Content Analysis of Family Policy Instruments to Promote the Sustainable Development of Families in China from 1989–2019

: Family policy involves a combination of policies enacted to address various family problems and improve the sustainable development of families. Evaluating family policy by considering policy instruments is conducive to optimizing policy allocation and promoting this sustainable development. This study constructs a two-dimensional analysis framework of policy instruments and policy themes and employs content analysis to conduct a quantitative analysis of 112 family policy texts issued by the Chinese government. The results show that the policy instruments used in China are not e ﬀ ective. The study also shows that environmental policy instruments are most frequently used, but the internal structure is unbalanced; supply-side policy instruments are moderately used; and the use of demand-side policy instruments is obviously limited. Policy themes focus excessively on “safeguard measures” and pay less attention to “parental welfare and protection”. Overall, China’s family policy is still in its infancy, as it focuses mainly on assistance and remains incomplete. Therefore, the parties responsible for the formulation of family policy should adjust and optimize the combinations of policy instruments that are employed are required to consider “general welfare”, and promote the two-dimensional integration of policy instruments and policy themes.


Introduction
Regardless of the type of political system, the level of economic development, or religious beliefs of a country, family is emphasized as being important for existence and is an institution that supports people's life development [1]. However, the process of modernization has had a great impact on families; especially since the 1960s and 1970s, the arrival of the "second demographic transformation" has fundamentally affected the establishment and breakdown of families [2]. The forms of marriage and family have become diversified, such as the increase in cohabitation and single-parent families [3], an increasing number of women have entered the labor market, resulting in paid work behavior that has greatly eroded the male breadwinner model, and the adult worker model has become increasingly prominent [4]. A strong individualization trend [5], together with the increasing instability and economic vulnerability of families [6,7], has led to new divisions between individuals and families. Social policy has had to respond to these new social risks, particularly by considering valuing and supporting the family as a social institution [5]. Therefore, since the mid-1990s, families and family-related policies have gained increased public attention in Western welfare states [8].
Scholars have become interested in family policy research in high-income countries. Many studies focus on various trends, results, and development directions of family policies under the family policy and to draw a comprehensive picture of China's family policy. This paper constructs a two-dimensional analysis framework of policy instruments and policy themes and uses a content analysis method to conduct a quantitative analysis of China's family policy texts from 1989-2019. The aim of this study is to reveal the shortcomings of China's family policy in terms of the choice and allocation of policy instruments, identify the positive policy implications of reforming family policy in China, and detail China's experience to help developing countries in similar situations.
This article is organized as follows. Section 2 presents the literature review. Section 3 summarizes the sample selection and method application. The fourth section presents the analysis based on the quantitative results of policy samples. The conclusion of this study and effective suggestions for promoting policy sustainability are presented in Section 5.

Literature Review
This paper analyzes family policy based on policy documents, so it is necessary to clarify what is meant by "family policy".
In terms of family policy, there is a broad sense and a narrow sense. From a broad perspective, family policies are instruments that have direct and indirect effects on families and family members [28]. On this basis, Kamerman and Kahn (1978) distinguished two types of family policies: explicit and implicit [29]. The former refers to governmental actions that directly target the family; the latter refers to the governmental actions that do not target the family but that ultimately have an impact on the family. Because the broad definition of family policy lacks pertinence and it is difficult to distinguish family policies from social policies, this study defines the explicit family policy in a restrictive way. That is, we define family policy as various institutional arrangements and related interventions made by the government directly for families and family members, with the ultimate goal of improving family welfare (or supporting family development). Although there are different definitions of family policy in the existing literature, there is still a certain consensus on the core areas that family policy should involve, such as financial support for families, services and welfare for working parents, and policies involving health, education, and family law [6].
Research shows that family policy in different countries is a heterogeneous field in which different objectives [30], and their development models differ considerably. According to Hantrais (2004), although family changes have similar characteristics in all European countries, there are substantial differences in family policies [31]. Bable (2008) suggests that the diversity of family policy models depends on the specific historical-cultural context and can distinguish different types of family policy according to the two key dimensions of "national policy on family-work relations" and "state-provided income benefits for families" [32]. In general, Western Europe can be divided into four models: Nordic, Mediterranean, liberal, and two different subtypes at the center of the continent exemplified by France and Belgium, and Germany and Austria [12]. For a long time, the classification model and comparative analysis of the family policies of these welfare countries have been strongly influenced by the work "Three Worlds of Welfare Capitalism" written by Esping Anderson (1990) [33]. However, some scholars still have questioned the rationality and applicability of this typology [34,35]. On the one hand, whether these types can be modified to include other countries with meaningful analysis results remains a controversial issue [36]. On the other hand, Ferragina and Seeleib-Kaiser (2011) confirmed the validity of the welfare state regime theory, but more importantly, they observed a dual transformation of the welfare country through a dynamic time method instead of a quite static approach to the welfare regime typology analysis [34,37].
Regardless of the development model, the main function of family policies in all countries is to coordinate and meet various goals and needs, such as the balance of work and family life [38], solving poverty [39], improving fertility levels [10], and promoting gender equality [40]. This means that family policy is a cross-cutting policy area [41]. Therefore, when analyzing family policy, we need to consider it as a multidimensional concept, rather than a unidimensional concept [42]. Reasonable Sustainability 2020, 12, 693 4 of 20 measurement tools based on multidimensional thinking and indicators are needed to evaluate the effectiveness of family policies [43].
As effective subjects of policy analysis, policy instruments, also regarded as "governing instruments", are the bridge between policy objectives and policy outcomes [44]. This area has been an important part of public policy research since the 1980s [45]. Policy instrument analysis, a systematic methodology, evaluates the application of various governance techniques and measures used by governments [46]. This type of analysis evaluates whether and how policies are implemented, revealing the core value and focus of policy and how much effort should be made by policy makers to achieve the goals [47]. In the field of family policy, the most widely used analysis in the OECD countries considers the diverse mix of policy instruments around cash benefits, childcare services, and generous leaves [43,48,49], which has resulted in both the convergence and variation of family policy development among countries.
However, there is no consensus in the literature on which specific policy instruments can be used to best describe a country's family policies. Many studies measure the strength of family policy instruments by constructing a series of proxy variables (family-related leaves such as maternity leave, parental leave, and paternity leave, children's allowance, and family service expenditures, etc.) [13,43,50]. Although multiple dimensions of family policy are considered in the selection of variables [12], this analytical method tends to ignore the internal characteristics of the policy itself, and few studies have conducted an overall analysis at the policy level. Most importantly, the selection of indicators is easily limited by data availability [6,51], which has a negative impact on the scope and depth of the analysis. As Lohmann and Zagel argue, policy-level analysis is important for the field of family policy [51]. The use of appropriate qualitative indicators can greatly improve the analysis of family policies [12]. To systematically describe the Chinese government's efforts to support the family, this paper proposes a new analytical framework for analyzing the promulgation of China's family policy that considers the content of the policy text and aims to capture policy characteristics at different levels by using a large number of broad indicators.
Many scholars have studied the choice of indicators, which is based on the classification of the policy instruments [52][53][54]. This paper uses the classification proposed by Rothwell and Zegveld and divides family policy instruments into three categories: supply-side policy instruments, environmental policy instruments and demand-side policy instruments [55]. Although this classification originated in the field of science and technology innovation policy [56,57] and has no obvious connection to the field of family policy, it addresses the mandatory and authoritative features of policy instruments. Moreover, this categorization considers the role of the government in policy promotion, rather than simply intervention and control. At the same time, this classification highlights the role of supply and demand in family development, indicating that the government should not only support families but also let society and the market "invest" in families and promote the internal sustainable development of families ( Figure 1). In addition, this classification has strong applicability due to the integrity and methodical nature of the system. In addition, the use of categorization is not limited to a particular industry.
Supply-side family policy instruments that directly promote the sustainable development of families are divided into five categories: information technology support, public service supply, infrastructure construction, capital investment, and education and training [58]. Environmental family policy instruments generally focus on laws and regulations, goal programming, tax concessions, financial policies and strategic measures [58]. These instruments play an indirect role in influencing and creating a family-friendly policy environment. Demand-side family policy instruments reflect the direct driving force for family development and reduce the vulnerability of families in society. Specifically, demand-side family policy instruments involve cooperation and outsourcing, government purchase, reward and encouragement, and market control [58].
highlights the role of supply and demand in family development, indicating that the government should not only support families but also let society and the market "invest" in families and promote the internal sustainable development of families ( Figure 1). In addition, this classification has strong applicability due to the integrity and methodical nature of the system. In addition, the use of categorization is not limited to a particular industry.  The theme of policy is identified through a high-level summary and refinement of the core content and is used to identify the basic purpose of a policy and the goal of the text [59], which should indicate that the focus and resources of the policy are related to family policy. As early as the late 1980s, the European Union has identified four common issues considered in family policy: reconciling work and family, assisting poor families, developing community policies for families, and promoting child protection [60]. However, recent studies have found that some changes and a convergence have occurred in the family policy of the welfare states, which are gradually focusing on the three themes of work/family reconciliation, gender equality, and family income protection [11]. Another important form of family policy addresses aging and health and seeks to help families deal with the current and future responsibilities of caring for their elderly or sick family members [1]. Due to their different cultures, institutional configurations and economic development stages, countries have different concerns about family policies [30,61], but on the whole, they mainly focus on themes of marriage and its diversified family forms, childbearing and childcare, work-family balance, gender equality, elderly care, and healthcare [62].
Considering the abovementioned family policy themes and the content of China's family policy, this paper divides the family policy themes into six categories: economic support, parental welfare and protection, child development, elderly care, special family support, and safeguard measures ( Table 1). It should be noted that the indicator used to measure policy on "parental welfare and protection" is the concentrated embodiment of gender equality and work-family balance, which are two different but inseparable concepts [43]. To help women achieve gender equality in employment, corresponding measures need to be taken to ensure their work-family balance, such as changing the image of women as traditional caregivers, and increasing opportunities for male spouses to participate in family care [15,63]. Although these "safeguard measures" are not directly related to family support, they can only be realized through government guidance, overall planning, management, and close cooperation, which are vital for supporting the effective implementation of family policy.
Based on research on policy instruments and the themes of family policy, this paper constructs a two-dimensional theoretical analysis framework to analyze family policy texts. This framework is shown in Figure 2.

Policy Themes Description
Economic support Family income protection [11], mainly cash transfers for family-related tax breaks, welfare benefits, poverty reduction measures, etc. [64] Parental welfare and protection Balance family and work and eliminate gender discrimination in employment-related policy arrangements, such as leave systems, labor protection, etc. [65] Child development Various efforts made by the government to promote child wellbeing and reduce child poverty and maltreatment, including early childhood care and education, health care and related community services [66,67].
Elderly care Policies that protect the informal support system, as well as formal support services that complement family care to promote the aging of the elderly at home and to ensure the function of family support for the elderly [68].

Special family assistance
Relevant services and support for people with disabilities, low-income families, family-planning families (which refers to the family having only one child or legally adopting one child), military families, etc. [69] Safeguard measures Ensure the smooth implementation of policies and formulate relevant safeguard measures, such as strengthening leadership, responsible implementation of relevant functional departments, capacity building, overall planning, supervision, etc. Based on research on policy instruments and the themes of family policy, this paper constructs a two-dimensional theoretical analysis framework to analyze family policy texts. This framework is shown in Figure 2.

Data Collection
To ensure that the analysis reflects the whole picture of China's family policy as much as possible, certain aspects are considered in selection of the samples. The first aspect is authority. The policies are mainly obtained from the official websites of the State Council and its directly related ministries (such as the Ministry of Education, National Health Commission, Ministry of Human Resources and Social Security, etc.) and involve laws, regulations, methods, opinions, and notices. The second aspect is the timeline. Because some laws and policies are constantly revised and improved, the selected policies are based on the most recently revised and released policies, and previously repealed policy documents are not considered. The third aspect is accuracy. Although some policies are closely related to family policy, the provisions in the text are either categorized as "regulating family responsibilities and obligations" or family support is provided only at the spiritual level. For example, marriage law states that "family members should respect the old and love the young" and aims to "protect the legitimate rights and interests of women, children and the

Data Collection
To ensure that the analysis reflects the whole picture of China's family policy as much as possible, certain aspects are considered in selection of the samples. The first aspect is authority. The policies are mainly obtained from the official websites of the State Council and its directly related ministries (such as the Ministry of Education, National Health Commission, Ministry of Human Resources and Social Security, etc.) and involve laws, regulations, methods, opinions, and notices. The second aspect is the timeline. Because some laws and policies are constantly revised and improved, the selected policies are based on the most recently revised and released policies, and previously repealed policy documents are not considered. The third aspect is accuracy. Although some policies are closely related to family policy, the provisions in the text are either categorized as "regulating family responsibilities and obligations" or family support is provided only at the spiritual level. For example, marriage law states that "family members should respect the old and love the young" and aims to "protect the legitimate rights and interests of women, children and the elderly". According to the narrow definition used for the abovementioned family policies, such policies are not truly "support policy". Therefore, only policy samples that clearly reflect government support measures are used as the research objects. A total of 112 policy texts were ultimately selected as the analysis samples. It should be noted that policies have been enacted since the founding of the People's Republic of China, and there have been several institutional adjustments; therefore, some relevant policy samples were not identified during the selection process and thus are not analyzed here. However, the selection process considered policies enacted by various policy departments, and the collected samples varied in terms of scope and type. Therefore, the existing policy samples can meet the needs of the research.

Content Analysis Coding
The content analysis method employed here is a combination of systematic quantitative and qualitative analyses on the content of policy documents. Using this method, the non-quantitative and unstructured information provided in the policy documents is transformed into quantitative data, and meaningful categories are established to decompose the policy content [70]. Based on the proposed analysis framework and samples, the analysis unit is defined, the data categories are established, the text is coded, the reliability and validity are tested, and finally, to reveal the effect of policy implementation, the data are analyzed and the results are explained [71]. In this paper, the content analysis method is used to extract the policy provisions related to family support from 112 policy sample texts; these are the most basic units of analysis. To avoid omission bias, all analysis units in the sample are sorted by employing word frequency analysis and the node coding function of the qualitative analysis software Nvivo 12 (QSR International Pty Ltd, Doncaster, Victoria, Australia). Based on the dates that the policies were enacted, the code is divided into five levels according to the form of "policy number-specific chapters/clauses-key points", with a total of 1756 policy code points (as shown in Table 2).
According to the established two-dimensional analysis framework, the "supply-side policy instruments", "environmental policy instruments", "demand-side policy instruments", and "policy themes" are set as tree nodes in Nvivo software. There are subnodes under the nodes in the tree. For example, the supply-side policy instruments include subnodes such as education and training, and capital investment. Finally, all policy coding points are classified into corresponding nodes to form the node table shown in Table 3.  Due to the subjectivity of the quantitative information provided in the policy texts, to avoid errors and improve accuracy, two trained coders were selected to independently code the information. The coding comparison function of Nvivo 12 software was employed to ensure the consistency of the coding results and enhance the reliability of the coding process. The final results show that the consistency of all nodes is higher than 0.82, which meets the reliability requirements (consistency of higher than 0.8 indicates that the reliability result is acceptable, and consistency higher than 0.9 indicates higher reliability). Therefore, the coding results are reliable.

Descriptive Statistics
As shown in Figure 3 Note: Due to limited space, only some coding information is presented.
Due to the subjectivity of the quantitative information provided in the policy texts, to avoid errors and improve accuracy, two trained coders were selected to independently code the information. The coding comparison function of Nvivo 12 software was employed to ensure the consistency of the coding results and enhance the reliability of the coding process. The final results show that the consistency of all nodes is higher than 0.82, which meets the reliability requirements (consistency of higher than 0.8 indicates that the reliability result is acceptable, and consistency higher than 0.9 indicates higher reliability). Therefore, the coding results are reliable.

Descriptive Statistics
As shown in Figure 3 Table 4). The Ministry of Civil Affairs issued 38 policies either independently or jointly with other departments. The Ministry of Civil Affairs is mainly responsible for supporting vulnerable and poor groups; thus, its policy covers a narrow population and provides very limited family security. Except for the Ministry of Civil Affairs, which issues more documents independently than jointly, departments mainly issue documents jointly, although the number of documents issued by the Ministry of Finance, the National Health Commission, the Ministry of  Table 4). The Ministry of Civil Affairs issued 38 policies either independently or jointly with other departments. The Ministry of Civil Affairs is mainly responsible for supporting vulnerable and poor groups; thus, its policy covers a narrow population and provides very limited family security. Except for the Ministry of Civil Affairs, which issues more documents independently than jointly, departments mainly issue documents jointly, although the number of documents issued by the Ministry of Finance, the National Health Commission, the Ministry of Human Resources and Social Security is not small. Due to the wide scope of family policies and the decentralization of family-related functions, the introduction of policies cannot be separated from multisectoral cooperation. However, in the case of unclear cross boundaries of department functions, multi-department cooperation easily leads to "buck passing" among departments, which greatly reduces the effectiveness of the policy. In addition, due to the lack of top-level design, these family policies are still in a fragmented state, so it is necessary to establish a department in charge of family affairs to integrate the relevant functions and resources between different departments and then provide the family with the greatest support [25].

X Dimension: Policy Instruments
Overall, China's application of policy instruments supporting families involves the use of supply-side, environmentally and demand-side instruments. These three types of policy instruments are obviously different in terms of their specific applications (Table 5): environmental policy instruments are the most frequently used, accounting for nearly one-half (49.83%) of the instruments; supply-side policy instruments account for 36.73%; and the application of demand-side policy instruments is obviously insufficient, at only 13.44%. An excessive number of environmental policy instruments are used, and there is a significant imbalance in the internal structure. In the specific application of the instruments, more emphasis is placed on strategic measures, which rank first and accounting for 27.05% of the policy instruments (Table 5) and 54.29% of the environmental policy instruments (Figure 4). Goal programming instruments rank second at 24.34%, and laws and regulations instruments rank third at 14.40%. These findings indicate that the government prefers to use these normative and leading policy instruments to create a family-friendly environment, but these instruments focus on long-term behaviors, and the effectiveness of policies takes a relatively long time to be realized. Tax concessions and financial policy instruments are seriously insufficient, at only 4.34% and 2.43%, respectively. Thus, the Chinese government does not create opportunities for families to improve their economic security through taxation and financial support. As a result, environmental policy instruments play the role of only direction and guidance. They do not truly encourage the relevant interest groups to act, and the impact of these instruments on families is very limited.
The frequency at which supply-side policy instruments is applied is relatively moderate, and these instruments focus on the direct supply of supportive resources. In terms of specific measures, public service supply accounted for the largest share of this type of instrument at 29.51%, followed by capital investment at 23.26%. The remainder is ranked as follows: education and training (18.14%), infrastructure construction (16.59%), and information technology support (12.40%). Clearly, the government is accustomed to strengthening and supporting families through more direct ways, such as service supply and cash transfers. However, the use of different types of policy documents indicates that the government mainly focuses on "laws" and "opinions" that have strong macro effects and guidance and are affected by poor policy implementation. The coverage targets are limited to poor children, poor elderly individuals, people with disabilities, low-income families, and other groups with difficulties, and the policies overlap to a high degree. In addition, the use of policy instruments that provide information technology support is relatively weak. The urbanization of the population has enhanced the mobility and dispersion of households, and it will become increasingly difficult to effectively obtain data on families. Therefore, it is necessary to track families over time and build a family information platform, such as a family management information system, which would provide the key foundation for the government to effectively respond to the needs of families. Work on this system needs to be further strengthened. The frequency at which supply-side policy instruments is applied is relatively moderate, and these instruments focus on the direct supply of supportive resources. In terms of specific measures, public service supply accounted for the largest share of this type of instrument at 29.51%, followed by capital investment at 23.26%. The remainder is ranked as follows: education and training (18.14%), infrastructure construction (16.59%), and information technology support (12.40%). Clearly, the government is accustomed to strengthening and supporting families through more direct ways, such as service supply and cash transfers. However, the use of different types of policy documents indicates that the government mainly focuses on "laws" and "opinions" that have strong macro effects and guidance and are affected by poor policy implementation. The coverage targets are limited to poor children, poor elderly individuals, people with disabilities, low-income families, and other groups with difficulties, and the policies overlap to a high degree. In addition, the use of policy instruments that provide information technology support is relatively weak. The urbanization of the population has enhanced the mobility and dispersion of households, and it will become increasingly difficult to effectively obtain data on families. Therefore, it is necessary to track families over time and build a family information platform, such as a family management information system, which would provide the key foundation for the government to effectively respond to the needs of families. Work on this system needs to be further strengthened.
The overall use of demand-side policy instruments is insufficient, and these instruments have little effect on improving family development. Among all the indicators used for the three categories of policy instruments, the use frequencies of cooperation and outsourcing, market control, rewards and encouragement, and government purchases are all at a relatively low level, and the proportion of the policy codes represented by these categories is only 4.90%, 4.56%, 2.11% and 1.88%, respectively (Table 5). In the process of supporting families, it is impossible to achieve perfection by the power of the government alone, and the role of the market is indispensable. The purpose of employing cooperation and outsourcing is to promote exchanges and cooperation with social forces such as scientific research institutions and enterprises to solve the government's shortage of professionals and issues regarding single service supply and other problems, and to expand family support channels. Reward and encouragement are used to encourage family members to help each  The overall use of demand-side policy instruments is insufficient, and these instruments have little effect on improving family development. Among all the indicators used for the three categories of policy instruments, the use frequencies of cooperation and outsourcing, market control, rewards and encouragement, and government purchases are all at a relatively low level, and the proportion of the policy codes represented by these categories is only 4.90%, 4.56%, 2.11% and 1.88%, respectively (Table 5). In the process of supporting families, it is impossible to achieve perfection by the power of the government alone, and the role of the market is indispensable. The purpose of employing cooperation and outsourcing is to promote exchanges and cooperation with social forces such as scientific research institutions and enterprises to solve the government's shortage of professionals and issues regarding single service supply and other problems, and to expand family support channels. Reward and encouragement are used to encourage family members to help each other and involve family allowances and flexible working hours, which will enhance intergenerational solidarity and the family care function [27,72]. Government purchases include family care services and public goods that will help meet the basic needs of families. However, at present, the weak dependence of the Chinese government on these three kinds of measures will not lead to the development of a system in which multiple parties, the government, market organizations, communities, and families cooperate to establish an effective system. The current system cannot effectively guide the market-oriented development of family services. The lack of market regulation policy means that the government's market regulation in family-related fields is not in place. For example, the government department responsible for family services has not yet formed an effective supervision mechanism or service standards, and the service system is highly flawed. Table 6 shows that although existing policies consider all six themes of family policy, there are obvious differences in the concerns of individual policies. First, most of the policies, 34.17%, involve safeguard measures; this proportion is far greater than that of other policy themes. This result shows that the government maximizes the mobilization of various institutions and ensures the implementation of policies, but it also reflects the characteristics of "multiple departments of government, division of responsibilities, and multiple coordination matters" under the multi-policy subjects. Second, China's family policy still pays more attention to childhood development (21.98%). As some studies have shown, family policy often focuses on children's well-being, which is crucial for addressing child poverty [67]. Nevertheless, the supply of formal child care and childhood education in China remains insufficient, and the government's universal subsidy scheme is almost nonexistent [73]. Third, special family assistance (16.46%) includes all kinds of plans to assist special and poverty-stricken families. This assistance is generally provided to four categories of families, families with difficulties, families with disabilities, military families, and families involved in family planning, reflecting the government's desire to care for those in great need. Notably, China has specifically formulated population and family planning law (2002) to help those involved in family planning [22]. This represents a special part of China's family policy. Existing family policy is insufficiently concerned with the care of elderly individuals, economic support, and parents' welfare and protection and must be further strengthened. Policies focusing on the care of elderly individuals are mostly issued in the form of guidance. Although these policies involve the supply of services to elderly individuals and the construction of infrastructure for elderly individuals, the policies are not effective, and there is no policy support for people who care for elderly family members. Most economic support is provided in the form of relief subsidies, and there is a lack of inclusive family cash benefits. Most notably, the proportion of instruments focusing on parental welfare and protection ranks last and is only 7.57%, which indicates that China's policy support for balancing work-family life and gender equality is relatively weak. At the same time, this lack of support is the reason for the increasing prevalence of an alternative care mode in which grandparents are caretakers [74,75]. As Bordene et al. (2016) argue, in countries with a large early care gap, due to the low level of service provision, the scarcity of childcare services, the lack of paid leave, and the lack of part-time opportunities for women, the highly intensive participation of grandparents is caused by mothers' need to balance family and work (the family's welfare is not met) [76]. Despite the continuous calls to address for "gender equality" and "support for women's return to work" in policy planning, especially in the context of "the two-child policy", China has not introduced supportive measures (such as parental leave) to ease women's burden of providing care to their families, except for necessary maternity leave and health care services.

Cross Analysis of Policy Instruments and Policy Themes
The two-dimensional analysis (see Table 6) indicates that economic support, child development, elderly care, and family special assistance are mostly addressed in supply-side policy instruments, while parental welfare and protection and safeguard measures are mostly addressed in environmental policy instruments. This difference reflects not only the preference of the government but also the ineffective use of policy instruments. Instead of using demand-side policy instruments, the Chinese government relies more on the use of supply-side policy instruments to support families, emphasizing the role of push policies, that is, intervention rather than investment in families. Supply-side policy instruments and environmental policy instruments have been widely used to address the two policy themes of child development and special family assistance, reflecting the considerable amount of attention that China pays to protecting children and families with difficulties.
The policy theme of parents' welfare and protection tends to use environmental policy instruments, while supply-side policy instruments and, especially demand-side policy instruments are not applied enough. This phenomenon reflects that the support of family policy for this theme is still involves indirect goal planning and lacks direct and effective support for family responsibilities for members of both sexes [26]. In contrast, employment-oriented family policies in Western countries have continued to expand for decades [77], such as generous child allowances, higher expenditure on childcare services, and extended leaves (maternity leave, parental leave, and paternity leave) [13,15], which can help mothers to work and relieve the pressure on their family care [78]. Although this expansion trend is still seen as an "incomplete revolution" in women's roles [38], it should be acknowledged that greater efforts have been made in gender equality and work-family balance. However, China's existing family policy cannot provide more targeted support for women working in the family, resulting in work family imbalance and other problems, which produce increasing risks for families. This situation indicates that China's family policy needs to be adjusted to meet the "new" and urgent family needs.
Finally, in terms of safeguard measures, the serious imbalance in the use of policy tools is mainly manifested in the excessive use of environmental policy instruments. The use of environmental policy instruments does not mean that more is better. Because the environmental policy instruments mainly have an indirect impact [55]. The effect is relatively slow; and the purpose of safeguard measures is to promote the implementation of family policy quickly and effectively, so an excessive number of environmental policy instruments and fewer supply-demand policy instruments will affect the effectiveness of safeguard measures. Thus, it is difficult for these instruments to play their due role in family policy. There are still some irrationalities in the choice of policy instruments in China's current family policies. Therefore, the choice of policy instruments should fully consider the objectives and contents of policy themes [79].

Conclusions
This study constructs a two-dimensional analytical framework of policy instruments and policy themes to examine the effectiveness and sustainability of family policies. Using the content analysis method, we analyze the application of family policy instruments in China from 1989 to 2019. The results show that the existing family policy system is still imperfect and needs to be further expanded. The main conclusions are as follows.
After 2012, the Chinese government truly began to attach importance to the development of family policy. In contrast to the development of family policy and subsequent prosperity in Europe and the United States over the past 50 years [12,80], China's family policy is seriously lagging behind and has a long way to go.
From the perspective of policy instruments, the use of the three types of policy instruments is unbalanced, and the internal structure is significantly uneven. Wolff and Schönherr (2011) believe that the basic element of a tool is that it is a means to achieve behavioral and sustainable effects [81]. The highest proportion of environmental policy instruments reflects the government's determination to support families in their efforts to create a friendly environment [58]. However, the irrationality of its internal structure, that is, the excessive use of strategic measures and the low use of financial policies and tax preferences, lead to the limited role of environmental policies. The supply-sided policy instruments are moderate, but insufficient attention is paid to information technology support, which needs to be strengthened. The severe shortage of demand-based policy instruments, especially the severe shortage of government purchase and reward and encouragement, indicates that the current family policy is not able to play a driving role in the family [55]. In general, the application of the three policy instruments in family policy is not ideal. Studies have shown that the use of multiple policy instruments has proven to be best in a wide range of situations [82]. Therefore, it is necessary to promote the coordinated development of three policy instruments: environment, supply and demand.
In terms of policy themes, China's family policy has a clear tendency to focus on values. China's family policy primarily pays attention to supporting vulnerable groups, such as poor children and special families, but it mainly addresses emergencies rather than providing consistent support [25]. China's support for families is not as strong as that of conservative groups. That is, the conservative model still pays attention to the general needs of all families on the basis of helping families in difficulty and specifically focuses on active policy efforts that will promote work-family balance [11]. Obviously, there is not only a lack of policies to promote work/family life balance in China, which makes it difficult for women to strike a balance between supporting a family with pursuing a career and earning money [24], but also a lack of institutional arrangements to support ordinary families in the whole policy system [26]. For example, the lack of social welfare or institutional support for child care means that families need to continue to bear the burden of raising children [83]. Therefore, the system of family policy is unable to actively respond to new challenges that families face.
Overall, China's family policy system is not effective, and the role of policy instruments in addressing various policy themes is not sufficient. Family policy is strongly dependent on the use of supply-side policy instruments but weakly dependent on the use of demand-side policy instruments, which is manifested in the government's tendency to support individuals with specific needs through direct transfers. This indicates that family policy is currently still in the "foundational phase", and development is extremely limited; that is, there is a "deficit intervention", and China is far from reaching the "consolidation phase" of family investment [12].

Policy Implications
Behavioral intention is the initial determinant of behavior [84]. To support the development of Chinese families, this study highlights the following implications.
First, the "family" perspective should be introduced when making family policy [39], and a competent department should be responsible for family affairs. Currently, China's family policy is managed by different functional departments. The decentralization and fragmentation of the policy affect its implementation. Focusing on families is more cost-effective than focusing only focusing on individuals [80]. Therefore, to improve family policy, the administrative system needs to be overhauled, the functions and resources related to family affairs need to be integrated, the top-level design needs to be strengthened, and the integrity of the policy system needs to be improved.
Second, the structure of policy instruments needs to be optimized, and demand-side policy instruments need to be employed more often. The use of environmental policy tools should be reduced, and the effectiveness of policies needs to be improved. China should pay more attention to the supporting role of information technology, increase data collected on families, and improve the data collection processes through an increase in monitoring and the use of sharing platforms, and lay the foundation for targeting and tracking different types of families. The government should increase the overall use of demand-side policy instruments.
The role of the government should change from that of a direct provider of family welfare to an indirect investor, guide, and regulator. In addition, the government should mobilize all parties to work together to address family issues. Third, the government should promote the overall use and "general welfare" of family policies, especially those focusing on work-family balance. Family policies focus on meeting the needs of all families by supporting both poor families and ordinary families [85]. More importantly, due to the reality of a high female labor participation rate [28], especially in China, a combination of policy instruments should be used that can provide cash and services and can address issues such as paid leave, such as allowing for leave to care for children and providing childcare services, subsidies and benefits to support family caregivers. This will help fill the gaps in relevant national policies and improve women's work-life balance. In addition, parental leave will encourage fathers to take responsibility for family care [63].
Finally, the government should focus on the two-dimensional integration of policy instruments and policy themes to enhance the effectiveness of the family policy system. For a policy instrument to be effective, there must be cooperation between the party responsible for the policy instrument and the party responsible for the policy objective, object and environment [79]. The analysis of the sub-objectives and contents of different policy themes shows that different types of policy instruments should be selected to improve the scientificity of family policy.

Limitations and Future Research Directions
This study is affected by some limitations and there is a need to improve future research. Because there is currently no consensus on the definition of family policy, there is a certain amount of subjectivity involved in the process of selecting policy samples. At the same time, there are a large number of policy samples, policy has been developed over a long time period, and there are many departments involved in the policy-making process, all of which led to the omission of certain policies in this study. In addition, this article only analyzes the use of policy instruments by evaluating the policy text and does not consider the actual effect of policy implementation.