《财经研究》
2022第48卷第5期

Uncertainty and Macroeconomic Fluctuations: From the Perspectives of Precautionary Pricing and Unemployment Risk
Wang Bo, Xu Piaoyang
School of Finance，Nankai University，Tianjin 300350，China
Summary: Since the outbreak of COVID-19 epidemic, the global economy has fallen into recession and the unemployment rate of all countries has risen sharply. Although the epidemic has been effectively controlled in China, the epidemic abroad is still severe, and China’s economy still faces great uncertainties in the future. How to reduce the impact of the uncertainty caused by extreme events on China’s macro economy and promote the steady recovery of China’s economy under the condition of uncertainty has gradually become the focus of attention. This paper analyzes the impact of uncertainty on China’s macro economy by constructing TVP-SV-VAR, and then makes clear the influence mechanism of uncertainty on the macro economy by constructing the DSGE model. The results show that: Firstly, the impact of uncertainty will lead to inflation in the short term and deflation in the long term. Short-term inflation is mainly due to precautionary pricing, while long-term deflation is mainly due to the decline of effective demand caused by unemployment risk. Secondly, the consumption of residents with unemployment risk recovers slowly, while that of residents without unemployment risk recovers quickly. Thirdly, easy monetary stimulus can alleviate the negative impact of uncertainty on the economy more effectively, while supportive fiscal policy can alleviate the welfare loss among heterogeneous residents more accurately and effectively. The following policy suggestions are put forward: Firstly, government departments should take into account the impact of uncertainty caused by the epidemic. They should adhere to a positive and prudent monetary policy, strengthen communication with central banks, and stabilize the confidence of market subjects, so as to reduce the impact of uncertainty caused by the epidemic on the macro economy. Secondly, under the impact of uncertainty, due to the short-term inflation caused by precautionary pricing, monetary policy should reduce the reserve ratio and interest rate in a timely manner to slow down the cost of enterprises, stabilize the earnings of enterprises, and then reduce the incentive of enterprises to prevent pricing, so as to stabilize inflation. Thirdly, when dealing with the impact of uncertainty, fiscal policy should also play a role, especially through the targeted issuance of consumer vouchers to alleviate the budget constraints of middle and low income groups, stimulate consumer demand, and alleviate the consumption polarization under the impact of uncertainty. Fourthly, the government should further strengthen the construction of the labor market, improve the employment service system, and establish a reasonable unemployment insurance system, so as to shorten the time for residents to find a job and reduce the impact of the uncertainty of future income on resident consumption. The contributions of this paper are mainly reflected in the following aspects: Firstly, existing studies on China’s uncertainty believe that the economy will immediately enter a state of deflation after the impact of uncertainty; while this paper finds that the economy will briefly enter the state of inflation, and then enter the state of deflation. It explains this phenomenon by establishing a DSGE model. Secondly, existing studies do not clarify the role of pricing decision in the impact of uncertainty on China’s macro economy; while this paper clarifies the role of pricing decision in the impact of uncertainty on the macro economy, and further analyzes the pricing mechanism. Thirdly, this paper conducts a policy analysis based on a theoretical model containing unemployment risk and uncertainty impact, and finds that monetary policy can better mitigate the negative impact of uncertainty on the macro economy, and fiscal policy can better balance the welfare loss among different residents.
Key words: uncertainty    economic fluctuations    precautionary pricing    unemployment risk

Altig等（2020）以及Baker等（2020）从多个角度测度了新冠肺炎疫情所引起的不确定性，发现COVID-19所引起的不确定性将对经济造成极大的负面影响。在此之前，已有较多文献探讨了不确定性对宏观经济的影响及其机制。在不确定性会引起产出下降方面，现有文献已经达成共识，但对于不确定性的影响具体是通缩还是滞胀，却没有形成一致的结论。Leduc和Liu（2016）、Basu和Bundick（2017）以及许志伟和王文甫（2018）的研究表明，由于存在价格黏性，不确定性会导致市场不能及时出清，经济陷入通缩状态。而Mumtaz和Theodoridis（2015） 的跨国研究则表明，在遭遇不确定性冲击后，企业和工人为了抑制Calvo定价所带来的负面影响，会要求提高价格和工资，进而导致经济在短期内陷入滞胀状态。朱军和蔡恬恬（2018）研究发现，政策不确定性冲击会导致经济在短期内呈现滞胀状态，长期则呈现通缩状态。他们认为这主要是因为在短期内实物期权效应占据了主导，而长期则是预防性储蓄动机占据了主导。此外，Carriero等（2018）以及Katayama和Kim（2018）则认为，不确定性虽然对产出和就业具有显著的负向影响，但对通货膨胀率的影响却不显著。

 图 1 中国经济不确定性指数

 图 2 经济不确定性对宏观经济的影响

（一）居民部门

1. 工人。假设所有存在失业风险的工人在区间[0,1]上均匀分布，通过调整消费和储蓄来最大化总效用。工人家庭的目标函数为：

 $\begin{array}{c}{V}_{I,t}=\underset{{\left\{{a}_{I,t+1}(N),{c}_{I,t}(N)\right\}}_{N\in {Z}_+}}{\mathrm{max}}\left[\displaystyle\sum _{N\ge 0}{n}_{I,t}(N){u}_{I,t}(N)+{\beta }_{I}E\left({V}_{I,t+1}\right)\right]\end{array}$ (1)

 $\begin{array}{c}(1+{\tau }_{{c}_{I}}){c}_{I,t}(0)+{a}_{I,t}(0)\leqslant (1-{\tau }_{I,w}){w}_{t}+{A}_{t}{R}_{t-1}/[{n}_{t}(0){\pi }_{t}]\end{array}$ (2)

 $\begin{array}{c}{A}_{t}=(1-{s}_{t}){n}_{I,t-1}(0){a}_{I,t-1}(0)+{f}_{t}{\displaystyle\sum }_{N\geqslant 1}{a}_{I,t-1}(N){n}_{I,t-1}(N)\end{array}$ (3)

 $\begin{array}{c}(1+{\tau }_{{c}_{I}}){c}_{I,t}(N)+{a}_{I,t}(N)\leqslant {b}_{u}+{a}_{I,t-1}(N-1){R}_{t-1}/{\pi }_{t}\end{array}$ (4)

 $\begin{array}{c}{M}_{I,t,t+1}(0)={\beta }_{I}{E}_{t}\left[(1-{s}_{t+1}){u}_{c,I,t+1}^{\text{'}}(0)+{s}_{t+1}{u}_{c,I,t+1}^{\text{'}}(1)\right]/{u}_{c,I,t}(0)\end{array}$ (5)

 $\begin{array}{c}{M}_{I,t,t+1}(N)={\beta }_{I}{E}_{t}\left[(1-{f}_{t+1}){u}_{c,I,t+1}^{\text{'}}(N+1)+{f}_{t+1}{u}_{c,I,t+1}^{\text{'}}(0)\right]/{u}_{c,I,t}(N)\end{array}$ (6)

2. 企业家。假设所有企业家在区间[0,1]上均匀分布，通过调整消费和储蓄来最大化总效用。企业家的目标函数为：

 $\begin{array}{c}{V}_{P,t}=\underset{{\left\{{a}_{P,t+1},{c}_{P,t}\right\}}_{N\in {Z}_+}}{\mathrm{max}}\left[{u}_{P,t}+{\beta }_{P}E\left({V}_{P,t+1}\right)\right]\end{array}$ (7)

 $\begin{array}{c}(1+{\tau }_{{c}_{p}}){c}_{P,t}+{a}_{P,t+1}\leqslant (1-{\tau }_{P,w}){w}_{P,t}{n}_{P,t}+{a}_{P,t}{R}_{t-1}/{\pi }_{t}+{\mathrm{\Pi }}_{t}\end{array}$ (8)

 $\begin{array}{c}{M}_{P,t,t+1}={\beta }_{P}{E}_{t}({u}_{c,P,t+1}^{\text{'}}/{u}_{c,P,t}^{\text{'}})\end{array}$ (9)

（二）企业

1. 最终产品生产商。最终产品生产商主要负责对零售商生产的零售品 ${y}_{t}(j)$ 进行打包并出售，最终产品的生产函数为：

 $\begin{array}{c}{y}_{t}={[{\int }_{0}^{1}{{y}_{t}(j)}^{\frac{\varepsilon -1}{\varepsilon }}dj]}^{\frac{\varepsilon }{\varepsilon -1}}\end{array}$ (10)

 $\begin{array}{c}{y}_{t}(j)={[{P}_{t}(j)/{P}_{t}]}^{-\varepsilon }{y}_{t}\end{array}$ (11)

 $\begin{array}{c}{P}_{t}={[{\int }_{0}^{1}{{P}_{t}(j)}^{1-\varepsilon }dj]}^{\frac{1}{1-\varepsilon }}\end{array}$ (12)

2. 零售商。参照Calvo（1983），假定每一期零售商有 $1-\theta$ 的概率可以将价格调整为最优，以 $\theta$ 的概率价格保持不变。零售商选择最优价格来最大化利润：

 $\begin{array}{c}\mathrm{max}\;\;{E}_{t}\sum _{k=0}^{\mathrm{\infty }}{\theta }^{k}{M}_{P,t,t+k}\left[\dfrac{{P}_{t}^{*}(j)}{{P}_{t+k}}-\dfrac{{P}_{m,t+k}}{{P}_{t+k}}\right]{y}_{t+k}(j)\end{array}$ (13)

 $\begin{array}{c}{\mathrm{\Delta }}_{t}=(1-\theta ){({P}_{t}^{*}/{P}_{t})}^{-\varepsilon }+\theta {({P}_{t}/{P}_{t-1})}^{\varepsilon }{\mathrm{\Delta }}_{t-1}\end{array}$ (14)

3. 中间产品生产商。假设所有中间产品生产商在区间[0,1]上均匀分布，处于完全竞争市场，他们通过雇佣劳动来生产中间产品。中间产品的生产函数为：

 $\begin{array}{c}{y}_{m,t}={z}_{t}{\stackrel{~}{n}}_{t}\end{array}$ (15)

4. 就业中介。就业中介同时雇用工人和企业家，将其打包后为中间产品生产商提供劳动服务。假设每一期都有一个外生的概率 $\rho$ ，使上一期处于就业状态的居民失业。与此同时，就业中介会提供 ${\upsilon }_{t}$ 个工作岗位，假设每提供一个工作岗位所需付出的成本为 $\kappa$ 。就业中介每雇用一单位工人所需付出的工资为 ${w}_{t}$ ，则就业中介雇用工人的值函数 ${J}_{I,t}$ 为：

 $\begin{array}{c}{J}_{I,t}=(1-{\tau }_{F}){p}_{m,t}{z}_{t}-{w}_{t}+{E}_{t}\left[(1-\rho ){M}_{P,t,t+1}{J}_{I,t+1}\right]\end{array}$ (16)

 $\begin{array}{c}{J}_{P,t}=\psi (1-{\tau }_{F}){p}_{m,t}{z}_{t}-\psi {w}_{t}+{E}_{t}\left[(1-\rho ){M}_{P,t,t+1}{J}_{P,t+1}\right]\end{array}$ (17)

 $\begin{array}{c}{m}_{t}=\mu {u}_{t}^{\chi }{\upsilon }_{t}^{1-\chi }\end{array}$ (18)

 $\begin{array}{c}{w}_{t}={w}_{t-1}^{{\gamma }_{w}}{[w{({n}_{t}/n)}^{{\varphi }_{w}}]}^{1-{\gamma }_{w}}\end{array}$ (19)

（三）政府部门

 $\begin{array}{c}{c}_{G,t}+\Omega {b}_{u}(1-{n}_{t})+{a}_{G,t}{R}_{t-1}/{\pi }_{t}\leqslant {a}_{G,t+1}+{T}_{t}\end{array}$ (20)

${T}_{t}=\mathrm{\Omega }({\tau }_{I,w}{w}_{t}{n}_{I,t}+{\tau }_{{c}_{I}}{c}_{I,t})+(1-\mathrm{\Omega })({\tau }_{P,w}\psi {w}_{t}{n}_{P,t}+{\tau }_{{c}_{P}}{c}_{P,t})+{\tau }_{F}{p}_{m,t}{y}_{m,t}$

 $\begin{array}{c}{c}_{G,t}={c}_{G}{({a}_{G,t}/{a}_{G})}^{{\varphi }_{Gb}}{({y}_{t}/y)}^{{\varphi }_{Gy}}\end{array}$ (21)

 $\begin{array}{c}{R}_{t}=R{({\pi }_{t}/\pi )}^{{\varphi }_{\pi }}{({y}_{t}/y)}^{{\varphi }_{y}}{\xi }_{R,t}\end{array}$ (22)

（四）市场出清

 ${\stackrel{~}{n}}_{t}=\mathrm{\Omega }{n}_{I,t}+(1-\mathrm{\Omega })\psi {n}_{P,t}$

 $\begin{array}{c}\Omega {A}_{t}+(1-\mathrm{\Omega }){a}_{P,t}+{a}_{G,t}=0\end{array}$ (23)

 $\begin{array}{c}{\mathrm{\Delta }}_{t}{y}_{t}={y}_{m,t}-{\kappa }_{y}\end{array}$ (24)

 $\begin{array}{c}{c}_{t}+\kappa {\upsilon }_{t}={y}_{t}\end{array}$ (25)

 $\begin{array}{c}{c}_{t}=\Omega [{n}_{I,t}(0){c}_{I,t}(0)+{n}_{I,t}(1){c}_{I,t}(1)+{n}_{I,t}(N){c}_{I,t}(N)]+(1-\mathrm{\Omega }){c}_{P,t}\end{array}$ (26)

（一）居民部门的参数校准

（二）企业部门的参数校准

（三） 政府部门的参数校准

（四）其他参数校准

 图 3 同质性居民模型和异质性居民模型下的不确定性冲击

（一）企业预防性定价

 $\begin{array}{c}{mp}_{certainty}=[(1-\varepsilon ){({P}_{certainty}^{\mathrm{*}}/P)}^{1-\varepsilon }+\varepsilon {({P}_{certainty}^{\mathrm{*}}/P)}^{-\varepsilon }{P}_{m}/P]{y}_{m}\end{array}$ (27)

 $\begin{array}{c}{mp}_{uncertainty}=q[(1-\varepsilon ){({P}_{uncertainty}^{\mathrm{*}}/{P}^{l})}^{1-\varepsilon }+\varepsilon {({P}_{uncertainty}^{\mathrm{*}}/{P}^{l})}^{-\varepsilon }{P}_{m}/P]{y}_{m}\\ +(1-q)[(1-\varepsilon ){({P}_{uncertainty}^{\mathrm{*}}/{P}^{h})}^{1-\varepsilon }+\varepsilon {({P}_{uncertainty}^{\mathrm{*}}/{P}^{h})}^{-\varepsilon }{P}_{m}/P]{y}_{m}\end{array}$ (28)

 图 4 确定性稳态和不确定性稳态下企业边际利润与最优价格的关系图

 图 5 不同的价格黏性和价格加成下不确定性对宏观经济的影响

（二）工人失业风险

Basu和Bundick（2017）的研究表明失业率上升，会导致居民的预防性储蓄动机增强，这一点与中国遭遇新冠疫情冲击后，居民储蓄率上升相一致。但新冠疫情之后中国也出现了一些不一样的情况，根据贝恩对中国市场奢侈品的调查报告可以看出，疫情之后，中国市场对奢侈品的需求很快恢复，同时在全球奢侈品销售额下降的情况下，中国市场的奢侈品销售额实现了大幅上涨，而中国社会消费品零售总额复苏缓慢。表明中国在遭遇不确定性冲击后，居民的消费出现了一定的分化，以往的同质性居民模型在解释这一点上存在缺陷。故接下来我们将在异质性居民模型中讨论居民的失业风险。

 $\begin{array}{c}{M}_{I,t,t+1}(0)={\beta }_{I}{E}_{t}\left\{(1-{s}_{t+1}){u}_{c,I,t+1}^{\mathrm{\text{'}}}[{c}_{I,t+1}(0)]+{s}_{t+1}{u}_{c,I,t+1}^{\mathrm{\text{'}}}[{c}_{I,t+1}(1)]\right\}/{u}_{c,I,t}[{c}_{I,t+1}(0)]\end{array}$ (29)

 $\begin{array}{c}{M}_{P,t,t+1}={\beta }_{P}{E}_{t}（{u}_{c,P,t+1}^{\text{'}}/{u}_{c,P,t}^{\text{'}}）\end{array}$ (30)

 图 6 不同工人占比下不确定性对宏观经济的影响

 图 7 不同的财政及货币政策下不确定性对宏观经济的影响

 $\begin{array}{c}E\left[{u}_{I}({c}_{I,t})\right]={u}_{I}[(1+{\mathrm{\Delta }}_{I}){c}_{I}];E\left[{u}_{P}({c}_{P,t})\right]={u}_{P}[(1+{\mathrm{\Delta }}_{P}){c}_{P}]\end{array}$ (31)

 $\begin{array}{c}{L}_{wel}={E}_{0}\sum _{t=0}^{n}{\beta }_{p}^{t}\left({\widehat{\pi }}_{t}^{2}+{\lambda }_{wel}{\widehat{y}}_{t}^{2}\right)\end{array}$ (32)