been decided in the state parliament.

6.2 Reasons for Support or Rejection of Fee Abolition

As described above, respondents who indicated that they are against or in favor of a fee abolition were then asked to state their reasons. Of the 267 respondents who indicated that they wanted to abolish the fees, 91.4 % (244 respondents) agreed with the argument that “Education must be accessible to all” and 32.2 % (86 respondents) stated that “Abolition strengthens distributive justice”. Another topic that was mentioned in the answers provided in the open text field was fee abolition as “support for families”. As it was mentioned by only 13 participants we will not analyze this reason further.

Examining which politicians subscribe to the frequently mentioned reasons for favoring the abolition of childcare fees (see Figure 2a and Table D1), the only significant result is that left-leaning persons have a stronger propensity (by 15 percentage points) than all others to accept the principle that “education should be accessible to all”. Other characteristics of the respondents such as age, gender and education do not seem to play a role. Table D2 in the Appendix shows that when using the continuous variable for political orientation, also the argument that abolition strengthens distributive justice finds significantly more support among left-leaning than among right-leaning respondents. Apart from this result, the statistical patterns stay the same when using our alternative specifications of political orientation.

Figure 2:

Coefficient plot: reasons in favor or against fee abolition. (a) Reasons in favor of abolition. (b) Reasons against abolition. Notes:Point estimates based on linear regression models. Dependent variables are binary variables indicating agreement with the respective statement. See Table D1 and Table D3 for the corresponding regression tables.

Of the 248 who did not want to abolish the fees, 46.4 % (115 respondents) agreed with the argument that “Only those families who pay day-care fees will benefit from abolition, and these are not the poorest”. 17.3 % (43 respondents) stated that “There are more sensible uses for taxpayers’ money”, 26.6 % (66 respondents) saw fees as necessary for the “Maintenance or improvement of childcare quality”, and 13.7 % (34 respondents) saw fees as a way to express “Service valuation”. Another topic that was occasionally mentioned was that the abolition is not financially feasible for the municipality or federal states. Again, as this topic was mentioned by only 17 people, we will not analyze it further.

Turning to reasons for rejecting the abolition of parental fees (see Figure 2b and Table D3), it is not surprising to observe that left-leaning politicians accept the argument that an abolition would mainly benefit the rich by a significantly higher percentage than right-leaning ones.^[18] Legislators who are new to parliament instead are less likely to support this argument. Of the remaining reasons, right-leaning politicians are more likely to accept the argument that “fees are justified as a compensation for the utilization of the service” (by 16 percentage points more than left-leaning ones). The argument that fees are necessary to sustain service quality is supported significantly more by left-leaning than by right-leaning politicians. While the effect does not persist when additional sociodemographic variables are included, Table D4 in the Appendix shows that when classifying people as left and right based on their self-classification as well as on preferences for taxes versus government spending, left-leaning politicians again turn out as more supportive of the quality argument. Moreover, we see that the argument of quality is mainly supported by female politicians, which might point to mothers having to bear the consequences of low-quality childcare (e.g. insufficient opening hours or unreliable service) more than fathers.^[19] Apart from the effect on the quality argument, all other statistical patterns remain the same when using our alternative specifications of political orientation.

7 Discussion

In an attempt to make sense of the results of the survey in the light of the theoretical considerations discussed in the Introduction, we can distinguish between two narratives on the equity implications of the childcare financing system and in particular the use of parental fees as an important financing instrument:

Narrative 1 (pro user fees) departs from the observation that public funds are always scarce and in particular there are limits to government revenues from income taxation. Therefore it is desirable to have additional instruments for redistribution such as income-dependent user fees for childcare, university education and potentially even more public services such as transport.^[20] In the case of childcare, user fees can also be justified by the fact that public childcare facilities enable parents of young children to participate in the labor market, earn incomes and maintain their human capital. From this point of view, it seems only fair that they bear a share of the costs of these facilities, where this share should increase in the “value” of their services to the parents, which can be measured by their earnings. An additional argument for these fees is potentially more controversial: As public funds are always scarce, user fees can help maintain a high quality of these facilities.

Narrative 2 (contra user fees) considers childcare facilities as an important part of the education system of a country and maintains that education is one of the key fields of government responsibility. According to this view, free and unlimited access, potentially even with mandatory enrolment at least for the last year before school starting age, is more important than income redistribution using fees as an instrument, in particular given that the revenue from these fees covers only a small part of costs anyway.

The results of our survey suggest that left-leaning politicians have a stronger tendency to subscribe to the second narrative, whereas more conservative politicians tend to the views described by the first narrative.

In 2013, Germany introduced a legal entitlement to a place in a childcare facility starting with the first birthday, but even 10 years after this date, places in childcare facilities are still scarce, and not every child whose parents demand a place can get one. This fact is especially worrying as a recent study (Hermes et al. 2023) shows that children from disadvantaged families (notably those with a migration background) tend to be discriminated against in the allocation process. Similarly, Huebener et al. (2023) provide survey results that show that low-income families are disadvantaged at the allocation of scarce places.

If the scarcity of places were at least partly caused by an insufficient financial endowment of the facilities and the levying of user fees from middle and high-income parents could alleviate this problem, leftist politicians might want to reconsider their rejection of these fees.

8 Concluding Remarks

The topic of user fees for educational services offered by governments, notably childcare and universities, turns out to be an interesting object for studying the political economy of fiscal policy, in particular the difference between the normative and the positive view of political decision-making on these matters. From the normative point of view, leftist politicians who tend to deplore rising inequality of disposable incomes in society should seize every opportunity to lower inequality by raising income-dependent user fees for these facilities. The positive view, in contrast, states that politicians compete for office and thus have to cater for the interests of clearly defined voter groups such as university students and parents of young children – with the result that more redistributive modes of financing will not survive once they get into the focus of an election battle.