New Zealand Government Spending and the Inflation Target

The RBNZ would be happy when inflation is at target. The nominal interest rate has been close to zero for a long time. This is called the Zero Lower Bound, where monetary policy cannot lower the interest to stimulate the economy and push inflation up to the target unless going negative. 

The Governor of the RBNZ called upon the government to spend more money. Since the interest rate is low, the RB is obviously encouraging spending. The government responded later by an investment package. 

Spending will stimulate aggregate demand and buy votes. However, If the government wants to expand demand and increase inflation to meet the inflation target of 2 percent, then the question is how much money it has to spend to achieve that?

I will show here that the government must spend a huge amount of money to achieve a 2 percent inflation.

We know from John Taylor and Valerie Ramey research that the government spending multiplier is small in the U.S. and the Obama fiscal package did not have a significant impact.

Today, I answer a more specific question. Suppose that the Zero Lower Bound constraint is binding so the nominal interest rate is zero. Hence, monetary policy is ineffective. The inflation target is 2 percent. How much the government have to spend in order to restore the inflation target?

To answer this question we have to have a theory for the natural rate of interest. I am working on this  now and I thought it would be a good idea if I do the calculations for NZ and share it with you. This is not the place to do all the math, and there is a lot of it in this research, but I will explain and hope that it would be clear.

The natural rate is an unobservable variable, Wicksell (1898. Here is a translation of the main point regarding price change[Boldface and italic is my emphasis].

“… These two rates of interest, the natural rate and the money rate, which is quoted on the market, tend of course, to coincide. If the former differs from the latter, money can no longer be said to be “neutral,” and monetary consequences in the shape of change in prices are bound to ensue. If the money rate were kept below the natural rate prices would rise, if above they would fall.”

So that means if r* is the natural rate of interest, and i is the nominal rate, r* - i = inflation.

The story goes like this. Let the economy be made of three optimizing agents, a household, a firm, and a government. The household maximizes a utility function subject to a budget constraint. The utility function has two arguments: a consumption good and leisure. The household makes a decision about saving-consumption, and consumption-leisure. The household holds bonds and stocks, and owns the capital stock, which is rented out to a firm that uses it along with labor to produce output. The household pays taxes on consumption, on investments, on income from capital, and on labor income. All tax revenues, except those used to finance pure public consumption good (e.g., education, defense, etc.) are given back to the household in the form of lump sum transfers from the government. This is a simple model that covers the basics.

The solution of this simple model results in a parsimonious equation for the natural rate of interest, which could be easily computed (no estimation) using observable raw data. The natural rate r* would be zero when consumption grows at the same rate of leisure and capital grows at the same rate of labor. Once these gaps open up, r* changes. The natural rate depends positively on the growth rate of consumption,  negatively on the growth rate of leisure, negatively on the growth rate of the stock of capital, and positively on the growth rate of labor.

Note that there is a great degree of uncertainty about the value of r* because people have different ways of modelling Wicksell's idea. Thus, r* is model-dependent. Therefore, what I have may well be very different from what the RB has in mind, and both of us are different from what Wicksell had in mind.

Here are Some measurements.

Consumption is private consumption plus government consumption less military spending less indirect taxes on consumption. Military spending is trivial in NZ. 

Labor is measured by working age population (15-64 years). I assume that the household has 100 hours a week available for work in the market to make money and pay taxes. Therefore, leisure is 100 minus the average weekly hours worked. For example, if a household average weekly hours worked are 30 hours, leisure time would be 70 hours. That is all we need to measure r*, but I am asking about the projection of r*, the future.

I look at the period 2018 to 2024 as the projection horizon because the IMF  has this horizon and some data are taken from the IMF - World Economic Outlook. I assume that consumption follows a random walk(Robert Hall). 

I assume that military spending and the indirect taxes on consumption are unchanged from 2018 to 2024, which is also reasonable. 

Capital stock evolves according to a typical Perpetual Inventory equation, (1-depreciation rate)*last period capital stock  plus investment.  The stock of capital and the depreciation rate are published by the World Penn Table 9.1. The investment forecasts are reported by the IMF, except that they are nominal percentages of current GDP. Given current GDP, we compute the level and deflate it by the GDP implicit price deflator to obtain real investments.

Working age population is from the OECD population projections. Leisure requires measuring average weekly hours worked. This is hours worked by employed people from the OECD statistics times (multiplied by) employment/working age population ratio. Then we compute 100- average weekly hours worked. 

I solve the model for the level of government sending needed to make r*=2, which is needed to achieve the 2 percent inflation target when nominal interest rate is zero. 

The average growth rate of consumption over the period 2018 - 2019 is projected to be 9 percent. It is high because it includes the increase in government spending that we need to achieve the inflation target. The stock of capital average growth rate is projected to be 7.8 percent, also high. Working age population growth rate average is 0.8, and leisure average growth rate is zero because there is no change in future average weekly hours worked. Hence,r* is 2. To get to this, government spending must increase by 20% on average over the period 2018 t 2024. This is a relatively very high growth rate knowing that the average growth rate between 2000 and 2017 was 3% only. I conclude that it is very costly to use fiscal policy to achieve the inflation target.

  

  

Is The Unemployment Rate in New Zealand Really Low?

Statistics NZ published the latest unemployment rate; it is 3.9 percent (June 2019). Excellent. 

However, higher for females, 4.2 percent. Male unemployment rate is 3.6 percent.

So even in New Zealand, which is a very progressive liberal country, the unemployment rate is higher for women; it always has been.

Female wage rate is also lower than males. Male and female get paid the same, on average, up to age 30 or so, then males get paid more. Pay equalized again later in life. I have not seen much research on this in NZ. It could be the effect of child-rearing or something else, i.e., some might say discrimination. 

That's not the whole story.

Minorities such as Maori, Pacific People, and Asians have higher unemployment rates than Europeans. Here is the full picture.

The rates in 2019 are 3.4, 8.7, 8.2 and 4.2 percent for Europeans, Maori, Pacific People, and Asians respectively. Much higher than 3.9 percent! Maori and Pacific People unemployment rates are more than double the rate of the Europeans?

The averages over the past 10 years, are worse: 4.2, 11.6, 11.4 and 6.3 percent for Europeans, Maori, Pacific People, and Asians respectively.

Maoris and Pacific People participation rate is not significantly different from the rest.

The participation rate trend for these groups is constant. However, participation rates have been always, slightly, lower for the minorities than the Europeans. Really, if Maori and Pacific People are not looking for jobs, their unemployment rate might even be higher than the posted numbers.

So they look for jobs and don't get any, how do they live?

The Ministry of Social Development's Benefit Fact Sheet for 2019 shows that the total recipients of benefits, as a percent of working age population, are 37.9 percent Europeans, 36.9 percent are Maori, and 8.2 percent are Pacific People!

On average over the past 5 years, Europeans are 39.9 and Maori and 35.5 percent. Pacific Islanders are receiving much less benefits. Puzzling. Again, how do they get by?

Remember that Maoris make up less than 17 percent of the total population but they receive just as much benefit as the Europeans!Pacific People make up about 7.5 percent of the population. Problem.

Maori and Pacific People unemployment rates have been higher than 8 percent for decades. These are problems?

New Zealand is not a high-tech country really. We are a farming,  commodity-exporting, and services economy. Does it make sense that Maori and Pacific People cannot find jobs in such economy? How come their unemployment rates are persistently high?

On the other hand, work is a European concept, hence the importance of the statistics about the labor market such as unemployment and participation rates. The Polynesian culture has been a hunting and fishing then an agricultural culture. Maybe Polynesians do not enjoy a 9-5 type of work in an office, a warehouse, bank, factory or hospital...etc.Could that be part of the story? I don't know. Is there research on these issue? I could not find any.

Using a European yard-stick to measure their labor market performance might be useless. Of course I am not talking about all people of such heritage; I am talking in general. Either way one need to know what is going on with the high unemployment rate. 

          

RBNZ Bank Capital Story

The RBNZ "bank capital requirement" review has become contentious. Yesterday, Michael Reddell (https://croakingcassandra.com/) showed three convincing graphs to debunk the RBNZ claim that the main reason for wanting to raise bank capital requirements is that our macro volatility is high.

Grant Spencer spoke about the process to review bank capital in October 2017. You could find his speech online. It is a nice speech. Thus, this is not something the new Governor invented for sure. The RBNZ thought about it earlier.

Here, I want to add one more graph to Reddell's three graphs to show that New Zealand's macro volatility is not significantly different from Australia's, not at all.

I decompose the real interest rate differentials r-r*, where r is New Zealand's real interest rate and r*is the U.S. real interest rate to Country Risk and a Currency Risk.

r-r* = (i - i*-f)  + [(f-ds) + dq]

The first term is the Country Risk; the second and the third are the Currency Risk.

i is New Zealand's 3-month nominal interest rate; i* is the U.S. 3-month T-Bill interest rate; and f is the forward exchange rate of the NZD-USD. ds is the nominal exchange rate depreciation rate, which is zero under perfect foresight. dq is the expected real exchange rate depreciation rate. I do not have actual data for f  so I proxy it by s(t) - s(t-12). I do the same for Australia and focus on the Country Risk - the first term in the equation because the currency risk is almost the same for both the Kiwi dollar and the AUD.

The country risk for Australia and New Zealand are almost identical. The Standard Deviations are 11.4 for Australia and 11.9 for New Zealand.

There is no convincing macro indicator to use as a reason to raise bank capital. 

Maybe the RBNZ has other reasons, not macro ones.

The IMF newly published World Economic Outlook increased our growth rate forecast. Our banks withstood the massive Global Financial Crisis. Reddell also mentioned that we do not even have data for banking crisis that occurred once in 200 years.

The RBNZ must come up with a more reasonable story for its capital level review.

The RBNZ's OCR Cut

Business NZ criticized the RBNZ for cutting the OCR by 50 basis point recently. It said that the current economic conditions do not warrant such a large cut.

The governor of the RBNZ issues a statement saying that the RB makes forward looking policy decisions, i.e., it is based on forecasts of domestic and global economic conditions, and not on current conditions.

Makes sense to me. Obviously, monetary policy affects the economy with "variable and long lags," i.e., it takes time for the OCR cut to work through.

That being said, there is an issue with the RBNZ OCR cut. It assumes that its forecast is reliable. No forecast is reliable. A forecast error could translate into a policy error. Policy errors are persistent. They are costly to undo. In this case, large policy move is riskier than small moves.

Economists usually argued for interest rate smoothing. A large shock, if indeed it is large, which requires a 50 basis points cut in OCR could have been smoothed out, 25 followed by 25.

Shocks are hard to identify ex-post let alone predict ex-ante.

But let's examine the RBNZ forecasts, which is posted in the MPS, chapter 4, pp. 26 - 35.

They do not indicate, to me at least, that there are problems that warrant a 50 basis point, surprise, cut in OCR!

Begin from the bottom page, forecasts up to 2020 show that, tradeables inflation is low - near zero; CPI inflation returning to the target; non-tradebales inflation is forecast to "dip" slightly. It looks like it increasing before dipping. This is basically house price inflation, asset price inflation. With lower OCR, house price inflation should increase!

Wage inflation is expected to increase!


The output gap is positive but they have a small decline in 2019!

These forecasts do not indicate a need for this large cut in OCR.

Employment does not look like falling. They say it is near maximum.

Unemployment increases very little over the forecasting horizon. These forecasts do not seem to warrant a 50 basis points cut in interest rate either.

Business investments up over the forecasting horizon, but the forecast is revised downward relative to last MPS forecast.

Fiscal stimulus supports a lift in GDP growth. Government consumption is up over the forecasting period and higher than that made in May MPS.

Households consumption is higher than the previous forecast, albeit weak, they said, 2 percent growth on average perhaps!

Residential investments do not indicate weakness, rather strong. House price inflation is up.

Overall, real GDP is projected to be "subdued" .

Net immigration remains elevated but is expected to decline.

Demand for New Zealand’s exports is slowing.

Import price expected to be lower.

Net immigration remains elevated but is expected to decline

They say that global conditions have continued to weaken, but I wonder how good is the RB forecast of global conditions.
 
I do not see anything alarming in these projections really. Therefore, the 50 basis point cut remains controversial.


 

Price Stability and Inflation in New Zealand

Many people ask why prices increase year after year while the Reserve Bank says the inflation rate is low. I have heard economists who cast doubt about the merit of inflation targeting citing rising prices as a reason!

I bet that if we conduct a survey we would be surprised to know that most people think the same way.

A layman’s explanation first

This is just an example using arbitrary numbers. Table (1) has four columns. Column 1 has four years, 1 to 4. There is a price level in the second column and the percentage change of the price level, i.e. the inflation rate in the last column.

Table 1

So in this example the price increased every year from 100 to 102 to 104... The inflation rate, however, remained stable, 2 percent every year. This is what an inflation-targeting central bank desires to achieve. 

Inflation-targeting central banks treat bygones as bygones. Essentially, the central bank ignores the effects of the various shocks that increase the price level and focuses on maintaining the inflation rate target period by period. The price level is not the concern of the inflation-targeting central bank.Under inflation targeting, the price level will drift up forever. That is why prices keep increasing every year and the inflation rate does not.

The sketch below describes it in a different way. A shock at period 1 increases the price from p0 to p1. The price stays there, but the inflation rate, which is the slope (i.e. the change in the price over time), remained unchanged.

Academic talk - not for the average reader 

We could go further and use academic language for those who are interested. I tested the data for NZ as everyone in this field does routinely. The price level measured by the CPI has a unit root; inflation does not.

Figure (2) 

Thus, the forecast of the price level is meaningless because the forecast error variance grows to infinity as the forecast horizon increases. The mean, variance...of the price level are all functions of time.These are the results of targeting the inflation rate. The RB could target the price level if they want to. I have shown that (with Eric Hansen) in 1995 in an RB conference. No takers. Many scholars have written about price level targeting. Still No takers.

As long as the public demand for money (the stock of money, e.g., M1)keeps increasing, which is the case in NZ, the price level will  increase in proportion because people spend the money (could be spent on assets and housing too and increase asset prices). Money and prices are highly correlated. Inflation and money growth, however, are not correlated at all because the RBNZ rendered inflation stationary, but not money.I studied this issue in 2001 when I was in the RB https://www.rbnz.govt.nz/research-and-publications/discussion-papers/2001/dp2001-02-2 and I don't think anything has changed since. 

If the CPI inflation is under control then What kind of inflation do we have? 

Average annual money growth over the past since 1989 is about 7 percent. Lending to housing plus personal lending average growth rate over the period from Dec 1990 is 9.6 percent. People spend the money. They buy things, but they also buy assets and housing. The CPI inflation does not measure that. Hence, inflation is absent. 

We have much higher housing price inflation than CPI inflation. Figure (3) plots the two inflation rates. Over the same period from Dec 1990 to Dec 2018, the average CPI inflation was 2 percent while housing price inflation, HPI, was 6.2 percent! 

Figure (3)

Low interest rate and the next recession

Recently, more people wonder how the central banks would deal with the next recession given that the interest rate is too low.

Central bankers, and many economists, believe that the short-term nominal interest rate is the main or only policy instrument. However, this may not be entirely true. Central banks could use the money supply, by they never did.

Not many central bankers care about money anymore when setting up policy. Certainly, money does not even play a role in their models. I remember Stanley Fischer giving a lecture in Wellington about 20 years ago on the Russian monetary policy after the collapse of the USSR. I wrote in my notes that he said that the central bank brought down inflation by "controlling money and credit."

People may worry about recessions, but they are impossible to predict. Shocks are random. The propagation mechanism of these shocks - the data generating process - is unknown. Even identifying the shocks ex-post is very hard.

The question though is, what happens to real GDP growth (or the output gap) if the short-term nominal interest rate is literally zero? (In fact Milton Friedman argued that the short-term nominal interest rate should be set to zero such that the marginal cost of producing money is equal to the price.)

Consider this simple counterfactual.

In an open-economy - Keynesian - IS curve (i.e., the good market schedule), real GDP growth could depend on its own past (persistence), real interest rate, real exchange rate, and foreign real GDP growth. To carry out the counterfactual, replace the real interest rate with the negative of expected inflation (lag inflation for simplicity) if the nominal interest rate is zero (the Fisher equation). One could choose / estimate the coefficients and calibrate this IS equation with a random normal errors. One could even use a more elaborate Dynamic Stochastic General Equilibrium model to do this counterfactual experiment. Here is one simple counterfactual for New Zealand under such scenario.

The counterfactual suggests that:

-  Real GDP growth could have been higher the actual on average, 4.1 percent compared to 2.6 percent.

-  Real GDP growth could have been less volatile than actual. The variance is 3.1 compared to 4.5

-  The peaks could have been higher and the troughs could have been lower.

-  No recessions detected in the counterfactual.

Genes, and early childhood interventions

A long time ago I wrote several technical notes about early childhood intervention to the deputy and the secretary of labour when there was still "a Department of Labour" . I am almost certain they did not read any of these notes. James Heckman, Nobel Laureate, econometricain at Chicago wrote extensively on this subject and showed the efficacy of early childhood intervention. Intervention would save the country a lot, and reduce private and social cost over time. I argued that it is the right way to go for New Zealand.

I have been reading a very informative book, which I highly recommend because it has very well written interesting stories. The book is The Gene, An Intimate History, by Siddhartha Mukherjee who is an American cancer physician and researcher at Columbia University. My daughter, who is a University of Cambridge PhD in Chemistry and a Chief Editor at Nature Science in the UK, and I had an argument on the dinner table about the gene and gene-environment effects on student's ability (ability is an unobservable variable) so she recommended that I read this book.

Here is something from page 459, which reminded my of the notes that I wrote back then at the DoL.

He says, "In 2010, a team of researchers launched a research study, called the Strong African American Families Project, or SAAF, in an impoverished rural belt n Georgia. It is a startlingly bleak place overrun by delinquency, alcoholism, violence, mental illness, and drug use. Abandoned clapboard houses with broken windows dot the landscape; crime abounds; vacant parking lots are strewn with hypodermic needles. Half of the adults lack high school education, and nearly half the families have single mothers.

Six hundred African-American families were randomly assigned to two groups. In one group, the children and their parents received seven weeks of intensive education, counseling, emotional support, and structured social interventions focused on preventing alcoholism, binge behaviors, violence, impulsiveness, and drug use. In the control group, the families received minimal  interventions. Children in the intervention group and in the control group had the 5HTTLPR gene sequenced (It encodes a molecule that modulated signaling between certain neurons in the brain.It was found that it is associated with response of psychic stress. The gene comes in two forms or alleles - a long variant and a short variant. The short variant called 5HTTLPR/short and it is carried by about 40 percent of the population and seems to produce significantly lower levels of the protein. The short variant has been repeatedly associated with anxious behavior, depression, trauma, alcoholism, and high-risk behaviors. The link is not strong, butbit is broad: the shorter allele has been associated with increased suicidal risk among German alcoholics, increased depression in American college students, and higher rate of PTSD among deployed soldiers).

The first results of this randomized trail was predictable from prior studies: in the control group, children with short variant, i.e. the high risk form of the gene - were twice as likely to veer toward high-risk behaviors, including binge drinking, drug use, and sexual promiscuity as adolescences, confirming earlier studies that had suggested an increased risk within this genetic subgroup. The second result was more provocative: these very children were also the most likely to respond to the social interventions. In the intervention group , children with high-risk allele were most strongly and rapidly "normalized"  - i.e., the most drastically affected subjects were also the best respondents. In a parallel study, orphaned infants with the short variant of 5HTTLRP appeared more impulsive and socially disturbed than their long-variant counterparts at baseline - but were also the most likely to benefit from placement in a more nurturing foster-care environment."                 

Enjoy reading the book