Communication with the market to guide public expectations has become a pivotal monetary policy instrument for central banks worldwide. Therefore, assessing the efficacy of communication in influencing personal expectations is essential for central banks. To examine the effectiveness of communication by the People's Bank of China (PBC), we conduct a randomized experiment in which subjects receive excerpts from China's Monetary Policy Implementation Report (MPIR) for the third quarter of 2022 as information interventions. This study exploits the exogenous variation generated by the treatment to explore how central bank communication influences individuals' expectations regarding the macroeconomy. The key findings are as follows: First, positive communication from the central bank can steer individuals' expectations about the likelihood of a recession in line with the central bank's objectives, and this effect is more pronounced among those who read financial news less frequently. Second, clear and specific retrospective communication is more effective in guiding expectations about the macroeconomy than vague and complex forward-looking communication. Third, PBC communication affects individuals' consumption plans, supporting the central bank's economic growth objectives.

1 Introduction

Following the 2008 financial crisis, conventional monetary policy in advanced economies has encountered significant obstacles, prompting central banks to rely on communication as a key policy tool increasingly (Blinder et al. 2008). Former Chair of the Federal Reserve Ben Bernanke once remarked that “Monetary policy is 98 percent talk and only 2 percent action.”¹ For instance, central banks issue statements about the future path of their policy tools to influence public expectations and economic behavior directly when interest rates approach the effective lower bound, known as forward guidance (Bassetto 2019). In China, the role of expectation management has steadily gained prominence since the phrase “stabilizing expectations” first appeared in the Government Work Report in 2009. In 2015, China's central bank, the People's Bank of China (PBC), introduced a dedicated section on “Central Bank Communication and Expectation Management” in its quarterly Monetary Policy Implementation Report (MPIR). The significance of expectation management was further underscored in July 2018 when the Political Bureau of the Communist Party of China (CPC) Central Committee put forward the goal of stabilizing expectations. This emphasis was reaffirmed in the 2021 release of the 14th 5-Year Plan, which stressed the necessity of focusing on expectation management and guidance. Since then, the annual Central Economic Work Conference has consistently highlighted the critical role of social expectations in driving economic development².

Exploring the impact of central bank communication and expectation management in China is crucial. For one thing, as the world's second-largest economy with increasing capital openness, China is expected to exert growing global spillover effects in policy. For another, PBC communication exhibits certain characteristics that distinguish it from those of developed countries, implying that findings from those economies may not be directly applicable to China. As noted by the previous PBC Governor Zhou Xiaochuan, the PBC's policy has multiple objectives. These include four annual goals: maintaining price stability, promoting economic growth, fostering employment, and ensuring a balanced balance of payments. There are two dynamic objectives: financial reform and opening-up, as well as the development of financial markets³. In practice, the PBC sometimes communicates to achieve various conflicting objectives simultaneously. This may make it difficult for the public to understand the purpose of central bank communication, thus undermining its effectiveness. Furthermore, PBC communication remains in its early stages. In terms of forward guidance, it primarily provides ambiguous descriptions of future economic conditions and monetary policy deployment (Guo and Zhou 2018).

In China, emphasizing central bank communication and expectation management has practical significance today. On the one hand, expectations affect macroeconomic fluctuations by influencing the behavior of market participants such as individuals and enterprises. According to the self-fulfilling prophecy theory (Merton 1948), a widespread belief in an economic downturn can result in reductions in consumption and investment, ultimately leading to an actual economic recession. On the other hand, given the complexity of China's current monetary policy environment, communication is essential for the PBC to enrich its monetary policy toolkit and improve policy effectiveness. First, the ability of conventional monetary policy to stabilize macroeconomic fluctuations has weakened with China's economic growth slowing down (Guo et al. 2016), and the normal monetary policy space is also gradually shrinking. For example, data from Wind show that the deposit reserve ratio of small and medium-sized deposit financial institutions in China has fallen from 18% in 2014 to 7% at the beginning of 2024, and the space for interest rate cuts has been greatly narrowed. Therefore, as Yi Gang, former governor of the PBC, noted in 2019, it is necessary to cherish the normal monetary policy space to support the sustainable development of the economy⁴. Second, the transmission of conventional monetary policy to the real economy in China is not smooth (Wu and Lian 2016), and monetary policy actions affect economic conditions only after a long and variable lag (Friedman 1961). Besides, prolonged loose monetary policy can lead to an accumulation of excess liquidity, which may cause inflation. Third, as the PBC continues to innovate its monetary policy tools, certain monetary policy operations in practice sometimes serve purposes that diverge from their theoretically noticeable effects (Wang et al. 2019)⁵. As such, it is essential for the central bank to communicate the purpose and function of these tools to the market to mitigate the public's understanding bias and secure the cooperation of market participants effectively. In summary, regular information releases alongside sporadic policy communication can help the PBC preserve normal monetary policy space and enhance policy effectiveness by directly guiding market participants' expectations and economic behavior.

Existing research on PBC communication and expectation management primarily falls into two categories. One focuses on simulating the effects of expectation management by constructing theoretical models such as DSGE (Guo et al. 2016; Guo and Zhou 2018). Another uses text analysis or event study methods to examine the impact of PBC communication on macroeconomic variables like asset prices, the RMB exchange rate, or inflation expectations (Xiong and Wang 2012; Zhu et al. 2016; McMahon et al. 2018). There are relatively few empirical studies on the micro-level effects of PBC communication, which mainly focus on enterprises (including banks). For example, Wang and Wang (2015) explore the relationship between central bank expectation management and bank risk-taking, while Wang et al. (2019) examine the impact of central bank verbal communication and actual actions on corporate investment behavior. The role of PBC communication in guiding individual expectations has received limited attention in the literature, largely due to the lack of individual expectation data before and after the central bank communication⁶. However, research on this topic holds significant practical importance. On the one hand, as the scope for conventional monetary policy narrows, China is increasingly moving toward structural adjustments in monetary policy. Understanding the heterogeneous effects of PBC communication on individuals can aid in more targeted communication efforts, referred to as structural communication, thereby improving communication efficiency. On the other hand, public expectations affect their consumption behavior, and consumption is playing a growing role in economic growth. According to data from the National Bureau of Statistics of China, in 2023, final consumption expenditure, gross capital formation, and net exports of goods and services contributed 82.5%, 28.9%, and –11.4% to economic growth, respectively, making consumption the dominant driver of GDP growth. Therefore, comprehending how PBC communication influences individuals' expectations and consumption behavior driven by expectations is crucial for effectively expanding domestic demand and thus fostering economic growth in China.

In addition to the difficulty in obtaining data, identifying the causal impact of China's central bank communication on individuals' expectations presents another key challenge. Specifically, central bank communication usually occurs alongside other macroeconomic events or policies, making it challenging to disentangle their influence on individuals' expectations. Recent years have witnessed a credible revolution in causal inference, emphasizing a return to the experimentalist tradition (Li and Xu 2022). Chen et al. (2023) argue that randomized controlled trials are more advantageous for policy evaluation because of their randomness, in contrast to quasi-experiments. By comparing the performance of different experimental groups, we can get attributional feedback for policy implementation, such as identifying the mechanism of the policy's functioning and uncovering reasons for failure. In developed economies, researchers have tried to address issues like data unavailability and endogeneity in this area of study by using survey experiments. Some utilize survey data collected before and after actual central bank communication (Enders et al. 2019; Lamla and Vinogradov 2019), while others use experimental surveys embedded with communication messages to evaluate the effectiveness of central bank communication (Bholat et al. 2019; Beutel et al. 2021). These studies reach mixed conclusions on the efficacy of central bank communication.

To examine the impact of PBC communication on individuals' macroeconomic expectations, we conduct a randomized experiment referring to Beutel et al. (2021). We use the summary of China's MPIR released by the PBC for the third quarter of 2022 as the information intervention material for the treatment groups. The exogenous changes induced by the information intervention help assess the effectiveness of central bank communication at the individual level. Previous research on the impact of PBC communication on individuals' expectations often struggles to separate the effects of concurrent macroeconomic events or policies. This study overcomes the endogeneity issue by using an experimental design that ensures any systematic differences between randomly assigned treatment and control groups arise primarily from the information intervention. The experiment was conducted offline and online in December 2022, with participants mainly consisting of economic and financial practitioners and full-time students majoring in finance and economics⁷. The key findings of this study are: (1) Positive central bank communication reduces individuals' expectations of the likelihood of a recession. (2) Clear and specific retrospective communication is more effective than vague and complex forward-looking communication. (3) The impact of central bank communication on individuals' expectations may vary across different groups, with those who read financial news less frequently being more susceptible to its influence. (4) After guiding individuals' macroeconomic expectations, central bank communication further affects individuals' consumption plans, contributing to the central bank's economic growth objectives.

The rest of the article is structured as follows: Section 2 surveys the related literature. Section 3 describes the experimental design and data. Section 4 empirically analyzes the causal effect between central bank communication and individuals' macroeconomic expectations. The impact of central bank communication on public consumption plans is presented in Section 5. Section 6 concludes.

2 Literature Review

We contribute to the growing literature on the effectiveness of central bank communication, especially China's central bank communication. By event study and text analysis methods, existing literature has extensively investigated the impact of central bank communication in developed economies on financial markets, including monetary policy announcements, regularly published reports, and governor speeches (Hansen and McMahon 2016; Nakamura and Steinsson 2018; Cieslak and Schrimpf 2019; Ehrmann and Talmi 2020; Gardner et al. 2022). For example, Hansen and McMahon (2016) use computational linguistics tools to construct indicators for Federal Reserve communication. The empirical results based on the Factor-Augmented Vector Autoregressive (FAVAR) model show that communications regarding the current economic situation and forward guidance affect both the interest rate market and the stock market, but forward guidance is more effective. Cieslak and Schrimpf (2019) categorize the communications of the Federal Reserve, the European Central Bank, the Bank of England, and the Bank of Japan into monetary news and nonmonetary news, finding that nonmonetary news has a greater impact on the financial market during the financial crisis while monetary news has become more and more important since 2013. For China, abundant studies use text analysis methods to construct communication indices and investigate the effect of PBC communication on the capital market. They have explored the different impacts of various communication channels and monetary policy orientations. Zhu et al. (2016) employ the Exponential Generalized Autoregressive Conditional Heteroskedasticity (EGARCH) model to assess the impact of PBC communication on RMB exchange rate fluctuations and confirm the positive role of communication in exchange rate expectation management. They also find that written communication is more effective than oral communication, and among oral communications, those delivered by the governor are more effective than others. Using an event study approach, Zou et al. (2020) find that loose PBC communication positively influences China's stock market, while the impact of contractionary communication is negative. McMahon et al. (2018) provide empirical evidence that the release of quarterly MPIR reduces the price volatility of short-term notes.

However, there is limited research examining the impact of China's central bank communication on individuals' expectations, in contrast to its influence on financial markets. Most studies lack accurate data on individual expectations, relying instead on asset prices and economic models to estimate inflation expectations. For instance, Yan and Gao (2017) construct proxy variables for inflation expectations of the public and financial market participants based on survey indices and treasury bond yield data. They argue that information from PBC significantly influences inflation expectations, with an effect better than monetary policy actions. Wang and Wang (2015) estimate inflation expectations using central bank expectation management information and other economic indicators, discovering that PBC's expectation management policy helps reduce inflation expectations volatility. Nevertheless, Guo et al. (2023) highlight that asset prices only indirectly reflect expectations, subject to complex influencing factors and considerable noise. Hong et al. (2023) also point out that constructing an econometric model to extract effective information from the financial market to predict inflation is inferior to forecasting inflation expectations based on survey data.

In developed economies, a broad body of empirical literature assessing the influence of central bank communication on the beliefs and expectations of the general public or firms through survey experiments in recent years. Lamla and Vinogradov (2019) surveyed consumers' expectations about inflation and interest rates before and after the Federal Open Market Committee (FOMC) meetings. The empirical results indicate that although Federal Reserve communications have no measurable direct effect on personal economic expectations, they make the public more likely to receive the news released by the central bank. Enders et al. (2019) compare the changes in German enterprises' expectations before and after the monetary policy announcement. They find that a moderately contractionary monetary policy announcement lowers output and price expectations, while a moderately expansionary surprise raises them. D'Acunto et al. (2020) conducted a large-scale random questionnaire experiment, verifying that communication about targets is better than communication about policy instruments in managing expectations. Based on a random information experiment, Beutel et al. (2021)8 find that individuals expect the possibility of a financial crisis to increase after receiving financial risk warning information from the central bank.

The above studies use survey experiments to obtain real, rather than estimated, individual expectations across various categories, such as inflation, output, and financial conditions. Leveraging these real expectation data allows us to analyze the heterogeneous responses of different groups to central bank communication, thereby facilitating more tailored and structured policy implementation. Furthermore, survey experiments offer another distinct advantage: they allow for a clearer identification of the causal impact of central bank communication. In practice, central bank communication often responds to events in the real economy, which can lead to reverse causality issues in identification. Moreover, some macroeconomic events that coincide with communications may cause omitted variable bias. As such, isolating the effect of central bank communication itself is challenging. By a randomized controlled design with control and treatment groups, survey experiments minimize systematic differences between them other than the experimental intervention, effectively alleviating the endogeneity issue. Of course, every coin has two sides. While the advantages of survey experiments are clear, they also come with limitations: on one hand, unlike research based on existing data published by the authorities, investigating the effectiveness of central bank communication through survey experiments requires more sophisticated pre-designs and can be costly; on the other, due to the controlled nature of experiments, such studies often struggle to stimulate the complexity of the real world and typically focus only on a few specific aspects of central bank communication. Our study focuses on the impact of central bank communication content by designing treatment and control groups based on content variations rather than the mode or transmission of communication. To the best of our knowledge, we are the first to provide direct evidence on the effects of PBC communication on individuals' expectations about macroeconomy by survey experiment.

In addition, we contribute to relatively few studies on the impact of PBC communication on personal consumption. The relevant literature primarily simulates the effectiveness of PBC communication in influencing household consumption through the dynamic stochastic general equilibrium (DSGE) model or laboratory experiment (Guo et al. 2016; He and Zhao 2022). Guo et al. (2016) construct a DSGE model incorporating expectation error shocks and expectation management policy. The numerical simulations indicate that PBC's provision of information with high accuracy can reduce the deviation of public expectations, thereby reducing the fluctuation of household consumption and promoting its faster return to a steady state. He and Zhao (2022) carry out a laboratory experiment based on the intertemporal consumption-savings model, and find that forward guidance with a high commitment level, which provides guarantees, can guide individuals' expectations and smooth consumption.

In summary, this article contributes to existing literature in two ways. First, it fills a gap in the research on the effectiveness of PBC communication on individuals' expectations. By using a survey experiment to obtain individuals' expectations about the macroeconomy in China, we not only alleviate the potential endogenous problem in this area of research but also conduct a heterogeneity analysis. Second, this article provides the first empirical evidence of the causal effect of PBC communication on personal consumption planning, which lays a foundation for a deeper understanding of how PBC communication influences consumer behavior and the improvement of PBC communication.

3 Randomized Experiment

3.1 Experiment Design

We executed a randomized information experiment embedded in a survey online and offline from December 3 to December 12, 2022, during which time 656 valid questionnaires were gathered⁹. This experimental setup consists of three stages: a baseline stage, collecting respondents' prior macroeconomic expectations and consumption plans; a treatment stage, during which respondents are randomly assigned to either receive the summary of China's MPIR for the third quarter of 2022¹⁰ or receive no materials; a final stage, eliciting posterior expectations and consumption plans (see Figure 1).

Details are in the caption following the image — **Figure 1**
Open in figure viewer PowerPoint

Setup of the experiment. *Note:* The stars are randomization points, indicating that respondents are randomly assigned to one of four groups. About One-third of the respondents are allocated to the treatment group 1, in which they receive the full summary of the MPIR. About one-sixth of respondents receive the first half of the summary of the MPIR (treatment 2, retrospective communication), while one-sixth receive the second half of the summary of the MPIR (treatment 3, forward-looking communication). One-third of the respondents are assigned to the control group, receiving no extra information. This Figure is drawn with reference to Beutel et al. (2021).

3.1.1 Baseline Stage

In this stage, respondents are asked to answer questions about demographic characteristics such as gender, age group, and risk appetite. Then, similar to Roth and Wohlfart (2020) and Beutel et al. (2021), respondents receive a brief introduction on how to express expectations for future events in terms of probability values¹¹. Next, respondents are tasked with giving their expectations for personal disposable income in the next 12 months. Subsequently, we provide them with the concept of a recession and elicit their expectations for the probability of a recession in China in the next 12 months. In addition, respondents are asked about their total consumption expenditure plans in the next 12 months. The main questions in the questionnaire are shown in Supporting Information S1: Appendix A. The two questions most relevant to this study are presented below:

Macroeconomic expectation: What do you think is the probability of a “recession” occurring within the next 12 months in China?¹²

Consumption plan: What do you anticipate your consumption expenditure will be over the next 12 months?

3.1.2 Treatment Stage

During the treatment stage, individuals are randomly assigned to one of four subgroups. Respondents in the control group do not receive any additional reading materials. In contrast, respondents in treatment group 1 receive the full summary of the MPIR for the third quarter of 2022 (see Supporting Information S1: Appendix B), a highly condensed version of the full MPIR text. The first half focuses on the economic conditions and monetary policy since the beginning of the year, while the second half offers a perspective on the economic conditions and monetary policy in the future¹³. Respondents in treatment group 2 receive only the first half of the summary, which begins with macroeconomic data, such as GDP, highlighting the apparent rebound of China's economy in the first three quarters of 2022. In addition, this section reviews the major monetary policies implemented over the past three quarters, indicating a moderately loose stance through increased credit volumes and declining interest rates. Subjects in treatment group 3 receive the second half of the summary, which mainly contains the content that the long-term positive fundamentals of the economy will remain and that the support of sound monetary policy for the real economy will be enhanced in the future. However, this portion is expressed more generally and lacks specific commitments. Overall, the summary of the MPIR for the third quarter of 2022 conveys a positive outlook regarding macroeconomic conditions and monetary policy.

3.1.3 Final Stage

In the final stage, respondents are prompted to answer the same questions as in the baseline stage so that we can measure changes in their expectations regarding macroeconomic conditions, income levels, and consumption. Next, respondents need to complete a quiz designed to measure their economic and financial literacy. Finally, we gather information about their highest level of education, type of occupations, study experience in finance and economics, frequency of reading financial news, and investment conditions. Through this experiment embedded in the questionnaire, we can evaluate the effectiveness of PBC communication. See Supporting Information S1: Appendix C presents the principles of this experiment.

3.2 Data

3.2.1 Variable Definition and Descriptive Statistics

Table 1 shows the descriptive statistics for the main variables used in this study. When the subject is randomly assigned to treatment group 1, that is, he or she is expected to read the complete summary of the MPIR, ${T}^{{AL}}$ is equal to one, and zero otherwise. Similarly, for respondents in treatment group 2 or treatment group 3, ${T}^{{RP}}$ or ${T}^{{FW}}$ is equal to one, and zero otherwise. ${\text{Pr}({reces})}^{{pre}}$ denotes the expectation of the probability of a recession in China for the next year as reported by the respondent in the baseline stage, while ${\text{Pr}({reces})}^{{post}}$ reflects the expectation of a recession in the final stage, with a value from 0 to 10014. Variables about demographic characteristics (gender, age group, self-reported risk appetite, full-time student status, and academic degree) and economic knowledge level (financial literacy, study experience in finance and economics, frequency of reading financial news, and security investment conditions) are all binary variables. Specifically, the variable ${Female}$ equals 1 if the respondent is female. The variable ${Below}25$ is coded as 1 for respondents aged 18–25 and 0 for those aged 26–30, 31–40, or 41–50. The variable ${Risktaking}$ equals one when the respondent's reported risk appetite is greater than or equal to the average level of all the respondents, zero otherwise. If the individual is a full-time student, the ${Student}$ dummy variable is equal to 1. When the respondent is an undergraduate or holds a bachelor's degree, the variable ${Bachelor}$ equals 1. It equals 0 for master's students or those with a master's degree, as well as for PhD students or those with a doctoral degree. The variable ${Finaliteracy}$ is equal to 1 if the respondent answers all three financial literacy test questions¹⁵ in the questionnaire correctly. When the respondent has graduated or is studying a finance or economics major, the variable ${FEstudies}$ is equal to 1. The variable ${FEnewsreading}$ is coded as 0 for respondents who read financial news “hardly ever or never,” “once a month,” or “once a week,” and as 1 for those who read it “every day.” If the individual invests in individual stocks, funds, options, or futures, the ${Stockholder}$ dummy variable is equal to 1. $E({Income})$ and $E({Consum})$ denote the expected total personal disposable income and consumption expenditure for the upcoming 12 months, respectively.

Table 1. Descriptive statistics.

	Mean	SD	Min	Median	Max	Observations
${T}^{{AL}}$	0.35	0.48	0.00	0.00	1.00	656
${T}^{{RP}}$	0.16	0.36	0.00	0.00	1.00	656
${T}^{{FW}}$	0.18	0.38	0.00	0.00	1.00	656
${\text{Pr}({reces})}^{{pre}}$	46.06	28.58	0.00	46.50	100.00	656
${\text{Pr}({reces})}^{{post}}$	41.96	27.65	0.00	40.00	100.00	656
${Female}$	0.62	0.49	0.00	1.00	1.00	656
${Below}25$	0.52	0.50	0.00	1.00	1.00	656
${Risktaking}$	0.56	0.50	0.00	1.00	1.00	656
${Student}$	0.47	0.50	0.00	0.00	1.00	656
${Bachelor}$	0.27	0.44	0.00	0.00	1.00	656
${Finaliteracy}$	0.66	0.48	0.00	1.00	1.00	656
${FEstudies}$	0.78	0.41	0.00	1.00	1.00	656
${FEnewsreading}$	0.37	0.48	0.00	0.00	1.00	656
${Stock}h{older}$	0.70	0.46	0.00	1.00	1.00	656
${E({Income})}^{{pre}}$	166595.47	166726.85	15000.00	100000.00	600000.00	649
${E({Income})}^{{post}}$	164741.88	165759.77	18000.00	100000.00	600000.00	649
${E({Consum})}^{{pre}}$	85990.91	82315.63	10000.00	50000.00	300000.00	649
${E({Consum})}^{{post}}$	87523.57	83114.94	10000.00	50000.00	300000.00	649

Note: Variables ${\text{Pr}({reces})}^{{pre}}$ and ${\text{Pr}({reces})}^{{post}}$ are measured in %; Personal disposable income and consumption expenditure are both measured in RMB.

In all samples, 35% of the subjects are allocated to treatment group 1 to read the complete summary of the MPIR; 16% and 18% of the respondents are assigned to treatment groups 2 and 3, respectively, to read the first half and the second half of the MPIR, respectively. Additionally, 32% of individuals are assigned to the control group without reading any additional information. The proportion of participants in each group is consistent with our random distribution assignment. The respondents are mainly between 18 and 40 years old¹⁶. Nearly half of the respondents are full-time students, and 66% of them have obtained or are pursuing a master's degree. Generally, the respondents possess a high level of economic and financial knowledge: 66% of them answer all three financial literacy test questions correctly, 78% have experience in financial-related professional learning, 65% read financial news at least once a week, and 70% hold some investment in stocks, funds, options, or futures. To address outliers in the personal disposal income and consumption expenditure variables, we winsorize these relevant variables at the top and bottom five percent, following Roth and Wohlfart (2020).

3.2.2 A One-Way Analysis of Variance (ANOVA) of Equality Test

According to the experimental design, respondents are randomly assigned to the treatment and control groups to ensure that there are no systematic differences between the four subgroups, except for the different materials allocated to individuals during the treatment stage. We use a one-way ANOVA test to examine the equality among these subgroups for all pretreatment expectations and sample characteristics. Columns 2–5 of Table 2 present the mean values of the four groups for these variables, and column 6 reports the p-values of the test statistics. The null hypothesis that the mean values of these subgroups are equal cannot be rejected for all pretreatment and sample characteristic variables, which validates the randomness of our experimental design. Nevertheless, in the regression below, these variables are still used as control variables to mitigate potential endogeneity issues.

Table 2. One-way ANOVA of equality test.

	Control	Treatment 1	Treatment 2	Treatment 3	p value
${\text{Pr}({reces})}^{{pre}}$	48.34	44.70	43.68	46.71	0.455
${E({Income})}^{{pre}}$	152237.50	171863.11	183155.50	167578.95	0.428
${E({Consum})}^{{pre}}$	78918.27	88534.22	95661.76	85222.81	0.367
${Female}$	0.68	0.58	0.60	0.59	0.159
${Below}25$	0.53	0.49	0.52	0.54	0.791
${Risktaking}$	0.55	0.56	0.54	0.58	0.943
${Student}$	0.48	0.45	0.52	0.43	0.536
${Bachelor}$	0.28	0.24	0.28	0.28	0.716
${Finaliteracy}$	0.66	0.63	0.67	0.69	0.718
${FEstudies}$	0.81	0.77	0.78	0.76	0.609
${FEnewsreading}$	0.32	0.37	0.39	0.42	0.332
${Stockholder}$	0.68	0.67	0.73	0.75	0.418
Observations	210	227	103	116

Note: The last column reports the p values from a one-way ANOVA test assessing whether each row variable differs significantly across the four groups: Control, ${T}^{{AL}}$ , ${T}^{{RP}}$ , ${T}^{{FW}}$ .

4 Central Bank Communication and the Expectation About Macroeconomy

According to the macroeconomic models with incomplete information, the general public does not fully understand the current economic situation and monetary policy due to infrequent updates to their information sets or the presence of noisy information (Roth and Wohlfart 2020). Central banks usually invest more resources in estimating and predicting economic conditions and are considered to possess superior information about the economic environment (Blinder et al. 2008). Therefore, in theory, when the public believes that the central bank is credible, the central bank can provide information about economic conditions and monetary policy to effectively manage public expectations by prompting individuals to update their personal information sets and gain a better understanding of the economic situation and monetary policy (Morris and Shin 2002; Wang et al. 2019). In this section, we validate this theoretical conclusion using experimental data.

4.1 Simple Analysis

After combining the three treatment subgroups into a single treatment group, we present scatterplots with fitted lines in Figure 2 to compare the correlations between recession probability expectations at the baseline and final stages for the treatment and control groups. The slope of the fitted line for the treatment group is lower than that of the control group, which tentatively suggests that positive central bank communication can reduce individuals' expectations of the probability of a recession.

Furthermore, we conduct a simple differential calculation of the mean macroeconomic expectations for each treatment group compared with the control group. Table 3 shows the differences in the means of macroeconomic expectations within and between subgroups, as well as the difference-in-difference (DID) values that we focus on in the empirical analysis below. It is noteworthy that even respondents in the control group adjust their views on the future by lowering their expectations of the probability of a future recession. Zwane et al. (2011) argue that simply participating in a survey may prompt respondents to think more and change their expectations. However, this confounding effect is counteracted when considering the DID values for the three treatment groups. Compared with the control group, respondents in all three treatment subgroups significantly reduce their expectations of recession probability, with the most significant reduction observed among participants assigned to read only the first half of the MPIR summary and the smallest decrease found among those designated to read the second half of the MPIR summary.

Table 3. A simple analysis of the adjustment of individual macroeconomic expectations.

	$C$	${T}^{{AL}}$	${T}^{{RP}}$	${T}^{{FW}}$	${T}^{{AL}}-C$	${T}^{{RP}}-C$	${T}^{{FW}}-C$
	Levels				Differences between groups
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
${\text{Pr}({reces})}^{{pre}}$	48.340	44.705	43.680	46.707	–3.636	–4.661	–1.634
${\text{Pr}({reces})}^{{post}}$	47.836	38.480	37.058	42.474	–9.356***	–10.777***	–5.362*
$\unicode{x02206}\text{Pr}({reces})$	–0.505	–6.225	–6.621	–4.233	–5.720***	–6.117***	–3.728***

Note: $C$ represents the control group, and ${{\boldsymbol{\unicode{x02206}}}\text{Pr}({reces}){\boldsymbol{=}}\text{Pr}{\boldsymbol{(}}{reces}{\boldsymbol{)}}}^{{post}}-{\text{Pr}{\boldsymbol{(}}{reces}{\boldsymbol{)}}}^{{pre}}$ . Asterisks *** and * represent significance at the 1% and 10% levels, respectively.

The above results roughly suggest a negative causal relationship between positive central bank communication and individual expectations of recession probability. We explore it in more detail below.

4.2 Regression Analysis

4.2.1 Effectiveness of Central Bank Communication

Based on the Law of the PBC, the main goal of the PBC's monetary policy is to maintain the stability of the currency value and thereby promote economic growth. Given the content of the MPIR for the third quarter of 2022 and the downward pressure on macroeconomy at that time, it can be inferred that PBC communication was primarily focused on fostering economic growth. The individuals' macroeconomic expectations most pertinent to this growth objective are expectations regarding economic growth or recession. According to the self-fulfilling prophecy theory (Merton 1948), individuals with pessimistic economic expectations tend to adjust their behavior accordingly, such as reducing consumption and investment. These behavioral changes can lead to an actual recession. Thus, reducing public expectations of a recession is crucial for achieving economic growth. In this context, we examine the effectiveness of central bank communication in alleviating individuals' expectations of the probability of a recession. The regression model is as follows:

\unicode{x02206}{\text{Pr}({reces})}_{i}={\alpha }_{0}+{\alpha }_{1}{T}_{i}^{{AL}}+{\alpha }_{2}{T}_{i}^{{RP}}+{\alpha }_{3}{T}_{i}^{{FW}}+{\alpha }_{4}{\text{Pr}({reces})}_{i}^{{pre}}+{\boldsymbol{\theta }}{{\bf{X}}}_{{\bf{i}}}+{\epsilon }_{i},

()

where

\unicode{x02206}{\text{Pr}({reces})}_{i}={\text{Pr}({reces})}_{i}^{{post}}-{\text{Pr}({reces})}_{i}^{{pre}}

{\text{Pr}({reces})}_{i}^{{post}}

represents the subject

i

's expectation of the probability of a recession in the final stage, and

{\text{Pr}({reces})}_{i}^{{pre}}

represents the expectation in the baseline stage.

{T}_{i}^{{AL}}

{T}_{i}^{{RP}}

, or

{T}_{i}^{{FW}}

is a dummy variable that equals to 1 if respondent

i

is assigned to the treatment 1 group, the treatment 2 group or the treatment 3 group, respectively, and 0 otherwise¹⁷. Referring to Beutel et al. (2021), the vector

{{\bf{X}}}_{{\bf{i}}}

contains control variables for demographic characteristics (gender, age group, self-reported risk appetite, full-time student status, and academic degree) and economic knowledge level (financial literacy, study experience in finance and economics, frequency of reading financial news, and security investment conditions). In addition, we also include

{\text{Pr}({reces})}_{i}^{{pre}}

in the control variables to alleviate possible endogeneity problems. The residual term

{\epsilon }_{i}

contains measurement errors and any unobserved sample characteristic variables.

Table 4 reports the ordinary least squares (OLS) estimation results of Equation (1), showing that the coefficients ${\alpha }_{1}$ , ${\alpha }_{2}$ , and ${\alpha }_{3}$ are all negative and statistically significant at the 1% level¹⁸. This suggests that information intervention can effectively change individuals' expectations about the macroeconomy. That is, positive PBC communication reduces individuals' expectations of an economic recession. The results in columns 1 and 2 of Table 4 show that among the information intervention materials, the first half of the MPIR summary has the best effect, followed by the whole summary, and the second half demonstrates the least impact. This preliminarily indicates that clear and specific retrospective communication is more effective than vague communication, a notion we further investigate in the next section.

Table 4. Effects of central bank communication on expectations about macroeconomy.

	$\unicode{x02206}{\text{Pr}({reces})}_{i}$	$\unicode{x02206}{\text{Pr}({reces})}_{i}$	$\unicode{x02206}{\text{Pr}({reces})}_{i}$	$\unicode{x02206}{\text{Pr}({reces})}_{i}$
	(1)	(2)	(3)	(4)
${T}_{i}^{{AL}}$	–6.309***	–6.586***
${T}_{i}^{{AL}}$	(1.295)	(1.297)
${T}_{i}^{{RP}}$	–6.872***	–7.033***
${T}_{i}^{{RP}}$	(1.483)	(1.496)
${T}_{i}^{{FW}}$	–3.993***	–4.150***
${T}_{i}^{{FW}}$	(1.333)	(1.331)
${T}_{i}^{{Pool}}$			–5.833***	–6.056***
${T}_{i}^{{Pool}}$			(1.011)	(1.009)
${\text{Pr}({reces})}^{{pre}}$	–0.162***	–0.155***	–0.161***	-0.154***
${\text{Pr}({reces})}^{{pre}}$	(0.020)	(0.020)	(0.020)	(0.020)
${Constant}$	7.325***	9.157***	7.276***	8.986***
${Constant}$	(1.167)	(2.539)	(1.165)	(2.529)
${Demographics}$	No	Yes	No	Yes
${Economic\; knowledge}$	No	Yes	No	Yes
${R}^{2}$	0.139	0.151	0.134	0.147
$N$	656	656	656	656

Note: ${Demographics}$ contains individual-specific controls (gender, age group, self-reported risk appetite, full-time student status, and academic degree). ${Economic\; knowledge}$ includes controls indicating levels of economic knowledge (financial literacy, study experience in finance and economics, frequency of reading financial news, and security investment conditions). Asterisk *** denotes significance at the 1% level. The standard errors robust to heteroskedasticity are in parentheses.

In addition, we aggregate the three treatment subgroups to compare the expectation changes with the control group, as presented in columns 3 and 4 of Table 4. As long as the subjects are expected to read the MPIR summary in the treatment stage, the variable ${T}_{i}^{{Pool}}$ is equal to 1; otherwise, it is 0. The results in column 4 reveal that after the information intervention, individuals' expectations of economic recession decrease by an average of 6.06%¹⁹, which amounts to 13.16% of the mean prior belief of 46.06%. With rational inattention, the public has an incomplete understanding of the macroeconomic situation and monetary policy. By providing relevant positive information, PBC communication makes the public more optimistic about the future and narrows the information bias among them.

4.2.2 Comparison of the Effectiveness of Different Types of Communication

Furthermore, we compare the effects of different types of central bank communication on individuals' macroeconomic expectations. As previously noted, the first half of the MPIR summary is a type of retrospective communication, which mainly illustrates the gradual improvement in the macroeconomy and the moderately loose monetary policy over the first three quarters of 2022 through reviewing specific and clear macroeconomic data and monetary policy operations. In contrast, the second half of the MPIR summary prospectively communicates about the coming economic situation and policy deployment. That is, the future economic fundamentals will be strong in the long run, and the PBC will increase the strength of prudent monetary policy to promote economic growth. This part of the statement is more general and lacks specific data and policy details. To analyze how these different types of communications impact changes in individuals' macroeconomic expectations, we employ the following regression model:

\unicode{x02206}{\text{Pr}({reces})}_{i}={\beta }_{0}+{\beta }_{1}{{TR}}_{i}+{\beta }_{2}{\text{Pr}({reces})}_{i}^{{pre}}+{\boldsymbol{\delta }}{{\bf{X}}}_{{\bf{i}}}+{\varepsilon }_{i}.

()

Same as Equation (1), $\unicode{x02206}{\text{Pr}({reces})}_{i}={\text{Pr}({reces})}_{i}^{{post}}-{\text{Pr}({reces})}_{i}^{{pre}}$ is the change of individual $i$ 's expectation regarding a recession before and after the treatment stage. ${{TR}}_{i}$ is a binary variable that is equal to one if individual $i$ is supposed to read the first half of the MPIR summary and equals zero if $i$ is expected to read the second half of the summary. The other settings in Equation (2) are the same as those in Equation (1).

Columns 1 and 2 of Table 5 present the OLS estimates for Equation (2). The coefficient ${\beta }_{1}$ is negative and statistically significant at the 10% level, indicating that the first half of the summary (retrospective communication) has a greater negative effect on the recession expectation than the second half (forward-looking communication). The results show that retrospective communication by the PBC is more influential than forward-looking communication in guiding individuals' expectations regarding the macroeconomy. On the one hand, retrospective communication enhances public understanding of PBC's goals and actions (Kryvtsov and Petersen 2021) and alleviates individuals' pessimistic expectations about the economy by providing positive information on past economic conditions and monetary policy. On the other hand, although the forward-looking communication conveys an optimistic outlook for the future macroeconomy and PBC's monetary policy, it also describes the current challenging external environment and an unstable foundation for economic recovery, which may offset the effect of the positive outlook (Wang et al. 2019). In addition, forward-looking communication is ambiguous, while retrospective communication uses clearer and more specific economic and financial data that is easier to understand and more credible. The conclusion that clear and simple communication is preferable to complex and ambiguous communication aligns with existing literature (Guo and Zhou 2018; Bholat et al. 2019; Beutel et al. 2021). The experimental results of Kryvtsov and Petersen (2021) also support the idea that simple and relevant retrospective announcements from the central bank are more effective in reducing the volatility of individuals' expectations than forward-looking statements without specifying the timing of future policy changes. Additionally, the information on economic policy changes obtained by economic agents includes the policy change information itself and policy noise. Economic agents act in full accordance with this information when the noise is zero, while they do not react in any way when the noise is infinitely large and economic policy uncertainty is extremely high (Hong et al. 2023).

Table 5. Comparative effects of different central bank communications on macroeconomic expectations.

	$\unicode{x02206}{\text{Pr}({reces})}_{i}$	$\unicode{x02206}{\text{Pr}({reces})}_{i}$	$\unicode{x02206}{\text{Pr}({reces})}_{i}$	$\unicode{x02206}{\text{Pr}({reces})}_{i}$
	(1)	(2)	(3)	(4)
${{TR}}_{i}$	–2.895*	–3.032*
${{TR}}_{i}$	(1.700)	(1.761)
${{TA}}_{i}$			0.619	0.598
${{TA}}_{i}$			(1.651)	(1.694)
${\text{Pr}({reces})}_{i}^{{pre}}$	–0.167***	–0.166***	–0.217***	–0.210***
${\text{Pr}({reces})}_{i}^{{pre}}$	(0.033)	(0.034)	(0.033)	(0.033)
${Constant}$	3.577**	4.170	2.861*	1.880
${Constant}$	(1.528)	(3.641)	(1.622)	(4.345)
${Demographics}$	No	Yes	No	Yes
${Economic\; knowledge}$	No	Yes	No	Yes
${R}^{2}$	0.136	0.161	0.147	0.164
$N$	219	219	330	330

Note: ${Demographics}$ contains individual-specific controls (gender, age group, self-reported risk appetite, full-time student status, and academic degree). ${Economic\; knowledge}$ includes controls indicating levels of economic knowledge (financial literacy, study experience in finance and economics, frequency of reading financial news, and security investment conditions). Asterisks ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. The standard errors robust to heteroskedasticity are in parentheses.

Moreover, we replace ${{TR}}_{i}$ with ${{TA}}_{i}$ in Equation (2) to compare the influence of the full summary with that of only the first half. If the individual $i$ is allocated to read the full summary of the MPIR in the treatment stage, the variable ${{TA}}_{i}$ is equal to one. If $i$ is expected to read only the first half of the summary, ${{TA}}_{i}$ is equal to zero. The results in columns 3–4 indicate that there is no significant difference between these two types of communication. This finding can be attributed to information overload. Generally speaking, an individual's performance (i.e., the quality of decision-making or reasoning) correlates positively with the amount of information they receive. However, when the information received exceeds a certain threshold, additional content will no longer be integrated into the decision-making process, a phenomenon known as information overload (Chewning and Harrell 1990; Eppler and Mengis 2004). The threshold for information overload is influenced not only by information quantity but also by characteristics of the information itself, such as ambiguity, relevance, uncertainty, or complexity (Keller and Staelin 1987; Schneider 1987). In this study, after reading the first half of the summary, the complex and ambiguous messaging in the second half causes the subjects to experience information overload and has no additional impact on their expectations of a recession. Consequently, compared with respondents who only read clear and specific retrospective communication content, respondents who read the full MPIR do not reduce their recession expectations more largely.

4.2.3 Heterogeneity Analysis of Communication Effectiveness

Are there significant quantitative differences in the effectiveness of central bank communication among respondents with different characteristics? To answer this question, we examine the average treatment effects (ATEs) across samples categorized by gender, degree, financial literacy, major, and frequency of reading financial news. The regression results are presented in Table 6. Specifically, undergraduates or respondents with a bachelor's degree are in the “Yes” group, while the others are in the “Low” group. Participants who correctly answer all three financial literacy test questions are assigned to the “High” group, while those who fail to do so are placed in the “Low” group. Individuals who have graduated or are studying finance, economics, or related fields are classified into the “Yes” group; otherwise, they belong to the “No” group. In addition, daily readers of financial news are included in the “Yes” group, while those reading less frequently are placed in the “No” group. We aggregate the three treatment subgroups and compare them to the control group, with each regression controlling for respondents' prior expectations and the demographic and economic knowledge variables noted above.

Table 6. Effects of central bank communication on expectation about macroeconomy: heterogeneous effects.

	Female	Male	Yes	No	High	Low	Yes	No	Yes	No
	Gender		Bachelor		Financial Literacy		Finance/Economics Major		Reading FE News Every Day
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
${T}_{i}^{{Pool}}$	–7.377***	–3.864*	–4.189**	–6.810***	–5.082***	–7.796***	–6.140***	–5.706**	−3.565**	−7.559***
${T}_{i}^{{Pool}}$	(1.190)	(1.969)	(1.941)	(1.174)	(1.125)	(2.038)	(1.091)	(2.597)	(1.622)	(1.303)
${\text{Pr}({reces})}_{i}^{{pre}}$	–0.147***	–0.161***	–0.126***	–0.164***	–0.146***	–0.164***	–0.137***	–0.177***	–0.129***	–0.165***
${\text{Pr}({reces})}_{i}^{{pre}}$	(0.023)	(0.036)	(0.044)	(0.022)	(0.023)	(0.036)	(0.021)	(0.046)	(0.033)	(0.025)
${Constant}$	6.572**	9.712**	5.068	9.580***	4.095	15.574***	11.243***	6.542	10.160***	10.496***
${Constant}$	(2.632)	(4.461)	(3.325)	(3.117)	(3.115)	(4.621)	(3.150)	(4.143)	(3.828)	(3.299)
${Controls}$	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
${R}^{2}$	0.178	0.131	0.124	0.169	0.139	0.195	0.147	0.207	0.134	0.159
$N$	406	250	174	482	430	226	514	142	240	416
SUEST p-value	0.119		0.236		0.234		0.873		0.051

Note: ${Con}{trols}$ contain demographic characteristics (gender, age group, self-reported risk appetite, full-time student status, and academic degree) and economic knowledge level control variables (financial literacy, study experience in finance and economics, frequency of reading financial news, and security investment conditions). Asterisks ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. The standard errors robust to heteroskedasticity are in parentheses.

The regression results indicate that the absolute value of the average treatment effect of PBC communication on female respondents is greater than that on male counterparts, although the difference is not statistically significant. This finding is consistent with the conclusions drawn by Armantier et al. (2016) and Beutel et al. (2021). One possible reason is that women have less confidence in their prior beliefs than men, which leads to a pronounced impact of information intervention on their expectations. In addition, respondents who do not read financial news daily show a significantly larger decline in the recession expectations to daily readers. This may be attributed to the fact that less frequent readers obtained more positive new information from the summary of the MPIR for the third quarter of 2022, whereas daily readers may have already been aware of the information before participating in the experiment. However, the average treatment effects of PBC communication do not exhibit significant differences across participants with varying education levels (bachelor's degree or higher), financial literacy²⁰, or backgrounds in finance or economics. The following are some possible reasons for the above results: First, the key positive information about economic conditions in the PBC communication materials we provided is relatively easy to capture and understand²¹, requiring neither advanced education nor specialized financial knowledge. Second, regardless of education, financial literacy, or economic background, respondents generally have almost similar levels of trust in the PBC's MPIR summary. We examine the impact of these three variables on the treatment group's trust²² in the information from the PBC's MPIR summary. The results show that these variables have no significant effect on trust (see Supporting Information S1: Appendix E), which supports the second explanation. Third, prediction patterns may remain consistent across various groups, such as students, families, and financial market participants (Cornand and Hubert 2020; Kryvtsov and Petersen 2021), meaning that information interventions influence the expectations of these groups similarly, with no significant differences on average.

4.2.4 Robustness Checks

To exclude the influence of extreme values on our results, we winsorize or trim $\unicode{x02206}{\text{Pr}({reces})}_{i}$ and ${\text{Pr}({reces})}^{{pre}}$ at the top and bottom 2% or 5%. We obtain qualitatively similar results as above, showing that respondents reduce their expectations of recession probability after reading the summary of the MPIR for the third quarter of 2022 (see Table 7). Additionally, the treatment group participants may not have read the intervention materials we provided, that is, there are non-compliers. To test whether this issue could bias the estimation results of Equation (1), we restrict the treatment group to those who answer the question regarding the trust level in the information from the MPIR summary they have just read (see Question A.13 in Appendix E of Supporting Information S1)²³. The regression results are presented in Table 8, where the results in column (2) are nearly identical to those in column (4) of Table 4. Thus, the conclusion that PBC communication can effectively guide individuals' macroeconomic expectations is robust.

Table 7. Robustness checks: winsorization and trimming.

	Winsorize 2%	Winsorize 5%	Trim 2%	Trim 5%
	(1)	(2)	(3)	(4)
${T}_{i}^{{Pool}}$	–5.690***	–5.130***	–4.663***	–4.335***
${T}_{i}^{{Pool}}$	(0.860)	(0.738)	(0.765)	(0.703)
${\text{Pr}({reces})}_{i}^{{pre}}$	–0.130***	–0.116***	–0.096***	–0.081***
${\text{Pr}({reces})}_{i}^{{pre}}$	(0.015)	(0.014)	(0.014)	(0.014)
${Constant}$	7.212***	5.665***	4.434***	3.158*
${Constant}$	(1.937)	(1.630)	(1.711)	(1.646)
${Controls}$	Yes	Yes	Yes	Yes
${R}^{2}$	0.157	0.163	0.129	0.121
$N$	656	656	631	544

Note: ${Con}{trols}$ contain demographic characteristics (gender, age group, self-reported risk appetite, full-time student status, and academic degree) and economic knowledge level control variables (financial literacy, study experience in finance and economics, frequency of reading financial news, and security investment conditions). Asterisks *** and * denote significance at the 1% and 10% levels, respectively. The standard errors robust to heteroskedasticity are in parentheses.

Table 8. Robustness checks: considering potential noncompliance issues.

	$\unicode{x02206}{\text{Pr}({reces})}_{i}$	$\unicode{x02206}{\text{Pr}({reces})}_{i}$
	(1)	(2)
${T}_{i}^{{Pool}}$	–5.952***	–6.327***
${T}_{i}^{{Pool}}$	(1.055)	(1.056)
${\text{Pr}({reces})}_{i}^{{pre}}$	–0.162***	–0.153***
${\text{Pr}({reces})}_{i}^{{pre}}$	(0.021)	(0.021)
${Constant}$	7.303***	9.232***
${Constant}$	(1.211)	(2.693)
${Controls}$	No	Yes
${R}^{2}$	0.136	0.152
$N$	598	598

Note: ${Con}{trols}$ contain demographic characteristics (gender, age group, self-reported risk appetite, full-time student status, and academic degree) and economic knowledge level control variables (financial literacy, study experience in finance and economics, frequency of reading financial news, and security investment conditions). Asterisk *** denotes significance at the 1% level. The standard errors robust to heteroskedasticity are in parentheses.

5 Further Analysis

The empirical evidence presented above suggests that positive communication from the PBC about the macroeconomy and monetary policy can lower the public's expectations of the probability of a future recession. Roth and Wohlfart (2020) argue that an innovation to anticipated future economic resources should prompt households to make immediate adjustments to their consumption. Therefore, can positive central bank communication increase personal planned consumption in the future? To answer this question, we examine the effect of central bank communication on respondents' consumption plans using a regression model similar to Equation (1):

\unicode{x02206}{{Ycr}}_{i}={\gamma }_{0}+{\gamma }_{1}{T}_{i}^{{Pool}}+{\gamma }_{2}{Y}_{i}^{{pre}}+{\boldsymbol{\eta }}{{\bf{X}}}_{{\bf{i}}}+{\mu }_{i},

()

where

{Y}_{i}={E({Consum})}_{i}

represents the consumption plan of individual

i

in the next 12 months.

\unicode{x02206}{{Ycr}}_{i}={(Y}_{i}^{{post}}-{Y}_{i}^{{Pre}})/{Y}_{i}^{{Pre}}

is the rate of change in individual

i

's consumption plan from pretreatment to posttreatment.

{Y}_{i}^{{Pre}}

and

{Y}_{i}^{{post}}

denote consumption plans of individual

i

at the baseline stage and final stage, respectively. The vector

{{\bf{X}}}_{{\bf{i}}}

contains the same control variables that we included in our previous estimations. Since there are extreme values of consumption plans in the sample, we apply 5% winsorization to

\unicode{x02206}{{Ycr}}_{i}

and

{Y}_{i}^{{pre}}

Columns 1 and 2 of Table 9 show that an individual's planned spending for the upcoming year significantly increases by 3% on average after the information intervention. There are two possible reasons behind this: First, economic recessions accelerate labor market deterioration (Farber 2011), while positive central bank communication makes individuals more optimistic about the future economic situation and labor market, thereby raising their expectations of permanent income levels (Krueger et al. 2016; Yagan 2019). Consumption theories suggest that individual consumption increases with permanent income (Friedman 1957; Houthakker 1958). Second, positive central bank communication can reduce individuals' expectations of future income uncertainty, thereby lowering precautionary savings (Lusardi 1998; Parker and Preston 2005). This, in turn, can enhance individuals' future consumption plans. In addition, we examine the impact of central bank communication on the change rate of personal expected disposable income over the next 12 months by estimating Equation (3) where ${Y}_{i}={E({income})}_{i}$ . The results are represented in columns 3 and 4 of Table 8, and the impact of central bank communication on the income change rate in the next 12 months is not statistically significant. It is important to note that this result does not contradict our first hypothesis, namely the permanent income hypothesis, as short-term income differs from permanent income²⁴. On the contrary, this finding is consistent with the view that income exhibits stickiness in the short run.

Table 9. Effects of central bank communication on individuals’ consumption plans and income expectations.

	(1)	(2)	(3)	(4)
	Consumption change rate		Income change rate
${T}_{i}^{{Pool}}$	0.030*	0.034**	0.010	0.011
${T}_{i}^{{Pool}}$	(0.016)	(0.016)	(0.012)	(0.012)
${Constant}$	0.043***	0.039	0.015	0.045
${Constant}$	(0.015)	(0.036)	(0.011)	(0.031)
${E({Consum})}_{i}^{{pre}}$	Yes	Yes
${E({Income})}_{i}^{{pre}}$			Yes	Yes
${Demographics}$	No	Yes	No	Yes
${Economic\; knowledge}$	No	Yes	No	Yes
${R}^{2}$	0.020	0.034	0.032	0.039
$N$	649	649	649	649

Note: ${Demographics}$ contains individual-specific controls (gender, age group, self-reported risk appetite, full-time student status and academic degree). ${Economic\; knowledge}$ includes controls indicating levels of economic knowledge (financial literacy, study experience in finance and economics, frequency of reading financial news, and security investment conditions). Asterisks ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. The standard errors robust to heteroskedasticity are in parentheses.

6 Conclusion

Central bank communication is increasingly becoming a critical tool for managing public expectations and enhancing the effectiveness of monetary policy. Although research on the effectiveness of the PBC's communication has gained growing attention, few studies directly examine the impact of the PBC's communication on individuals' macroeconomic expectations. This study incorporates the summary of China's MPIR for the third quarter of 2022 as an information intervention in a survey for individuals. Using this random information experiment, we assess the effectiveness of the PBC's communication in influencing individuals' expectations of the probability of an economic recession. The findings of this article are as follows: (1) By providing positive information about past or future economic conditions and monetary policies, central bank communication can reduce individuals' expectations of an economic recession. (2) Clear and specific retrospective communication is more effective than vague and complex forward-looking communication. (3) The impact of central bank communication on individuals' expectations of an economic recession is heterogeneous. Individuals who read financial news less frequently are more likely to be influenced. (4) After affecting individuals' macroeconomic expectations, central bank communication further affects personal consumption plans, and the direction of influence is consistent with the central bank's goals.

In light of the conclusions above, we propose the following policy recommendations: (1) The PBC should strengthen its communication with the public by delivering information regarding past and anticipated macroeconomic conditions and monetary policies. This would enhance public comprehension of the macroeconomic environment, the central bank's goals, and policy actions, thereby guiding public expectations and spending behavior. (2) Central bank communication should be as clear as possible, providing more specific policy measures and relevant data. If the PBC struggles to offer definitive forward-looking commitments or forecasts owing to multiple objectives and the necessity of maintaining credibility, it should minimize vague forward-looking communications and concentrate on providing clear retrospective communications. (3) The central bank should pay more attention to communication with groups prone to modify expectations and behavior in line with policy objectives, such as individuals who are less attentive to financial news, to improve the efficacy of the communication to a greater extent. (4) To better assess the effectiveness of the central bank's expectation management and make timely policy adjustments, it is recommended that relevant Chinese authorities conduct surveys that regularly track residents' expectations about the macroeconomy. Examining the impact of central bank communication on public expectations is the most direct way to evaluate the effectiveness of central bank expectation management, and such research relies on accurately measuring public expectations. Currently, data on residents' expectations in China primarily come from the quarterly Urban Depositors Questionnaire Survey conducted by the PBC's Department of Investigation and Statistics. Nonetheless, the frequency of these surveys is low, and only aggregated indices rather than specific data on individual expectations are published. Developed countries often use survey-based expectation data to evaluate the effectiveness of expectation management (Guo et al. 2023), and their experience may serve as a reference for the PBC.

Future research will be expanded in the following areas. First, we need to examine the effectiveness of central bank communication using a broader sample to enhance the external validity of the quantitative conclusions in the study. Second, we will develop various types of communication materials based on existing central bank communications to further explore which communication content is more effective in guiding individual expectations. For instance, we will design materials where the only difference between the reading materials for the treatment and control groups lies in readability or visibility. Third, additional questions related to demographic characteristics will be introduced to explore the heterogeneity of the effectiveness of central bank communication, which can help the PBC communicate more targetedly. Fourth, we will conduct a comparative study of the impact of various central bank communication channels through a randomized controlled experiment, such as official statements, traditional media, and online media.

Author Contributions

Yuying Jin: conceptualization, formal analysis, funding acquisition, investigation, methodology, project administration, resources, supervision, writing – original draft, writing – review and editing. Sunyao Xia: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, software, validation, visualization, writing – original draft, writing – review and editing.

Acknowledgments

The authors acknowledge financial support from the Major Program of the National Social Science Foundation of China (No. 21&ZD082) “Theoretical and Experimental Research on Cross-cycle Design and Adjustment of Macro-regulation in China from the Perspective of the New Development Paradigm” and the Fundamental Research Funds for the Central Universities from Shanghai University of Finance and Economics (No. CXJJ-2023-416) “Exchange Rate Expectation Management, RMB Exchange Rate, and Firm Export.” The authors also thank the two anonymous reviewers for their constructive comments and suggestions. All remaining errors are our own.

Ethics Statement

This study involved an anonymous survey in which participants voluntarily took part after providing informed consent. The study did not involve any personally identifiable information or sensitive data, and no deception was used.

Conflicts of Interest

The authors declare no conflicts of interest.

Endnotes

1 For more information, see https://www.brookings.edu/articles/inaugurating-a-new-blog/.

2 China's 2022 Central Economic Work Conference emphasized the need to improve public expectations and boost confidence in development. The 2023 conference highlighted the introduction of more policies to stabilize expectations, growth, and employment. In 2024, the conference stressed the importance of strengthening expectation management and coordinating policy implementation with expectation guidance.

3 For more information, see http://www.pbc.gov.cn/goutongjiaoliu/113456/113469/3090366/index.html.

4 For more information, see http://www.pbc.gov.cn/goutongjiaoliu/113456/113469/3895219/index.html.

5 From the third quarter of 2006 to the second quarter of 2008, the PBC raised the reserve requirement ratio 18 consecutive times, not only to prevent the economy from overheating but also to withdraw the excess base money injected due to the surge in foreign exchange reserves. Since 2013, the PBC has introduced several monetary policy tools, including the Short-term Liquidity Operation (SLO), Standing Lending Facility (SLF), and Pledged Supplementary Lending (PSL).

6 Data on individual consumer expectations in China are mainly from the quarterly questionnaire surveys of urban depositors conducted by the PBC. However, the PBC only publicizes general indices, such as the price expectation and consumption intention index. We cannot obtain data on expectations and intentions specific to individuals.

7 This sample composition may raise concerns about the external validity of our conclusions. However, the heterogeneous results presented in our article indicate that there is no significant difference in the effectiveness of central bank communication across groups with different educational backgrounds (undergraduate or higher), financial literacy levels (high or low), or experience in financial professional learning (with or without). This result may be related to the participants' reception of the information, their level of trust, and their response patterns, which we have elaborated on in the heterogeneity analysis section of this article.

8 This article is the closest analog to our paper. We extend the research of Beutel et al. (2021) in two key areas. Regarding the research topic, this article investigates how the PBC communication about economic conditions and monetary policy influences subjects' expectations about the probability of a recession and their consumption plans. In contrast, Beutel et al. (2021) examine how the Bundesbank's financial stability communication affects individuals' expectations of a financial crisis and investment behavior. In terms of information intervention materials, this article employs a more comprehensive summary of the PBC's quarterly Monetary Policy Implementation Report (MPIR), while Beutel et al. (2021) use only a few sentences from the Bundesbank's annual financial stability report.

9 We distributed and collected online questionnaires via the professional data collection platform Credamo, while offline randomized questionnaires were mainly collected from part-time MBA and EMBA students at a university in China that focuses on finance and economics. A total of 483 valid online questionnaires and 173 offline questionnaires were gathered.

10 MPIR is the most widely used form of written communication from the PBC (Li et al. 2024), and the report published on November 16, 2022, was the most recent MPIR available when the experiment was conducted. Based on the work of Bholat et al. (2019) and Beutel et al. (2021), we select content from this report as the information intervention for the experiment. There may be concerns that the use of written communication materials could affect the external validity of our conclusions, as other communication modes, such as press conferences, might differ in effectiveness compared to written communication. In practice, the central bank usually communicates the same content through both written forms such as written reports and oral forms such as press conferences. However, by designing both treatment and control groups, our study focuses on assessing whether the content of central bank communication itself—rather than the specific mode of delivery—has an impact.

11 Specifically, 2% and 5% indicate “almost no chance”; 18% or so means “not much chance”; 47% and 52% signify “pretty even chance”; 83% or so means “very good chance”; and 95% or 98% indicates “almost certain.” These descriptions are based on the works of Roth and Wohlfart (2020) and Beutel et al. (2021). Beutel et al. (2021) focus on the probability of a financial crisis, while Roth and Wohlfart (2020) are concerned about the likelihood of a recession.

12 Referring to Roth and Wohlfart (2020), we use the expectation of the probability of a recession as a proxy variable for the macroeconomic expectation, allowing respondents to select any probability between 0% and 100%. Since the information materials used in this experiment contain data on the GDP growth rate, directly asking subjects about their expectations for economic growth rate may lead them to anchor their responses to the figure of the GDP growth rate presented in the MPIR summary, potentially introducing bias due to the anchoring effect. The concept of “recession” given in the questionnaire is as follows: A recession is a situation where a country's economic growth stagnates. Macroeconomists generally recognize that a recession occurs when a country's GDP growth falls for two or more consecutive quarters, leading to a contraction in the business cycle, an imbalance in the supply and demand network, rising unemployment, a decline in personal purchasing power, and corporate bankruptcies.

13 We collect the views of 20 students majoring in finance and economics or practitioners in the financial industry on the two parts of the summary. 90% of them think that although both parts are positive descriptions of the economic situation and monetary policy, the former is more specific, more precise, and thus more credible. The latter not only has no specific data and policy support but also contains a description of future uncertainty, so its credibility is slightly weak.

14 The unit is %. In this article, the variables labeled pre refer to the expectations or plans before the treatment stage, that is, the baseline stage; variables labeled post refer to expectations or plans after the treatment stage, that is, the final stage.

15 The three questions are structured as single-choice items and pertain to the concept of “Jiangzhun,” the effect of inflation on the real purchasing power of savings accounts, and the calculation of compound interest.

16 In Supporting Information S1 (Appendix D), we provide more detailed distributions of the sample by age group, academic degree, and frequency of reading financial news.

17 We estimate the intention-to-treat (ITT) effect, which evaluates the impact of the intervention based on participants' initial group assignments, rather than the intervention they actually received. Although we emphasized the importance of careful and focused participation to the subjects before the experiment, our ITT estimate is likely a lower bound of the actual treatment effect in magnitude (Kling et al. 2007; Jessoe and Rapson 2014). The reason is that noncompliance may occur within the treatment group, where some participants may not have engaged with the central bank communication materials we provided, leading them to exhibit little change in expectations between the baseline and final stages, similar to the control group. However, we believe that the probability of noncompliance among control group participants is very small. From the outset of the experiment, we explicitly stated that we were solely interested in participants' actual opinions and perceptions—regardless of their accuracy. Specifically, within the experimental questionnaire, only the financial literacy test questions have objectively correct or incorrect answers, but whether participants answer them correctly is unrelated to the experiment's incentives. Additionally, the time interval between the baseline and final stages for the control group is nearly zero, minimizing both the motivation and likelihood of control group members being exposed to external materials. Furthermore, in our robustness checks, we estimate the treatment effects for participants who are more likely to comply with the experimental intervention, and the results remain nearly identical to our baseline estimates.

18 Concerns may arise regarding the potential influence of the experimenter demand effect. That is, subjects may speculate about the experimenter's intentions and alter their responses accordingly to meet the experimenter's aims, which may distort the study result. It has been shown that the experimenter demand effect has no impact on the qualitative conclusions and that quantitative effects are small or nonexistent (De Quidt et al. 2018; De Quidt et al. 2019; Mummolo and Peterson 2019). An extended experiment conducted by Roth and Wohlfart (2020) shows that the subjects do not change their responses to meet the experiment organizers' needs, although the experiment organizers inform the subjects of their purpose.

19 Before the survey, some respondents may have read the MPIR for the third quarter of 2022, which would lead to a smaller change in their expectations induced by our information intervention. Therefore, the absolute value of this coefficient is the lower bound of the absolute value of the actual coefficient.

20 Roth and Wohlfart (2020) also find no difference in the impact of information intervention on the public's recession expectations across different financial literacy groups.

21 For example, the statement in the MPIR summary, “The economy rebounded remarkably, with the GDP recording year-on-year growth of 3.9 percent in Q3, an acceleration of 3.5 percentage points from Q2.” is relatively easy to understand.

22 In the questionnaires for the treatment group, we ask subjects about their level of trust in the information in China's MPIR summary they have just read (see Question A.13 in Appendix A of Supporting Information S1). 0 means no trust at all, and 10 means complete trust. The average trust level is 6.63, indicating that the subjects generally trust the information.

23 We consider that subjects in the treatment group who answer this question have likely read the intervention materials we provided. In this case, the treatment group consists of 388 participants, accounting for 87% of the treatment group in the baseline regression, which has 446 participants.

24 The result that an individual's income expectation will not change immediately does not imply that their permanent income will not increase. Unfortunately, our questionnaire only captures respondents' income expectations for the next 12 months, not their longer term or permanent income expectations. Consequently, we are unable to test whether central bank communication influences consumption plans through its impact on long-term or permanent income expectations.

Supporting Information

References

Armantier, O., S. Nelson, G. Topa, W. Van der Klaauw, and B. Zafar. 2016. “The Price Is Right: Updating Inflation Expectations in a Randomized Price Information Experiment.” Review of Economics and Statistics 98, no. 3: 503–523. https://doi.org/10.1162/REST_a_00499.
10.1162/REST_a_00499
Web of Science® Google Scholar
Bassetto, M. 2019. “Forward Guidance: Communication, Commitment, or Both?” Journal of Monetary Economics 108: 69–86. https://doi.org/10.1016/j.jmoneco.2019.08.015.
10.1016/j.jmoneco.2019.08.015
Web of Science® Google Scholar
Beutel, J., N. Metiu, and V. Stockerl. 2021. “Toothless Tiger With Claws? Financial Stability Communication, Expectations, and Risk-Taking.” Journal of Monetary Economics 120: 53–69. https://doi.org/10.1016/j.jmoneco.2021.03.003.
10.1016/j.jmoneco.2021.03.003
Web of Science® Google Scholar
Bholat, D., N. Broughton, J. Ter Meer, and E. Walczak. 2019. “Enhancing Central Bank Communications Using Simple and Relatable Information.” Journal of Monetary Economics 108: 1–15. https://doi.org/10.1016/j.jmoneco.2019.08.007.
10.1016/j.jmoneco.2019.08.007
Web of Science® Google Scholar
Blinder, A. S., M. Ehrmann, M. Fratzscher, J. De Haan, and D. J. Jansen. 2008. “Central Bank Communication and Monetary Policy: A Survey of Theory and Evidence.” Journal of Economic Literature 46, no. 4: 910–945. https://doi.org/10.1257/jel.46.4.910.
10.1257/jel.46.4.910
Web of Science® Google Scholar
Chen, Y., Y. Liu, and W. Yang. 2023. “Public Policy Evaluation and Randomized Controlled Trials: Implications and Prospects With China's Experience From the Perspective of Scalability.” Journal of Management World 39, no. 3: 172–196. https://doi.org/10.19744/j.cnki.11-1235/f.2023.0036.
10.19744/j.cnki.11-1235/f.2023.0036
Google Scholar
Chewning, Jr., E. G., and A. M. Harrell. 1990. “The Effect of Information Load on Decision Makers' Cue Utilization Levels and Decision Quality in a Financial Distress Decision Task.” Accounting, Organizations and Society 15, no. 6: 527–542. https://doi.org/10.1016/0361-3682(90)90033-Q.
10.1016/0361-3682(90)90033-Q
Web of Science® Google Scholar
Cieslak, A., and A. Schrimpf. 2019. “Non-Monetary News in Central Bank Communication.” Journal of International Economics 118: 293–315. https://doi.org/10.1016/j.jinteco.2019.01.012.
10.1016/j.jinteco.2019.01.012
Web of Science® Google Scholar
Cornand, C., and P. Hubert. 2020. “On the External Validity of Experimental Inflation Forecasts: A Comparison With Five Categories of Field Expectations.” Journal of Economic Dynamics and Control 110: 103746. https://doi.org/10.1016/j.jedc.2019.103746.
10.1016/j.jedc.2019.103746
Web of Science® Google Scholar
D'Acunto, F., D. Hoang, M. Paloviita, and M. Weber. 2020. “Effective Policy Communication: Targets Versus Instruments.” Becker Friedman Institute for Economics Working Paper No. 2020-148 (October). University of Chicago. https://doi.org/10.2139/ssrn.3712658.
10.2139/ssrn.3712658
Google Scholar
De Quidt, J., J. Haushofer, and C. Roth. 2018. “Measuring and Bounding Experimenter Demand.” American Economic Review 108, no. 11: 3266–3302. https://doi.org/10.1257/aer.20171330.
10.1257/aer.20171330
Web of Science® Google Scholar
De Quidt, J., L. Vesterlund, and A. J. Wilson. 2019. “ Experimenter Demand Effects.” In Handbook of Research Methods and Applications in Experimental Economics, edited by Arthur Schram and Aljaž. Ule, 384–400. Edward Elgar Publishing. https://doi.org/10.4337/9781788110563.
10.4337/9781788110563.00031
Google Scholar
Ehrmann, M., and J. Talmi. 2020. “Starting From a Blank Page? Semantic Similarity in Central Bank Communication and Market Volatility.” Journal of Monetary Economics 111: 48–62. https://doi.org/10.1016/j.jmoneco.2019.01.028.
10.1016/j.jmoneco.2019.01.028
Web of Science® Google Scholar
Enders, Z., F. Hünnekes, and G. J. Müller. 2019. “Monetary Policy Announcements and Expectations: Evidence From German Firms.” Journal of Monetary Economics 108: 45–63. https://doi.org/10.1016/j.jmoneco.2019.08.011.
10.1016/j.jmoneco.2019.08.011
Web of Science® Google Scholar
Eppler, M. J., and J. Mengis. 2004. “The Concept of Information Overload: A Review of Literature From Organization Science, Accounting, Marketing, MIS, and Related Disciplines.” Information Society 20, no. 5: 325–344. https://doi.org/10.1080/01972240490507974.
10.1080/01972240490507974
Web of Science® Google Scholar
Farber, H. S. 2011. “Job Loss in the Great Recession: Historical Perspective From the Displaced Workers Survey, 1984-2010.” NBER Working Paper No. 17040 (May). National Bureau of Economic Research. https://doi.org/10.3386/w17040.
10.3386/w17040
Google Scholar
Friedman, M. 1957. A Theory of the Consumption Function. Princeton University Press.
10.1515/9780691188485
Web of Science® Google Scholar
Friedman, M. 1961. “The Lag in Effect of Monetary Policy.” Journal of Political Economy 69, no. 5: 447–466. https://doi.org/10.1086/258537.
10.1086/258537
Web of Science® Google Scholar
Gardner, B., C. Scotti, and C. Vega. 2022. “Words Speak as Loudly as Actions: Central Bank Communication and the Response of Equity Prices to Macroeconomic Announcements.” Journal of Econometrics 231, no. 2: 387–409. https://doi.org/10.1016/j.jeconom.2021.07.014.
10.1016/j.jeconom.2021.07.014
Web of Science® Google Scholar
Guo, Y., W. Chen, and Y. Chen. 2016. “The Decreasing Effectiveness of China's Monetary Policy and Expectation Management.” Economic Research Journal 51, no. 1: 28–41+83.
Google Scholar
Guo, Y., H. Wang, and J. Guo. 2023. “Measurement of Monetary Policy Expectations and the Evaluation of the Central Bank Expectation Management Effectiveness.” Chinese Review of Financial Studies 15, no. 6: 47–71+123.
Google Scholar
Guo, Y., and X. Zhou. 2018. “Central Bank Communication, Adaptive Learning and Monetary Policy Effectiveness.” Economic Research Journal 53, no. 4: 77–91.
Google Scholar
Hansen, S., and M. McMahon. 2016. “Shocking Language: Understanding the Macroeconomic Effects of Central Bank Communication.” Journal of International Economics 99: S114–S133. https://doi.org/10.1016/j.jinteco.2015.12.008.
10.1016/j.jinteco.2015.12.008
Web of Science® Google Scholar
He, J., and Z. Zhao. 2022. “Forward Guidance, Expectation Management and Consumption Smoothing: A Study Based on Experimental Macroeconomics.” Nankai Economic Studies no. 9: 60–81. https://doi.org/10.14116/j.nkes.2022.09.004.
10.14116/j.nkes.2022.09.004
Google Scholar
Hong, Y., F. Liu, and J. Xue. 2023. “Measurement of Psychological Factors and Their Economic Impact in the Relationship Between Government and Market.” Journal of Management World 39, no. 3: 30–51. https://doi.org/10.19744/j.cnki.11-1235/f.2023.0044.
10.19744/j.cnki.11-1235/f.2023.0044
Google Scholar
Houthakker, H. S. 1958. “The Permanent Income Hypothesis.” American Economic Review 48, no. 3: 396–404. https://www.jstor.org/stable/1809775.
Web of Science® Google Scholar
Jessoe, K., and D. Rapson. 2014. “Knowledge Is (Less) Power: Experimental Evidence From Residential Energy Use.” American Economic Review 104, no. 4: 1417–1438. https://doi.org/10.1257/aer.104.4.1417.
10.1257/aer.104.4.1417
Web of Science® Google Scholar
Keller, K. L., and R. Staelin. 1987. “Effects of Quality and Quantity of Information on Decision Effectiveness.” Journal of Consumer Research 14, no. 2: 200–213. https://doi.org/10.1086/209106.
10.1086/209106
Web of Science® Google Scholar
Kling, J. R., J. B. Liebman, and L. F. Katz. 2007. “Experimental Analysis of Neighborhood Effects.” Econometrica 75, no. 1: 83–119. https://doi.org/10.1111/j.1468-0262.2007.00733.x.
10.1111/j.1468-0262.2007.00733.x
Web of Science® Google Scholar
Krueger, D., K. Mitman, and F. Perri. 2016. “On the Distribution of the Welfare Losses of Large Recessions.” NBER Working Paper No. 22458 (July). National Bureau of Economic Research. https://doi.org/10.3386/w22458.
10.3386/w22458
Google Scholar
Kryvtsov, O., and L. Petersen. 2021. “Central Bank Communication That Works: Lessons From Lab Experiments.” Journal of Monetary Economics 117: 760–780. https://doi.org/10.1016/j.jmoneco.2020.05.001.
10.1016/j.jmoneco.2020.05.001
Web of Science® Google Scholar
Lamla, M. J., and D. V. Vinogradov. 2019. “Central Bank Announcements: Big News for Little People?” Journal of Monetary Economics 108: 21–38. https://doi.org/10.1016/j.jmoneco.2019.08.014.
10.1016/j.jmoneco.2019.08.014
Web of Science® Google Scholar
Li, B., Y. Lei, and J. Wang. 2024. “Fed Monetary Policy Spillovers, Central Bank Expectation Management and Chinese Asset Prices.” Journal of World Economy 47, no. 01: 57–85. https://doi.org/10.19985/j.cnki.cassjwe.2024.01.008.
10.19985/j.cnki.cassjwe.2024.01.008
Google Scholar
Li, W., and W. Xu. 2022. “Policy Evaluation Research Based on Causal Inference: A Unified Framework for Experimental and Quasi-Experimental Designs.” Journal of Management World 38, no. 12: 104–123. https://doi.org/10.19744/j.cnki.11-1235/f.2022.0179.
10.19744/j.cnki.11-1235/f.2022.0179
Google Scholar
Lusardi, A. 1998. “On the Importance of the Precautionary Saving Motive.” American Economic Review 88, no. 2: 449–453. https://www.jstor.org/stable/116965.
Web of Science® Google Scholar
McMahon, M., A. Schipke, and X. Li. 2018. “China's Monetary Policy Communication: Frameworks, Impact, and Recommendations.” IMF Working Paper No. 2018/244 (November). International Monetary Fund.
Google Scholar
Merton, R. K. 1948. “The Self-Fulfilling Prophecy.” Antioch Review 8, no.2: 193–210. https://doi.org/10.2307/4609267.
10.2307/4609267
Google Scholar
Morris, S., and H. S. Shin. 2002. “Social Value of Public Information.” American Economic Review 92, no. 5: 1521–1534. https://doi.org/10.1257/000282802762024610.
10.1257/000282802762024610
Web of Science® Google Scholar
Mummolo, J., and E. Peterson. 2019. “Demand Effects in Survey Experiments: An Empirical Assessment.” American Political Science Review 113, no. 2: 517–529. https://doi.org/10.1017/S0003055418000837.
10.1017/S0003055418000837
Web of Science® Google Scholar
Nakamura, E., and J. Steinsson. 2018. “High-Frequency Identification of Monetary Non-Neutrality: The Information Effect.” Quarterly Journal of Economics 133, no. 3: 1283–1330. https://doi.org/10.1093/qje/qjy004.
10.1093/qje/qjy004
Web of Science® Google Scholar
Parker, J. A., and B. Preston. 2005. “Precautionary Saving and Consumption Fluctuations.” American Economic Review 95, no. 4: 1119–1143. https://doi.org/10.1257/0002828054825556.
10.1257/0002828054825556
Web of Science® Google Scholar
Roth, C., and J. Wohlfart. 2020. “How Do Expectations about the Macroeconomy Affect Personal Expectations and Behavior?” Review of Economics and Statistics 102, no. 4: 731–748. https://doi.org/10.1162/rest_a_00867.
10.1162/rest_a_00867
Web of Science® Google Scholar
Schneider, S. C. 1987. “Information Overload: Causes and Consequences.” Human Systems Management 7, no. 2: 143–153. https://doi.org/10.3233/HSM-1987-7207.
10.3233/HSM-1987-7207
Web of Science® Google Scholar
Wang, L., and X. Wang. 2015. “Central Bank Expectation Management, Inflation Volatility and Bank Risk-Taking.” Economic Research Journal 50, no. 10: 34–48.
Google Scholar
Wang, Y., G. Zhou, T. Wu, and C. Fan. 2019. “Central Bank's Communication, Monetary Policy Action and Corporate Investment Behavior.” China Industrial Economics no. 5: 118–135. https://doi.org/10.19581/j.cnki.ciejournal.2019.05.007.
10.19581/j.cnki.ciejournal.2019.05.007
Google Scholar
Wu, G., and F. Lian. 2016. “Research on China's Monetary Policy Transformation: An Exploration Based on the Mixed Rules of Quantity and Price.” Journal of World Economy 39, no. 3: 3–25. https://doi.org/10.19985/j.cnki.cassjwe.2016.03.002.
10.19985/j.cnki.cassjwe.2016.03.002
Google Scholar
Xiong, H., and Z. Wang. 2012. “Monetary Policy Surprise, Interest Rate Term Structure and Inflation Expectation Management.” Journal of World Economy 35, no. 6: 30–55. https://doi.org/10.19985/j.cnki.cassjwe.2012.06.003.
10.19985/j.cnki.cassjwe.2012.06.003
Google Scholar
Yagan, D. 2019. “Employment Hysteresis From the Great Recession.” Journal of Political Economy 127, no. 5: 2505–2558. https://doi.org/10.1086/701809.
10.1086/701809
Web of Science® Google Scholar
Yan, X., and W. Gao. 2017. “Central Bank Information Disclosure and Inflation Expectation Management: Construction of China's Central Bank Information Disclosure Index and Empirical Analysis.” Journal of Financial Research no. 8: 35–49.
Google Scholar
Zhu, N., Y. Xu, and G. Qiu. 2016. “The Influence of Central Bank Communication on RMB Exchange Rate Volatility.” Journal of Financial Research, no. 11: 32–46.
Google Scholar
Zou, W., X. Wang, and X. Xie. 2020. “Financial Market Response to Central Bank Communication: Event Study on Stock Market.” Journal of Financial Research no. 2: 34–50.
Google Scholar
Zwane, A. P., J. Zinman, E. Van Dusen, et al. 2011. “Being Surveyed Can Change Later Behavior and Related Parameter Estimates.” Proceedings of the National Academy of Sciences 108, no. 5: 1821–1826. https://doi.org/10.1073/pnas.1000776108.
10.1073/pnas.1000776108
CAS PubMed Web of Science® Google Scholar

Early View

Online Version of Record before inclusion in an issue

Can Central Bank Communication Guide Individuals' Expectations About the Macroeconomy? Evidence From a Randomized Information Experiment in China

ABSTRACT

1 Introduction

2 Literature Review