Hello Stata wizards

I am trying to append around ~35 excel files into 1 dta files. Luckily all the files are cleaned and mirror one another in terms of how they are structured. I figured because of this I can loop a code to append all of the files for me. All of the files are named ENR_YEAR_MONTH.xlsx so a sample coould potentially be named ENR_2017_01.xlsx.

My first step was to try my code out to get my desired result (to ensure it works):

Code:
set more off
tempfile in
import excel "ENR_2016_01.xlsx", sheet("MCHIP") cellrange(A10:K103) firstrow case(lower) clear
rename a county_name
drop anthem mdwise mhs total f g
drop mchip schip
gen medicaid_enroll_ch= county+ expansion
gen month="01"
gen year="2016"
save `in', replace
It worked out beautifully and I got the following output (don't worry, this is all public data):

Code:
* Example generated by -dataex-. To install: ssc install dataex
clear
input str15 county_name long county int expansion float medicaid_enroll_ch str2 month str4 year
"01-ADAMS       "   342   39   381 "01" "2016"
"02-ALLEN       "  5666  453  6119 "01" "2016"
"03-BARTHOLOMEW "   815   81   896 "01" "2016"
"04-BENTON      "   121   10   131 "01" "2016"
"05-BLACKFORD   "   184   16   200 "01" "2016"
"06-BOONE       "   515   50   565 "01" "2016"
"07-BROWN       "   209   20   229 "01" "2016"
"08-CARROLL     "   263   32   295 "01" "2016"
"09-CASS        "   683   36   719 "01" "2016"
"10-CLARK       "  1619  133  1752 "01" "2016"
"11-CLAY        "   405   34   439 "01" "2016"
"12-CLINTON     "   562   53   615 "01" "2016"
"13-CRAWFORD    "   113    5   118 "01" "2016"
"14-DAVIESS     "   496   61   557 "01" "2016"
"15-DEARBORN    "   471   46   517 "01" "2016"
"16-DECATUR     "   344   40   384 "01" "2016"
"17-DEKALB      "   584   48   632 "01" "2016"
"18-DELAWARE    "  1267  110  1377 "01" "2016"
"19-DUBOIS      "   367   42   409 "01" "2016"
"20-ELKHART     "  4358  398  4756 "01" "2016"
"21-FAYETTE     "   435   28   463 "01" "2016"
"22-FLOYD       "   815   74   889 "01" "2016"
"23-FOUNTAIN    "   174    4   178 "01" "2016"
"24-FRANKLIN    "   284   12   296 "01" "2016"
"25-FULTON      "   349   42   391 "01" "2016"
"26-GIBSON      "   344   25   369 "01" "2016"
"27-GRANT       "   949   78  1027 "01" "2016"
"28-GREENE      "   468   48   516 "01" "2016"
"29-HAMILTON    "  2374  283  2657 "01" "2016"
"30-HANCOCK     "   712   71   783 "01" "2016"
"31-HARRISON    "   468   54   522 "01" "2016"
"32-HENDRICKS   "  1514  184  1698 "01" "2016"
"33-HENRY       "   746   72   818 "01" "2016"
"34-HOWARD      "   908   63   971 "01" "2016"
"35-HUNTINGTON  "   456   37   493 "01" "2016"
"36-JACKSON     "   507   35   542 "01" "2016"
"37-JASPER      "   381   44   425 "01" "2016"
"38-JAY         "   334   26   360 "01" "2016"
"39-JEFFERSON   "   356   40   396 "01" "2016"
"40-JENNINGS    "   390   37   427 "01" "2016"
"41-JOHNSON     "  1891  185  2076 "01" "2016"
"42-KNOX        "   449   42   491 "01" "2016"
"43-KOSCIUSKO   "  1182  109  1291 "01" "2016"
"44-LAGRANGE    "   383   32   415 "01" "2016"
"45-LAKE        "  5608  469  6077 "01" "2016"
"46-LAPORTE     "  1339  105  1444 "01" "2016"
"47-LAWRENCE    "   745   54   799 "01" "2016"
"48-MADISON     "  1750  128  1878 "01" "2016"
"49-MARION      " 18125 1102 19227 "01" "2016"
"50-MARSHALL    "   923   74   997 "01" "2016"
"51-MARTIN      "   150    7   157 "01" "2016"
"52-MIAMI       "   464   26   490 "01" "2016"
"53-MONROE      "  1164  158  1322 "01" "2016"
"54-MONTGOMERY  "   383   34   417 "01" "2016"
"55-MORGAN      "   916   68   984 "01" "2016"
"56-NEWTON      "   171   14   185 "01" "2016"
"57-NOBLE       "   696   62   758 "01" "2016"
"58-OHIO        "    26    3    29 "01" "2016"
"59-ORANGE      "   378   33   411 "01" "2016"
"60-OWEN        "   401   42   443 "01" "2016"
"61-PARKE       "   215   20   235 "01" "2016"
"62-PERRY       "   175   17   192 "01" "2016"
"63-PIKE        "    99    5   104 "01" "2016"
"64-PORTER      "  1279  121  1400 "01" "2016"
"65-POSEY       "   202   20   222 "01" "2016"
"66-PULASKI     "   163   19   182 "01" "2016"
"67-PUTNAM      "   391   43   434 "01" "2016"
"68-RANDOLPH    "   335   29   364 "01" "2016"
"69-RIPLEY      "   307   28   335 "01" "2016"
"70-RUSH        "   240   22   262 "01" "2016"
"71-ST. JOSEPH  "  3641  283  3924 "01" "2016"
"72-SCOTT       "   276   19   295 "01" "2016"
"73-SHELBY      "   617   53   670 "01" "2016"
"74-SPENCER     "   181   14   195 "01" "2016"
"75-STARKE      "   354   27   381 "01" "2016"
"76-STEUBEN     "   451   36   487 "01" "2016"
"77-SULLIVAN    "   258   17   275 "01" "2016"
"78-SWITZERLAND "   100    6   106 "01" "2016"
"79-TIPPECANOE  "  1855  150  2005 "01" "2016"
"80-TIPTON      "   215   21   236 "01" "2016"
"81-UNION       "   114    8   122 "01" "2016"
"82-VANDERBURGH "  2059  155  2214 "01" "2016"
"83-VERMILLION  "   235   19   254 "01" "2016"
"84-VIGO        "  1352  107  1459 "01" "2016"
"85-WABASH      "   450   59   509 "01" "2016"
"86-WARREN      "   120   20   140 "01" "2016"
"87-WARRICK     "   639   70   709 "01" "2016"
"88-WASHINGTON  "   392   35   427 "01" "2016"
"89-WAYNE       "   822   67   889 "01" "2016"
"90-WELLS       "   384   40   424 "01" "2016"
"91-WHITE       "   328   25   353 "01" "2016"
"92-WHITLEY     "   370   41   411 "01" "2016"
"Total:"          88111 7307 95418 "01" "2016"
end
When I tried to create the looping code came up with the following:

Code:
forval yr = 16/19 {
forval mnth = 01/09 {
import excel "ENR_20`yr'_0`mnth’.xlsx", sheet("DISTRICT") cellrange(A10:K103) firstrow case(lower) clear
rename a county_name
drop anthem mdwise mhs total f g
drop mchip schip
gen medicaid_enroll_ch= county+ expansion 
gen month="0`mnth'"
gen year="20`yr'"
append using `in’ 
save `in’, replace
}
}
All is well but I keep getting this error, and cannot see where I am short/have added an extra "}"

too few quotes
r(132);