The Inflation Outlook
At the School of International and Public Affairs, Columbia University, New York, New York
哥伦比亚大学国际与公共事务学院，纽约州纽约市
Thank you, Mr. Secretary.
谢谢你，国务卿先生。
I appreciate the opportunity to speak here today.
我很高兴有机会今天在这里发言。
1
1
When I offered my assessment of policy in my first speech as a Federal Reserve Governor in September, I assumed inflation for core goods and nonhousing core services would continue to run at current rates.
当我在九月份作为美联储理事的第一次演讲中提出政策评估时，我假设核心商品和非住房核心服务的通胀将继续以目前的速度运行。
2 While that was a useful working assumption at the time, today I will break down my inflation outlook in more detail.
2 虽然这在当时是一个有用的工作假设，但今天我将更详细地分析我的通胀前景。
Specifically, I'll share my interpretation of inflation's components and how this influences my perception of progress toward our 2 percent target.
具体来说，我将分享我对通货膨胀组成部分的解释，以及这如何影响我对实现 2% 目标的进展的看法。
Shelter Inflation Shelter costs are top of mind for households, and a large component of inflation indices.
住房通胀 住房成本是家庭最关心的问题，也是通胀指数的重要组成部分。
But calculating shelter inflation is far from simple.
但计算住房通胀远非简单。
For instance, the personal consumption expenditures (PCE) price index that the Fed targets includes housing costs for all households in the economy.
例如，美联储目标的个人消费支出（PCE）价格指数包括经济中所有家庭的住房成本。
That seems appropriate for capturing prices associated with all consumer spending but not ideal as a measure of current supply and demand pressures.
这似乎适合捕捉与所有消费者支出相关的价格，但作为衡量当前供需压力的指标并不理想。
Figure 1 makes that clear: As the economy emerged from the pandemic, demand for housing outstripped supply, and market rents for new tenants jumped.
图1清楚地表明了这一点：随着经济摆脱疫情影响，住房供不应求，新租户的市场租金大幅上涨。
But the PCE shelter index for all households lagged, as rents only reset when people move or renew their leases.
但所有家庭的 PCE 住房指数都滞后，因为租金只有在人们搬家或续租时才会重置。
PCE will always lag market rents.
PCE将始终落后于市场租金。
As recently as last year, the PCE shelter index still had catching up to do, creating uncertainty as to how long measured shelter inflation would remain elevated.
就在去年，PCE 住房指数仍有待追赶，这给测得的住房通胀将在多长时间内保持高位带来了不确定性。
This is no longer the case.
现在情况已不再如此。
Indeed, a separate all-tenant rent index—drawn from the same microdata as PCE shelter—caught up to new-tenant rents, while PCE shelter has actually overshot new rents.
事实上，一个单独的全租户租金指数（根据与 PCE 庇护所相同的微观数据得出）赶上了新租户的租金，而 PCE 庇护所实际上超过了新租户的租金。
3 Based on this catchup being complete, I expect a faster fall in PCE shelter inflation.
3 基于这一追赶的完成，我预计 PCE 住房通胀率将更快下降。
The current elevated readings for shelter inflation are an after-echo of previous, rather than current, supply–demand imbalances in the economy.
目前住房通胀数据的上升是之前而非当前经济供需失衡的后果。
For the past two years, we've seen extremely low increases in new-tenant rents.
在过去的两年里，我们看到新租户的租金增幅极低。
Measures by Cotality and Apartment List show that's persisted in recent quarters.
Cotality 和 Apartment List 的衡量标准显示，这种情况在最近几个季度持续存在。
I therefore expect a faster decline in PCE shelter inflation.
因此，我预计个人消费支出住房通胀将更快下降。
Two factors give me additional confidence: first, the negative population shock resulting from a reversal in net migration, and second, an elevated ratio of nominal shelter services consumption relative to overall consumption, which has historically been mean-reverting.
有两个因素给了我额外的信心：首先，净移民逆转造成的负面人口冲击；其次，名义住房服务消费相对于总体消费的比例升高，而总体消费历来都是均值回归的。
Core Nonhousing Services Inflation Another important part of household budgets is the cost of services besides housing, food, and energy.
核心非住房服务通货膨胀 家庭预算的另一个重要部分是除住房、食品和能源之外的服务成本。
These services are about half of household spending and include childcare, education, entertainment, and medical expenses.
这些服务约占家庭支出的一半，包括儿童保育、教育、娱乐和医疗费用。
Core nonhousing services inflation has moved sideways this year, somewhat above the average level observed from 2002 to 2007, a useful reference period when core PCE inflation averaged 2 percent.
今年核心非住房服务通胀横向移动，略高于 2002 年至 2007 年观察到的平均水平，这是一个有用的参考期，当时核心 PCE 通胀平均为 2%。
However, I am not worried about nonhousing services.
不过，我并不担心非住房服务。
Wages are the primary driver of service inflation.
工资是服务业通胀的主要驱动力。
Many service prices are driven by labor costs; for instance, labor compensation is about 60 percent of value added in leisure and hospitality and 83 percent in education and health services.
许多服务价格是由劳动力成本决定的；例如，休闲和酒店业的劳动报酬约占增加值的60%，教育和医疗服务业的劳动报酬约占83%。
Over the medium term, nonhousing services will follow labor market pressures.
从中期来看，非住房服务将跟随劳动力市场的压力。
Unemployment has trended higher for over two years, and wage growth lower.
两年多以来，失业率一直呈上升趋势，而工资增长则较低。
Additional measures of labor market tightness—such as job openings, consumer surveys about the difficulty of finding jobs, long-term layoffs, and the duration of unemployment—have suggested a loosening labor market for several years, tilting nominal wage growth risks toward the downside.
衡量劳动力市场紧张程度的其他指标——例如职位空缺、有关找工作难度的消费者调查、长期裁员和失业持续时间——表明劳动力市场将持续数年宽松，导致名义工资增长风险偏向下行。
A notable aspect of core services inflation is that not all components are measured from directly observed transactions.
核心服务膨胀的一个值得注意的方面是，并非所有组成部分都是通过直接观察到的交易来衡量的。
Many services components in the index are imputed in various ways and do not capture information about supply–demand imbalances that are relevant for monetary policy.
该指数中的许多服务成分以各种方式进行估算，并且没有捕获与货币政策相关的供需失衡信息。
A prime example is portfolio management services, which have contributed a quarter-point to core PCE inflation in the last year, well above their long-run average.
一个典型的例子是投资组合管理服务，该服务去年对核心个人消费支出通胀贡献了四分之一个百分点，远高于其长期平均水平。
The Bureau of Economic Analysis measures portfolio management costs based on overall revenues generated by asset managers.
经济分析局根据资产管理公司产生的总收入来衡量投资组合管理成本。
If the stock market goes up and boosts assets under management, revenues move in tandem since they are typically a fixed percentage fee of assets.
如果股市上涨并增加管理的资产，收入就会同步变化，因为它们通常是资产的固定百分比费用。
What ought to be recorded as an increased quantity of services consumed is instead recorded as increased prices.
本应记录为消费服务数量增加的内容却记录为价格上涨。
There may, for national accounting purposes, be a good reason for doing it this way.
出于国民核算的目的，这样做可能有充分的理由。
But, for monetary policy purposes, there is none, and it would be foolish of us to chase statistical quirks rather than focus on actual consumer prices.
但是，就货币政策而言，不存在这种情况，如果我们追求统计怪异而不关注实际的消费者价格，那将是愚蠢的。
Remarkably, long-term trends in the asset management industry point toward fee compression, indicating trend deflation .
值得注意的是，资产管理行业的长期趋势指向费用压缩，表明趋势通货紧缩。
Morningstar found the average expense ratio paid by investors fell nearly 6 percent in 2024.
晨星公司发现，2024 年投资者支付的平均费用率下降了近 6%。
4 By contrast, PCE recorded a roughly 20 percent increase in portfolio management fees in this period, contributing about 30 basis points to core PCE.
4 相比之下，在此期间，PCE 的投资组合管理费增长了约 20%，为核心 PCE 贡献了约 30 个基点。
If PCE had instead matched industry data with a 6 percent decline, core PCE would have been about 40 basis points lower than officially reported.
如果 PCE 与行业数据相符，下降 6%，那么核心 PCE 将比官方报告的低约 40 个基点。
Fees have consistently fallen each year for the past two decades, now less than half of what they were in 2005, as seen in figure 2 .
过去 20 年来，费用逐年下降，目前还不到 2005 年的一半，如图 2 所示。
5 The PCE portfolio management measure contains no signal regarding tightness in the asset management industry, let alone the economy.
5 PCE 投资组合管理指标没有显示资产管理行业紧缩的信号，更不用说经济了。
It contains no signal for the inflation forecast going forward.
它没有包含未来通胀预测的信号。
It should not be used to frame underlying inflation.
它不应该被用来构建潜在的通货膨胀。
Yet here we are, keeping interest rates too high because of the phantom inflation of portfolio advisory fees.
然而，由于投资组合咨询费的幻影通胀，我们将利率维持在过高水平。
6
6
For these reasons, I do not take much signal from nonmarket-based components.
出于这些原因，我没有从非市场成分中获得太多信号。
When looking at market-based services, inflation in this category has come down dramatically from its peak and moved sideways since mid-2024, at around 3 percent.
就基于市场的服务而言，该类别的通胀率已从峰值大幅下降，并自 2024 年中期以来横盘整理，约为 3%。
That is about 10 basis points below its average from 2002 to 2007, which, again, was a period in which PCE inflation was broadly at target.
这比 2002 年至 2007 年的平均水平低约 10 个基点，而 2007 年 PCE 通胀率仍基本达到目标。
Core nonhousing services inflation only remains elevated compared with this period because of nonmarket components.
由于非市场因素的影响，核心非住房服务通胀与同期相比仍保持较高水平。
Core Goods Inflation Turning to goods prices, they have been relatively less likely to rise than other index components since the turn of the century.
核心商品通胀 就商品价格而言，自世纪之交以来，与其他指数成分相比，它们上涨的可能性相对较小。
That trend was abruptly disrupted in the aftermath of the pandemic.
大流行之后，这种趋势突然被打乱。
After that, core goods inflation fell into negative territory for a brief period of around one year.
此后，核心商品通胀在一年左右的短暂时间内陷入负值。
Over the past 12 months, core goods inflation has picked up.
过去12个月，核心商品通胀有所回升。
The dominant narrative explaining this pattern blames U.S. tariff policy.
解释这种模式的主流说法归咎于美国的关税政策。
I am not convinced.
我不相信。
Consider the timing, as shown in figure 3 .
考虑时序，如图3所示。
In core goods PCE, it appears that the increase coincides with the imposition of tariffs during this year.
在核心商品个人消费支出中，增长似乎与今年征收关税同时发生。
However, in the consumer price index, the upturn began in the middle of 2024.
然而，消费者价格指数从2024年中期开始回升。
We don't target CPI, but it contains useful information.
我们不以 CPI 为目标，但它包含有用的信息。
The evidence that tariff policy neatly coincides with the increase in core goods prices is conflicted.
关税政策与核心商品价格上涨完全一致的证据是相互矛盾的。
When thinking through the effect of tariffs on prices, we must remind ourselves that the larger share of incidence or burden of any economic policy—tax, subsidy, or tariff—falls on the more inelastic or inflexible party.
在思考关税对价格的影响时，我们必须提醒自己，任何经济政策（税收、补贴或关税）的影响或负担的大部分都落在缺乏弹性或不灵活的一方身上。
This is because the more elastic or flexible party can change its behavior to avoid the burden of the tax.
这是因为更具弹性或灵活性的一方可以改变其行为以避免税收负担。
As an importer, the U.S. is relatively more elastic, since we can substitute demand across borders or manufacture goods at home.
作为进口国，美国相对更有弹性，因为我们可以跨境替代需求或在国内制造商品。
But producers' factories are installed capital that is stuck in place, and their labor is generally immobile and not easily substitutable across tradable and nontradable sectors.
但生产者工厂的固定资本是固定的，他们的劳动力通常是固定的，并且不易在可贸易和不可贸易部门之间替代。
As the largest trade deficit country, there are few substitutes, if any, for American demand, but many substitutes for potential supply.
作为最大的贸易逆差国，美国的需求几乎没有替代品（如果有的话），但潜在供应有很多替代品。
Product-level estimates of demand and supply elasticities from Anson Soderbery can be used to examine goods-level incidence.
Anson Soderbery 对需求和供应弹性的产品级估计可用于检查商品级发生率。
7 Figure 4 draws predicted isoincidence lines.
7 图 4 绘制了预测的等重合线。
The 45-degree line characterizes goods which have equal supply and demand elasticities, meaning incidence is evenly split between exporter and importer.
45 度线代表具有相等供给和需求弹性的商品，这意味着出口商和进口商之间的发生率是平分的。
The overwhelming majority of goods lie below the 45-degree line, meaning importers bear less burden.
绝大多数商品位于45度线以下，这意味着进口商承受的负担较小。
Figure 5 shows the cumulative distribution of import value based on the incidence absorbed by exporters.
图 5 显示了基于出口商吸收的进口值的累积分布。
These elasticities imply that, for about 70 percent of import goods value, the exporter bears at least 70 percent of the incidence, and for about half of goods value, the exporter bears at least 80 percent of the incidence.
这些弹性意味着，对于约 70% 的进口商品价值，出口商承担至少 70% 的影响，对于约一半的商品价值，出口商承担至少 80% 的影响。
8
8
Of course, there are goods for which importers bear material incidence, and that will mean relative price changes—but whether they prove to be relevant for aggregate prices depends on other considerations.
当然，有些商品进口商要承担重大责任，这将意味着相对价格变化——但它们是否被证明与总价格相关取决于其他考虑因素。
Given the Soderbery (2018) elasticities and a split of the domestic incidence between wholesale importers and consumers, it is plausible that the ultimate effect of tariffs on consumer price levels will be in the neighborhood of two-tenths of a percent—noise.
考虑到 Soderbery（2018）的弹性以及批发进口商和消费者之间国内发生率的划分，关税对消费者价格水平的最终影响可能约为十分之二——噪音。
9
9
By contrast, many analyses for what tariffs would do to inflation were based on empirical studies of the U.S.–China trade conflict before the pandemic.
相比之下，许多关于关税对通胀影响的分析都是基于对大流行前中美贸易冲突的实证研究。
However, these studies suffer from bias from trade rerouting and de minimis exemptions.
然而，这些研究因贸易重新安排和最低限度豁免而存在偏差。
If a Chinese exporter can avoid bearing incidence by costly transshipment or exploitation of the de minimis loophole, the incentive to do so will be highly correlated with the incidence absorbed.
如果中国出口商可以通过昂贵的转运或利用微量漏洞来避免承担事故，那么这样做的动机将与吸收的事故高度相关。
In other words, goods for which China would have to absorb a greater share of the tariff cost drop from the sample, while goods for which China could push through tariff costs to importers remain in the sample, violating identification restrictions.
换句话说，中国必须承担更大份额关税成本的商品从样本中下降，而中国可以将关税成本转嫁给进口商的商品仍保留在样本中，这违反了识别限制。
Jackson Mejia has studied this bias in pass-through estimates and finds evidence that up to 40 percent of product categories exposed to tariffs are affected by transshipment, with trade volumes approaching 25 percent for intermediate and capital goods.
Jackson Mejia 研究了传递估算中的这种偏差，并发现证据表明，高达 40% 的关税产品类别受到转运的影响，中间产品和资本货物的贸易量接近 25%。
10 The de minimis exemption poses similar identification problems, meaning the transshipment effects are lower bounds on the identification problem in this literature.
10 最低限度豁免提出了类似的识别问题，这意味着转运效应是本文献中识别问题的下限。
11 Additionally, these studies are short term in nature, and incidence analyses are usually long term; a difference in short-run and long-run elasticities is not controversial.
11 此外，这些研究本质上是短期的，而发病率分析通常是长期的；短期和长期弹性的差异是没有争议的。
Moreover, any claim that tariffs are a current driver of inflation must be accompanied by a clear description of the counterfactual.
此外，任何声称关税是当前通胀推动因素的说法都必须附有对反事实的明确描述。
Several analyses use the two decades before the pandemic as the relevant counterfactual; I do not find these periods to be reasonable given large concurrent shocks that pollute the analysis.
一些分析使用大流行前的二十年作为相关的反事实；考虑到同时发生的巨大冲击会污染分析，我认为这些时期并不合理。
Attributing all excess inflation over the pre-trend to tariffs is something that could be done for many recent policies or shocks, such as an emphasis on supply chain resilience after the pandemic unrelated to tariffs.
对于许多最近的政策或冲击，可以将所有超额通胀归因于关税，例如在与关税无关的大流行之后强调供应链的弹性。
One counterfactual I think makes sense to use compares import-intensive core goods versus overall core goods, as shown in figure 6 .
我认为使用一个有意义的反事实将进口密集型核心商品与总体核心商品进行比较，如图 6 所示。
12 If tariffs were the driver of recent inflation, then one would expect import-intensive core goods to see substantially more inflation.
12 如果关税是近期通胀的驱动因素，那么人们预计进口密集型核心商品的通胀将大幅加剧。
In fact, total core goods prices have risen at approximately the same rate as import-intensive goods since the end of last year.
事实上，自去年底以来，核心商品总价格的上涨速度与进口密集型商品的上涨速度大致相同。
I also compare the level of inflation in the U.S. to that of other industrialized countries, displayed in figure 7 .
我还将美国的通货膨胀水平与其他工业化国家的通货膨胀水平进行了比较，如图 7 所示。
If U.S. tariff policy were a material driver of domestic inflation, then U.S. goods inflation would visibly stick out from major trading partners.
如果美国关税政策是国内通胀的实质性驱动因素，那么美国商品通胀将明显受到主要贸易伙伴的影响。
Yet U.S. inflation is comparable to that in Canada and the U.K., slightly above that in the European Union, and below that in Mexico; the U.S. doesn't stand out in any direction.
然而，美国的通胀率与加拿大和英国相当，略高于欧盟，低于墨西哥；美国在任何方面都不突出。
Neither counterfactual suggests tariffs are the primary driver of core goods inflation.
这两种反事实都表明关税是核心商品通胀的主要驱动因素。
Even if one believes that tariffs are driving goods inflation, the standard practice is for central bankers is to "look through" a transient shock, as a one-time increase in the price level differs from a persistent shift to inflation.
即使人们认为关税正在推动商品通胀，央行行长的标准做法也是“看穿”短暂的冲击，因为价格水平的一次性上涨与持续转向通胀不同。
This mirrors how value-added taxes are treated by central banks.
这反映了央行对待增值税的方式。
Monetary policy achieves stable prices through balancing aggregate supply and demand; changes in prices stemming from relative tax adjustments—incidence aside—are not indicative of supply–demand imbalances.
货币政策通过平衡总供给和总需求实现物价稳定；相对税收调整引起的价格变化（不考虑发生率）并不表明供需失衡。
While the burden from tariffs will ultimately be borne by the exporters, this analysis is drawn from long-term elasticities.
虽然关税负担最终将由出口商承担，但这一分析是根据长期弹性得出的。
In the short run, it may take time for threats of moving supply chains to become credible, creating a lag before exporters reduce selling prices.
从短期来看，转移供应链的威胁可能需要一段时间才能变得可信，从而在出口商降低售价之前产生滞后。
Although I do not yet see meaningful tariff-driven inflation, it may materialize.
尽管我还没有看到有意义的关税驱动的通胀，但它可能会成为现实。
But over time, short-run elasticities would converge with long-run elasticities.
但随着时间的推移，短期弹性将与长期弹性趋同。
Not only would the increase in inflation be transitory, but likely so would the increase in the price level, meaning subsequent offsetting deflation .
通货膨胀的上升不仅是暂时的，而且物价水平的上升也可能是暂时的，这意味着随后的通货紧缩会被抵消。
13
13
If tariffs are not the likely cause of the recent rise in goods inflation, then what is?
如果关税不是近期商品通胀上升的可能原因，那么什么才是呢？
One possibility is that the rise in prices is noise in a volatile series, though this is not a satisfying answer.
一种可能性是价格上涨是波动序列中的噪音，尽管这不是一个令人满意的答案。
A second possibility is that the U.S. and other advanced foreign economies are still following post-pandemic bullwhips of gradual oscillation around a lower mean level of core goods inflation.
第二种可能性是，美国和其他发达外国经济体仍在追随疫情后围绕较低的核心商品通胀平均水平逐渐振荡的趋势。
Under that scenario, last year's core goods inflation was "too low" and now it is "too high"; perhaps next year it will be "too low" again.
在这种情况下，去年的核心商品通胀“太低”，而现在却“太高”；也许明年又会“太低”了。
A third, unsavory, possibility is that goods inflation is settling in at a higher level than prevailed immediately pre-pandemic.
第三种令人讨厌的可能性是，商品通胀正处于比大流行前更高的水平。
After all, there were previous regimes with stable but higher core goods inflation.
毕竟，以前的政权有稳定但较高的核心商品通胀。
Higher global goods inflation could result from a longer-term trend of trade restructuring that encompasses much more than tariffs, including a reduced willingness to rely on unfettered access to exports in favor of national security and geoeconomic concerns.
全球商品通胀上升可能是由贸易结构调整的长期趋势造成的，这种结构不仅仅包括关税，还包括出于国家安全和地缘经济考虑而减少依赖不受限制的出口准入的意愿。
Attention to supply chain security and resilience—which predates this year's tariffs—may mean higher core goods inflation for a longer period, though perhaps not this high.
对供应链安全和弹性的关注（早于今年的关税）可能意味着核心商品通胀将在较长时期内保持较高水平，尽管可能不会这么高。
I accept I don't know what's driving higher goods inflation currently.
我承认我不知道目前商品通胀上升的原因是什么。
As former Bank of England Governor, Mervyn King, and economist, John Kay, urged in Radical Uncertainty , pretending we have more knowledge than possible will stymie our understanding of reality.
正如前英格兰银行行长默文·金 (Mervyn King) 和经济学家约翰·凯 (John Kay) 在《激进的不确定性》一书中所敦促的那样，假装我们拥有比可能更多的知识将阻碍我们对现实的理解。
14 Shedding that hubris will help us reach a clearer understanding of the dynamics and make better decisions.
14 摆脱傲慢态度将有助于我们更清楚地了解动态并做出更好的决策。
I do see reasons to be optimistic about goods inflation.
我确实认为有理由对商品通胀持乐观态度。
One is that the rapid deregulatory push to expand supply will, all else equal, decrease prices.
一是，在其他条件相同的情况下，迅速放松管制以扩大供应将降低价格。
Another is that work by Bloomberg Economics using AI to analyze thousands of public company earnings calls indicates price pressures in measured goods inflation will begin to decline within two quarters.
另一个原因是，彭博经济研究使用人工智能分析了数千家上市公司的财报电话会议，表明衡量的商品通胀的价格压力将在两个季度内开始下降。
15
15
Policy The lack of a clear downward forecast for core goods prices might suggest keeping interest rates elevated.
政策 由于缺乏对核心商品价格的明确下调预测，可能建议维持高利率。
However, I expect more than enough disinflation from housing services to counterbalance that possibility.
然而，我预计住房服务的通货紧缩足以抵消这种可能性。
Core goods represents just 25 percent of the core PCE index.
核心商品仅占核心PCE指数的25%。
Indeed, my earlier forecasts were conditioned on core goods inflation staying at this rate for longer.
事实上，我之前的预测是以核心商品通胀率长期保持在这一水平为条件的。
The risks to my forecast are if shelter inflation picks up again, or if core goods inflation remains well above 2 percent.
我的预测面临的风险是住房通胀是否再次上升，或者核心商品通胀是否仍远高于 2%。
I judge those cases to be unlikely.
我认为这些情况不太可能发生。
If I am right about shelter but wrong about tariffs, we will undershoot our inflation target.
如果我对住房的看法是正确的，但对关税的看法是错误的，那么我们将低于我们的通胀目标。
To summarize, we must be thoughtful in considering genuine underlying inflationary pressures.
总而言之，我们必须深思熟虑地考虑真正的潜在通胀压力。
Excess measured inflation is unreflective of current supply–demand dynamics.
过度的通货膨胀并不能反映当前的供需动态。
Shelter inflation is indicative of a supply–demand imbalance that occurred as much as two to four years ago, not today.
住房通胀表明供需不平衡的情况早在两到四年前就发生过，而不是现在。
Given monetary policy lags, we need to make policy for 2027, not 2022.
鉴于货币政策的滞后性，我们需要制定2027年的政策，而不是2022年的政策。
A better measure of underlying inflation would account for distortions from shelter and imputed prices.
更好地衡量潜在通胀可以考虑住房和估算价格造成的扭曲。
Removing imputed phantom inflation like portfolio management, market-based core inflation is running below 2.6 percent, as seen in figure 8 .
除去投资组合管理等估算的虚拟通胀，基于市场的核心通胀率低于 2.6%，如图 8 所示。
If we further remove housing and look at market-based core ex shelter, underlying inflation is running below 2.3 percent, within noise of our target.
如果我们进一步取消住房并关注基于市场的核心前住房，潜在通胀率将低于 2.3%，在我们的目标范围内。
Once shelter inflation has normalized from the anomalous post-pandemic experience, ordinary market-based core may be more appropriate.
一旦住房通胀从大流行后的异常经历中恢复正常，普通的基于市场的核心可能更合适。
Some might accuse me of cherry-picking a preferred measure, but my gauge of underlying inflation excludes less of the index than some other measures.
有些人可能会指责我挑选了一个首选指标，但我对潜在通胀的衡量标准排除的指数少于其他指标。
It is also easy to understand and has a straightforward rationale for discarding certain slices of the index.
它也很容易理解，并且对于丢弃索引的某些部分有一个简单的理由。
Market-based core includes 75 percent of overall PCE, and market-based core ex housing includes 60 percent; core services ex housing is only 51 percent of PCE and discards goods, which are salient.
市场化核心占整体PCE的75%，市场化核心除住房占60%；除住房外的核心服务仅占 PCE 的 51%，而废弃商品则很突出。
Moreover, James Stock and Mark Watson find that market-based prices are more tightly related to cyclical measures of the economy than poorly measured components like imputed prices.
此外，詹姆斯·斯托克（James Stock）和马克·沃森（Mark Watson）发现，与估算价格等衡量不善的组成部分相比，基于市场的价格与经济周期性指标的关系更为紧密。
16
16
Keeping policy unnecessarily tight because of an imbalance from 2022, or because of artifacts of the statistical measurement process, will lead to job losses.
由于 2022 年起的失衡或统计测量过程的人为因素而不必要地收紧政策，将导致失业。
There was a large bout of inflation that resulted in an increase in prices after the pandemic.
疫情过后，出现了大规模的通货膨胀，导致物价上涨。
While American families are still rightly distraught with that experience and unhappy with affordability, prices are now once again stable, albeit at higher levels.
尽管美国家庭仍然对这种经历感到心烦意乱，并对负担能力感到不满，但价格现在再次稳定，尽管处于较高水平。
Policy should reflect that.
政策应该反映这一点。
Fortunately, the shelter outlook appears relatively clear—because market rents lead measured inflation—and powerful enough to overwhelm even the possibility of sustained higher goods inflation.
幸运的是，住房前景似乎相对清晰——因为市场租金引领可衡量的通胀——并且强大到足以压倒商品通胀持续走高的可能性。
Underlying inflation is near, and further approaching, our target.
基本通胀已接近并进一步接近我们的目标。
On the other side of our mandate, experience suggests that labor market deterioration can occur quickly and nonlinearly and be difficult to reverse.
另一方面，经验表明，劳动力市场的恶化可能会迅速、非线性地发生，并且难以逆转。
In part because monetary policy lags several quarters, a quicker pace of easing policy—as I have advocated—would appropriately move us closer to a neutral stance.
部分原因是货币政策滞后几个季度，加快宽松政策的步伐——正如我所主张的——将适当地让我们更接近中性立场。
Recessions are an inevitable part of the business cycle, and at some point, we will suffer one.
经济衰退是商业周期不可避免的一部分，在某些时候，我们会遭受经济衰退。
We should strive to ensure that point is as far in the future and as shallow as possible by appropriately calibrating monetary policy.
我们应该通过适当调整货币政策，努力确保这一点尽可能在未来、尽可能浅。
1.
1.
The views expressed here are my own and are not necessarily those of my colleagues on the Federal Reserve Board or the Federal Open Market Committee.
这里表达的观点是我自己的观点，不一定是我在美联储委员会或联邦公开市场委员会的同事的观点。
Return to text
返回正文
2.
2.
Miran, Stephen I.
米兰，斯蒂芬一世。
(2025), "Nonmonetary Forces and Appropriate Monetary Policy," speech delivered at the Economic Club of New York, New York, N.Y., September 22.
(2025)，“非货币力量和适当的货币政策”，9 月 22 日在纽约州纽约经济俱乐部发表的演讲。
Return to text
返回正文
3.
3.
The main differences are a set of sample restrictions; for example, PCE shelter imputes many missing observations that are dropped from the other two indices.
主要区别在于一组样本限制；例如，PCE 庇护所估算了从其他两个指数中删除的许多缺失观测值。
A comparison of the different indices is available from the Bureau of Labor Statistics (CPI rents flow into PCE rents) at https://www.bls.gov/cpi/research-series/r-cpi-ntr.htm Return to text
不同指数的比较可从劳工统计局获取（CPI 租金流入 PCE 租金）：https://www.bls.gov/cpi/research-series/r-cpi-ntr.htm 返回文本
4.
4.
Morningstar Manager Research (2024), "How Fund Fees are Evolving in the US," July 16 Insights (blog), (updated October 8), https://www.morningstar.com/business/insights/blog/funds/us-fund-fee-study. Return to text
晨星经理研究 (2024)，“美国基金费用如何演变”，7 月 16 日洞察（博客），（10 月 8 日更新），https://www.morningstar.com/business/insights/blog/funds/us-fund-fee-study。返回正文
6.
6.
Monetary policy should not mechanically respond to asset prices vis-à-vis the inflation channel.
货币政策不应机械地对资产价格相对于通胀通道做出反应。
Financial markets reflect a host of factors—including technology, tax and regulatory policy, population, trade, and many other things—which may have separate or joint implications for neutral rates, the output gap and inflation but which require nuanced analysis and interpretation.
金融市场反映了许多因素，包括技术、税收和监管政策、人口、贸易和许多其他因素，这些因素可能对中性利率、产出缺口和通货膨胀产生单独或共同的影响，但需要细致入微的分析和解释。
Return to text
返回正文
7.
7.
Anson Soderberry (2018), "Trade Elasticities, Heterogeneity, and Optimal Tariffs," Journal of International Economics , vol. 114 (September), pp.
Anson Soderberry (2018)，“贸易弹性、异质性和最优关税”，《国际经济学杂志》，卷。 114（九月），第 114 页。
44–62.
44–62。
Return to text
返回正文
8.
8.
The weighted mean incidence drawn from the Soderbery (2018) elasticities is 70 percent borne by the exporter.
根据 Soderbery (2018) 弹性得出的加权平均发生率由出口商承担 70%。
This number is not far from what would be implied by using an aggregate demand elasticity of negative 3, in line with the trade literature, and a supply elasticity of 1.
这个数字与使用负 3 的总需求弹性（与贸易文献一致）和 1 的供给弹性所暗示的数字相差不远。
Unfortunately, much of the literature question begs large supply elasticities as a result of Dixit-Stiglitz style production characterized by constant markups, constant marginal cost, and free entry.
不幸的是，许多文献问题都回避了巨大的供给弹性，这是迪克西特-斯蒂格利茨式生产的结果，其特点是不断加价、恒定边际成本和免费进入。
Papers that move away from these assumptions find lower supply elasticities or pass-through of marginal cost shocks into prices; for instance, see Amiti, Itskhoki, and Konings (2019) without constant markups (Mary Amiti, Oleg Itskhoki, and Jozef Konings (2019), "International Shock, Variable Markups, and Domestic Prices," The Review of Economic Studies , vol. 86 (6), pp.
偏离这些假设的论文发现供给弹性较低或边际成本冲击传导至价格；例如，参见 Amiti、Itskhoki 和 Konings（2019 年），没有恒定加价（Mary Amiti、Oleg Itskhoki 和 Jozef Konings（2019 年），“国际冲击、可变加价和国内价格”，《经济研究评论》，第 86 卷（6），第 175 页）。
2356–402); Bergstrand, Cray and Gervais (2023) without constant marginal costs (Jeffrey H.
2356–402）； Bergstrand、Cray 和 Gervais (2023)，没有恒定的边际成本 (Jeffrey H.
Bergstrand, Stephen R.
伯格斯特兰德，斯蒂芬·R.
Cray, and Antoine Gervais (2023) "Increasing Marginal Costs, Firm Heterogeneity, and the Gains from 'Deep' International Trade Agreements," Journal of International Economics , vol. 144 (September)); or Alessandria, Choi and Ruhl (2021) without free entry (George Alessandria, Horag Choi, and Kim J.
Cray 和 Antoine Gervais (2023)“不断增加的边际成本、企业异质性以及‘深度’国际贸易协定的收益”，《国际经济学杂志》，第 1 卷。 144（九月））；或 Alessandria、Choi 和 Ruhl (2021)，无需免费入场（George Alessandria、Horag Choi 和 Kim J.
Ruhl (2021), "Trade Adjustment Dynamics and the Welfare Gains from Trade," Journal of International Economics , vol. 131 (July)).Therefore, I prefer to use the Soderbery (2018) estimates, which don't require these assumptions.
Ruhl（2021），“贸易调整动态和贸易带来的福利收益”，《国际经济学杂志》，卷。 131（7 月））。因此，我更喜欢使用 Soderbery（2018）的估计，它不需要这些假设。
Subsequent work by Soderbery, with Farid Farrokhi, employs some of these micro-founding assumptions on production, which, like in other papers, lead to large elasticities of supply.
索德伯里与法里德·法罗基 (Farid Farrokhi) 的后续工作采用了一些关于生产的微观基础假设，与其他论文一样，这些假设导致了供应的巨大弹性。
See Farid Farrokhi and Anson Soderbery (2024), "Trade Elasticities in General Equilibrium: Demand, Supply, and Aggregation," The Review of Economics and Statistics .
请参阅 Farid Farrokhi 和 Anson Soderbery (2024)，“一般均衡中的贸易弹性：需求、供给和聚合”，《经济学与统计评论》。
Working outside the trade general equilibrium literature, Horioka and Ford (2025) model capital mobility without comparative advantage, and if factor shares in this model are adjusted to reflect higher capital shares in surplus countries than deficit countries, incidence largely falls on the tariffed nation.
在贸易一般均衡文献之外，Horioka 和 Ford（2025）对没有比较优势的资本流动性进行了建模，如果对该模型中的要素份额进行调整以反映顺差国家比赤字国家更高的资本份额，那么影响力主要落在关税国家身上。
See Charles Yuji Horioka and Nicholas Ford (2025), "A New Modeling Approach to Help Address the Trump Tariffs," (PDF) Discussion Paper Series DP2025-31 (Japan: Kobe University, December).
请参阅 Charles Yuji Horioka 和 Nicholas Ford (2025)，“帮助解决特朗普关税问题的新建模方法”(PDF) 讨论论文系列 DP2025-31（日本：神户大学，12 月）。
Moreover, installed capital is usually the least elastic factor of production, and, with imperfect substitutability between tradable and nontradable labor, there is a real sense in which labor in those factories is similarly installed; welders do not easily become hairdressers.
此外，安装资本通常是弹性最小的生产要素，并且由于可贸易劳动力和不可贸易劳动力之间的不完全替代性，这些工厂中的劳动力实际上是类似安装的；焊工不容易成为理发师。
It is exceedingly strange that much of the literature on trade and tariff incidence neglects to study not only capital altogether but installed capital in particular, which, in many public finance settings, can serve as an incidence sink.
极其奇怪的是，许多关于贸易和关税发生率的文献不仅忽视了对资本的研究，而且特别忽视了安装资本，在许多公共财政环境中，安装资本可以充当发生率汇。
Return to text
返回正文
9.
9.
The calculation is the product of a 12.3 percent change in tariffs, a 30 percent domestic share of incidence, a 50 percent split between wholesale importers and consumers, and a roughly 10 percent import share of overall PCE.
该计算结果是关税变化 12.3%、国内发病率 30%、批发进口商和消费者之间比例 50%、以及个人消费支出总量中约 10% 的进口份额的结果。
If wholesale importers take all the tariff incidence on margin, the effect on consumer prices will be zero.
如果批发进口商将所有关税归入保证金，那么对消费者价格的影响将为零。
If they take none of it on margin, and are able to pass through the entirety of the incidence to consumers, the effect will be about 0.4 percent.
如果他们不采取任何边际措施，并且能够将整个事件转嫁给消费者，那么效果将约为 0.4%。
Two-tenths is the midpoint.
十分之二是中点。
Moreover, standard incidence calculations require competitive markets.
此外，标准发病率计算需要竞争性市场。
To account for potential imperfect competition, a broader set of parameters may be necessary, as in Weyl and Fabinger (2013), but identification would tax an already burdened empirical trade literature beyond what I fear it might bear.
为了解释潜在的不完全竞争，可能需要一套更广泛的参数，正如 Weyl 和 Fabinger (2013) 所指出的那样，但识别会给已经负担过重的实证贸易文献带来负担，超出我担心的负担。
E.
E.
Glen Weyl and Michal Fabinger (2013), "Pass-Through as an Economic Tool: Principles of Incidence under Imperfect Competition," Journal of Political Economy , vol. 121 (June), pp.437–641.
Glen Weyl 和 Michal Fabinger (2013)，“作为经济工具的传递：不完全竞争下的发生原理”，《政治经济学杂志》，卷。 121（六月），第 437–641 页。
Return to text
返回正文
10.
10.
Jackson Mejia (2025), "Selection into Tariff Avoidance and the Measured Welfare Cost of Tariffs," working paper, Massachusetts Institute of Technology, October, available at http://dx.doi.org/10.2139/ssrn.5653130 .
Jackson Mejia (2025)，“选择关税规避和可衡量的关税福利成本”，工作论文，麻省理工学院，10 月，可查阅 http://dx.doi.org/10.2139/ssrn.5653130 。
Return to text
返回正文
11.
11.
Though the exemption was eliminated in 2025, the bias will infect studies drawn from earlier data.
尽管豁免在 2025 年被取消，但这种偏见仍会影响根据早期数据得出的研究。
Return to text
返回正文
12.
12.
Federal Reserve Board staff replicated and extended a methodology described in The Council of Economic Advisers (2025), Imported Goods Have Been Getting Cheaper Relative to Domestically Produced Goods (PDF) (Washington: Council of Economic Advisers).
美联储工作人员复制并扩展了经济顾问委员会（2025）《进口商品相对于国内生产的商品越来越便宜》（PDF）（华盛顿：经济顾问委员会）中描述的方法。
On Figure 6, the blue import content measure uses import content data from 2019 and the red import content measure is more disaggregated and comes from recently released 2023 data.
在图 6 中，蓝色进口内容衡量标准使用 2019 年的进口内容数据，红色进口内容衡量标准更加细分，来自最近发布的 2023 年数据。
I included both aggregations to preserve comparability with the earlier CEA study which was produced when only the 2019 aggregation was available.
我纳入了这两种汇总，以保持与早期 CEA 研究的可比性，该研究是在只有 2019 年汇总可用时生成的。
Return to text
返回正文
13.
13.
For an example of recent research that uses dynamic elasticities to show that real wages can first decline and then go up (similar to inflation going up then down), albeit through a different channel than I'm describing, see Alessandria, George A., Jiaxiaomei Ding, Shafaat Y.
最近的研究使用动态弹性来表明实际工资可以先下降然后上升（类似于通货膨胀先上升然后下降），尽管通过的渠道与我所描述的不同，请参阅 Alessandria、George A.、Jiaxiaomei Ding、Shafaat Y.
Khan, and Carter B.
可汗和卡特 B.
Mix (2025), "The Tariff Tax Cut: Tariffs as Revenue," NBER Working Paper Series 33784 (Cambridge, Mass.: National Bureau of Economic Research, May).
Mix (2025)，“关税减税：关税作为收入”，NBER 工作论文系列 33784（马萨诸塞州剑桥：国家经济研究局，5 月）。
Return to text
返回正文
14.
14.
John Kay and Mervyn King (2020), Radical Uncertainty (New York: W.
约翰·凯和默文·金 (2020)，《激进的不确定性》（纽约： W.
W.
W.
Norton & Company) Return to text
诺顿公司）返回正文
15.
15.
Anna Wong, Joshua Danial, and Alex Tanzi (2025), "17,500 Earning Calls Greenlight Big Fed Cuts in 2026," Bloomberg Economics, December 12 Return to text
Anna Wong、Joshua Danial 和 Alex Tanzi (2025)，“17,500 次财报电话会议为 2026 年美联储大幅降息开绿灯”，彭博经济研究，12 月 12 日 返回文本
16.
16.
See James H.
参见詹姆斯·H.
Stock and Mark W.
斯托克和马克 W.
Watson (2019), "Slack and Cyclically Sensitive Inflation," NBER Working Paper Series 25987 (Cambridge, Mass.: National Bureau of Economic Research, June).
Watson (2019)，“松弛和周期性敏感的通货膨胀”，NBER 工作论文系列 25987（马萨诸塞州剑桥：国家经济研究局，6 月）。
See also the related work of the San Francisco Federal Reserve Bank of San Francisco on cyclical inflation, available at https://www.frbsf.org/research-and-insights/data-and-indicators/cyclical-and-acyclical-core-pce-inflation/ .
另请参阅旧金山联邦储备银行关于周期性通货膨胀的相关工作，请访问 https://www.frbsf.org/research-and-insights/data-and-indicators/circular-and-acyclal-core-pce-inflation/ 。
Return to text
返回正文