Inflation
Unless the government is literally printing money and handing out directly to people (which has happened in some countries), and/or supply-chain related issues are present, inflation is almost always highly correlated with wages. People need to have more money so they spend more, on an unchanging set of goods and services, which causes prices to increase.
Wages come from employers. Businesses work on credit, and credit is newly printed money. Therefore when businesses hand out salaries they are handing out new money. That could result in inflation, if there is too much credit out there, then the economy "stars to heat up" i.e. inflation.
Note that bank credit does not originate from bank customers' savings, they are printed money, so money base expands. The system works because even though the money base grows (through credit), the economy grows into that base (credit was used to invest, hire ppl etc); more products, services are circulating, but more money does not chase after the same amount of goods, it chases after more amt of goods - no inflation.
Bank credit is "printed money", brought into existence out of thin air.. Just type bunch of numbers on a computer, boom. New money. There is good reason for creating money and not simply lending out savings; If only savings were lent out, growth + same money base = deflation.
If the set of goods of services in an economy grows (more production) in tandem with the money base in that economy, there would not be inflation. Sadly keeping this due in perfect sync is an extremely tough job. The Central Bank has only one tool at its disposal to tune this - the (short term) interest rate. By raising rates during econ that is heating up, the idea is to raise borrowing costs and slow the economy. The effect is extremely indirect, the aim is companies to borrow less, to expand less, therefore pay employees less. It sounds harsh but this is the system in play.
Are government deficits inflationary? If new, more money reaches citizens directly, it could be... If new money goes to corporations, then no. Research on this area found a lagged response in inflation for government spending where it existed. The true determinant seemed to be unit labor costs.
Post 2008 bailouts provides the perfect example; money was dumped on corporations, their profits soared. But the end-result was not inflationary. People's wages were stagnating.
Take pre-1973. Some argue "government spending for the Vietnam war caused inflation".
See article here I took data from it directly, shifted source vars one year ahead to see causal effects better,
import pandas as pd, io
s = """
YEAR DEFICIT GROWTH INFLATION UNEMPLOYMENT
1965 $1B 6.5% 1.9% 4.0%
1966 $4B 6.6% 3.5% 3.8%
1967 $9B 2.7% 3.0% 3.8%
1968 $25B 4.9% 4.7% 3.4%
1969 -$3B 3.1% 6.2% 3.5%
1970 $3B 0.2% 5.6% 6.1%
1971 $23B 3.3% 3.3% 6.0%
1972 $23B 5.2% 3.4% 5.2%
1973 $15B 5.6% 8.7% 4.9%
"""
s = s.replace("%","").replace("$","").replace("B","")
df = pd.read_csv(io.StringIO(s),sep='\s*').set_index("YEAR")
df['GROWTH'] = df.GROWTH.shift(1)
df['UNEMPLOYMENT'] = df.UNEMPLOYMENT.shift(1)
df['DEFICIT'] = df.DEFICIT.shift(1)
df.corr()
DEFICIT GROWTH INFLATION UNEMPLOYMENT
DEFICIT 1.000000 0.123867 0.490311 0.264804
GROWTH 0.123867 1.000000 0.138173 -0.504883
INFLATION 0.490311 0.138173 1.000000 -0.138800
UNEMPLOYMENT 0.264804 -0.504883 -0.138800 1.000000
There is strong correlation with gov spending and inflation, but also between growth and unemployment which makes sense, wages are major contribution to inflation.
Doc mentions gov spending wasn't entirely for the war; LBJ did some social spending (money went to ppl direct).
Inflation wasn't too high pre-73. Starting 73 it was but oil shortages started then.
The draft might have played a role; >2 mil was drafted into Vietnam, 1% of population, that means more money was spent on people. Now population of >300 mil has mil active personnel less than 1.5 mil. Gov still deficit spends but if it all goes to Reytheon, Lockheed, from there to some offshore haven, then to stocks - no inflation.
In summary, more newly printed money in more hands can be inflationary. One rich person can only buy so many cars. But a million with increased wages can buy millions of products, bidding up their prices which for unchanged set of products will cause inflation.
Autocorrelation
There is another troubling aspect to inflation, by just being there, inflation can create more inflation. This is what economists mean when they say inflation time series is "autocorrelated", ie a value today is correlated with values from yesterday, signaling that value alone can effect inflation today. It is easy to see why, pricing is imperfect, producers, landlords see higher prices, they increase prices to match, they mostly overshoot. This is also why Central Banks raise rates to quash that self-reinforcing effect.
We can check for the presence of autocorrelation statistically using the Durbin-Watson test, or Ljung-Box Q-statistic. Investopedia: "The Durbin-Watson statistic will always have a value ranging between 0 and 4.. [v]alues from 0 to less than 2 point to positive autocorrelation".
from pandas_datareader.data import DataReader # get data from FRED
cpi = DataReader('CPIAUCNS', 'fred', start='1971-01', end='2016-12')
import statsmodels.formula.api as smf
from statsmodels.stats.stattools import durbin_watson
inf = np.log(cpi)
results = smf.ols('CPIAUCNS ~ 1', data=inf).fit()
print (durbin_watson(results.resid))
9.222098132626528e-05
Hints at autocorrelation
Ljung-Box
import statsmodels.tsa.stattools as tsa
acf,confint,qstat,pvalues = tsa.acf(results.resid, nlags=4, alpha=95,qstat=True, unbiased=True)
print (acf)
print (pvalues)
[1. 0.99585547 0.99164499 0.98737151 0.98305291]
[1.02286496e-121 5.63735022e-239 0.00000000e+000 0.00000000e+000]
Low pval means presence of autocorrelation.
Marginal Revolution University
Why inflate